JP3639030B2 - Image display system and image display method using the system - Google Patents

Description

なお、ここでは指定された断層像対はｓ対であり、ｚ_Ａ ^ｉとｚ_Ｂ ^ｉはｉ番目（ｉ＝１、２、…、ｓ）の断層像対のｚ座標とする。誤差Ｅを最小にするために、「ΔＥ／Δｚ_ｅ＝０」として（２）式を解くと、次式が得られる。
【０１０１】
【数３】

【０１０２】
したがって、１対のときと同様に、ｚ_Ｂの任意の座標はｚ_Ａ＋ｚ_ｅによりｚ_Ａの座標に変換される。この相対的なズレ量ｚ_ｅとその対象となる三次元画像（断層像Ａと断層像Ｂ）はすべて制御装置３内の主メモリ３ｂ内に記憶される（ステップ１１４）。そして、制御用ＣＰＵ３ａの処理はステップ１１５へ移行する。
【０１０３】
制御用ＣＰＵ３ａは、上述した処理を三次元画像「Ｂ，Ｃ」間、三次元画像「Ｃ，Ｄ間」…にも繰り返し行ない、すべての三次元画像「Ａ〜Ｆ」の座標の相対的な関係を間接的に関連付ける。なお、この「間接的に」とは、例えば三次元画像Ａのｚ座標と三次元画像Ｂのｚ座標、三次元画像Ｂのｚ座標と三次元画像Ｃのｚ座標、三次元画像Ｃのｚ座標と三次元画像Ｄのｚ座標、…、となって最後に三次元画像Ｆのｚ座標と三次元画像Ａのｚ座標と連鎖状に関連づけられることや、三次元画像Ａのｚ座標と三次元画像Ｂのｚ座標、三次元画像Ａのｚ座標と三次元画像Ｃのｚ座標、三次元画像Ｃのｚ座標と三次元画像Ｄのｚ座標、…のように、三次元画像Ａ〜三次元画像Ｆが互いに（相対的に）関連付けられることを意味している。なお、このとき、制御用ＣＰＵ３ａは、各三次元画像「Ａ〜Ｆ」の中で関連付やすい三次元画像のペアを当該三次元画像「Ａ〜Ｆ」のそれぞれのｚ座標に基づいて選択することもできる（ステップ１１５）。
【０１０４】
そして、制御用ＣＰＵ３ａは、すべての三次元画像「Ａ〜Ｆ」の座標の相対的な関係が間接的に関連付けられているか否かを判断する（ステップ１１６）。
【０１０５】
三次元画像「Ａ〜Ｆ」の座標が相対的に関連付けられていない場合、ステップ１１６の判断はＮＯとなり、上述したステップ１１５に戻り、上述した処理を繰り返す。
【０１０６】
このステップ１１６の判断の結果ＹＥＳの場合には、すべての三次元画像「Ａ〜Ｆ」のｚ座標が連鎖状に関連づけられているため、ステップ１１７の処理に進む。
【０１０７】
ステップ１１７の処理では、制御用ＣＰＵ３ａは、主メモリ３ｂに記憶された各三次元画像「Ａ〜Ｆ」のｚ座標間のズレ量に基づいて、このズレ量を基準三次元画像Ａを基準にしたズレ量に算出し直す。例えば、基準三次元画像Ａに対する三次元画像Ｂのズレ量をｇＡＢ、三次元画像Ｂに対する三次元画像Ｃのズレ量をｇ_ＢＣ、三次元画像Ｃに対する三次元画像Ｄのズレ量をｇ_ＣＤ、三次元画像Ｄに対する三次元画像Ｅのズレ量をｇ_ＤＥ、三次元画像Ｅに対する三次元画像Ｆのズレ量をｇＥＦとすると、基準三次元画像Ａの任意の位置（ｚ座標）から三次元画像Ｂの対応する位置を求める式は、「ｚ_Ａ＋（ｇ_ＡＢ）」となる。同様に、基準三次元画像Ａの任意の位置から三次元画像Ｃ，Ｄ，Ｅ，Ｆの対応する位置を求める式は、それぞれ、「ｚ_Ａ＋（ｇ_ＡＢ＋ｇ_ＢＣ）」、「ｚ_Ａ＋（ｇ_ＡＢ＋ｇ_ＢＣ＋ｇ_ＣＤ）」、「ｚ_Ａ＋（ｇ_ＡＢ＋ｇ_ＢＣ＋ｇ_ＣＤ＋ｇ_ＤＥ）」、「ｚ_Ａ＋（ｇ_ＡＢ＋ｇ_ＢＣ＋ｇ_ＣＤ＋ｇ_ＤＥ＋ｇ_ＥＦ）」となる。このような各相対関係から、全体の関係を求めることは広く一般的に行なわれている。また、コンピュータ上で実現するアルゴリズムも数多く提案されている。
【０１０８】
このようにして求められた基準三次元画像Ａの座標に対する各三次元画像「Ｂ〜Ｆ」の座標のズレ量は、対応する処理装置４ａ1〜４ａ6のメモリ４ｃ1〜４ｃ6に記憶される。基準座標に対応する各座標のズレ量は、前記の例では、括弧（）内のものになる。つまり、基準三次元画像の基準となる座標に対してこのズレ量を加味することにより、各三次元画像間「Ａ〜Ｆ」の位置合わせを行なうことができる（ステップ１１７）。
【０１０９】
続いて処理は図６のステップ１１８の処理に移行し、制御用ＣＰＵ３ａは、任意の基準となるスライス位置、すなわち、そのスライス位置に対応するｚ座標の断層像対を同期させて表示する旨を表す表示指令を各処理装置４ａ1〜４ａ6に送る。なお、このスライス位置は、予め定められていてもよいし、入力部３ｃから別個に指定されたものであってもよい。また、基準断層像Ａの最初のスライス位置としてもよい（ステップ１１８）。処理装置４ａ1〜４ａ6のＣＰＵ４ｂ1〜４ｂ6は、入力されたスライス位置、及びメモリ４ｃ1〜４ｃ6に記憶されたズレ量に基づいて各三次元画像間の位置合わせを行ない、それぞれ基準となるｚ座標（Ｚａ１、Ｚｂ１、…、Ｚｆ１）を算出する（ステップ１１９）。
【０１１０】
そして、各ＣＰＵ４ｂ1〜４ｂkは、算出されたｚ座標位置（Ｚａ１、Ｚｂ１、…、Ｚｆ１）に断層像データが存在するか否かを判断する（ステップ１２０）。この判断の結果ＮＯ、つまり断層像データが存在しなければ、この算出されたｚ位置（Ｚａ１、Ｚｂ１、…、Ｚｆ１）の断層像データを求める。
【０１１１】
この場合の考え方を図９を参照して説明する。図９（ａ），（ｂ）に示すように、三次元画像は、ピクセルを固定してｚ方向にピクセル値をプロットすると、一定間隔ｚd毎に値が記録されていくことになる。つまり、標本化されている状態で、標本点間の値を推定する問題となる。したがって、後述する文献（１）に示される補間処理を行なうことにより、あるピクセル位置の値を求めることができる。この処理を全てのピクセルに行なうことにより、断層像データが無いｚ座標位置での断層像データを新たに作成することができる。なお、図９（ｂ）の例では、補間手法として１次（線形）補間を用いた場合を紹介している。
【０１１２】
すなわち、ステップ１１４では、各処理装置のＣＰＵ４ｂ1〜４ｂ6は、算出されたｚ座標位置（Ｚａ１、Ｚｂ１、…、Ｚｆ１）の前後の断層像データ（図９（ｂ）では、ｚ0、ｚ0＋ｚd、ｚ0＋２ｚd、…、ｚ0＋（ｎ）ｚdの各データを用いた補間処理を行なって、算出されたｚ座標（Ｚａ１、Ｚｂ１、…、Ｚｆ１）に対応する新たな断層像データをそれぞれ作成し、画像記憶部４ｃ1〜４ｃ6に記憶する（ステップ１２１）。
【０１１３】
一方、ステップ１２０の判断の結果ＹＥＳであれば、ステップ１２２の処理に移行する。
【０１１４】
そして、各処理装置４ａ1〜４ａ6のＣＰＵ４ｂ1〜４ｂ6は、画像記憶部４ｃ1〜４ｃ6に記憶されている、基準ｚ座標（Ｚａ１、Ｚｂ１、…、Ｚｆ１）に対応する断層像データ「Ａ（Ｚａ１）、Ｂ（Ｚｂ１）、…、Ｆ（Ｚｆ１）」、すなわち略同一の解剖学的断層位置の断層像データ対「Ａ（Ｚａ１）、Ｂ（Ｚｂ１）、…、Ｆ（Ｚｆ１）」をそれぞれ読み出し、メモリ４ｃ1〜４ｃ6を参照しながら第１の表示条件である表示画素（ピクセル）サイズ、表示する断層像の間隔、ＷＷ（ウインドウ幅）、ＷＬ（ウインドウレベル）、座標系の被検体の絶対座標系に対する回転角、フィルタの種類等の設定値にしたがって画像処理を施す。なお、ピクセルサイズを一致させることは、図１０に示すように、同じ大きさの物体を画像上でも同じ大きさにすることに相当する。つまり、ある表示態様の基準となる三次元画像「Ａ」にすべての三次元画像「Ｂ〜Ｆ」を合わせることにより、作成される全ての三次元画像「Ａ〜Ｆ」が同一のピクセルサイズとなる。また、座標系を基準断層像Ａに一致させることは、図１１に示すように、同一構造を同一の角度から観察出来ることに相当する。つまり、ある基準三次元画像Ａの座標系に全ての三次元画像「Ｂ〜Ｆ」の座標系を合わせるべく、データの回転処理等を施しているわけである。但し、この図１１では、断層面内での回転しか行なっていないが、断層面の傾きを変えるような三次元的な回転も同様に処理できる。これらの処理を合わせて行なうと、ピクセル毎の比較までもが可能となる。ピクセルサイズや座標系を基準データに合わせる方法は、文献（１）等により広く知られている座標変換技術（回転・拡大・縮小）と補間技術とを組み合わせて実現できる。これらを組み合わせた例は、文献（２）に触れられている。また、同様の処理を行なう幾何学的歪に対する補正については、文献（３）に示されており、広く一般で使われている。
【０１１５】
なお、座標系の基準データに対する回転角情報だけでは、座標系の角度合わせが不十分な場合、または対象とする三次元画像に被検体の絶対座標系に対する回転角の情報が無い場合には、データ収集時の情報（記録や記憶等）や解剖学的知識を利用して、座標系の傾斜角を推定することもできる（ステップ１２２）。
【０１１６】
そして、各ＣＰＵ４ｂ1〜４ｂ6は、画像処理が施された断層像データ対「Ａ（Ｚａ１）、Ｂ（Ｚｂ１）、…、Ｆ（Ｚｆ１）」を割り当てられた表示装置５ａ1〜５ａ6に送る（ステップ１２３）。各表示装置５ａ1〜５ａ6は、送られた断層像データ「Ａ（Ｚａ１）、Ｂ（Ｚｂ１）、…、Ｆ（Ｚｆ１）」に対し第２の表示条件であるγ特性変換等の画像処理やブライトネス、コントラスト等の設定を行なった後、その断層像「Ａ（Ｚａ１）、Ｂ（Ｚｂ１）、…、Ｆ（Ｚｆ１）」をモニタＭ1〜Ｍ6に同一のタイミングで表示する（ステップ１２４）。
【０１１７】
このとき、操作者は、入力部３ｃのキーボードＫ、あるいは画像指定部Ｉ2の次断層像表示指令スイッチＮ（または、トラックボールＴ）を操作して次断層像対を表示する指令を入力する（シングルアクション）か、あるいは、入力部３ｃのキーボードＫ等から、動画のように予め定められたタイミングに基づいて、表示された断層像対から所定間隔Δｄ離れた次の断層像対を同期させて順次表示させる指令（シングルコマンド）を入力する。制御用ＣＰＵ３ａは、入力されたシングルアクション、あるいはシングルコマンドに応じて、表示された断層像対「Ａ（Ｚａ１）、Ｂ（Ｚｂ１）、…、Ｆ（Ｚｆ１）」から所定間隔離れた次の断層像対を同期させて表示させる指令を各処理装置４ａ1〜４ａ6のＣＰＵ４ｂ1〜４ｂ6に送る（ステップ１２５）。なお、この所定間隔は、この場合、基準断層像Ａの断層間隔Δｄである。なお、所定間隔Δｄは読影中に変更することができる。これは、注目する領域を細かく観察することにより、正確な診断を行なうことができると同時に、あまり、関心のない領域（例えば、疾病を起している可能性の低い部位）は粗く観察することにより、読影時間を節約することができる。
【０１１８】
各ＣＰＵ４ｂ1〜４ｂkは、次断層像対表示指令を受け取ると、現在表示画像から所定間隔Δｄ離れた位置のｚ座標（Ｚａ２、Ｚｂ２、…、Ｚｆ２）に基づいて断層像データ「Ａ（Ｚａ２）、Ｂ（Ｚｂ２）、…、Ｆ（Ｚｆ２）」を画像メモリ４ｅ1〜４ｅkから同一のタイミングで読み出す。もし、その位置に断層像データが無い場合は、ステップ１２１，ステップ１２２と同様の処理を行ない、断層像データを作成する）。この読み出された断層像データ「Ａ（Ｚａ２）、Ｂ（Ｚｂ２）、…、Ｆ（Ｚｆ２）」は、ステップ１２３、１２４と同様の処理が施された後、モニタＭ1〜Ｍ6に同一のタイミングで表示される（ステップ１２６）。
【０１１９】
そして、制御用ＣＰＵ３ａは、このまま所定間隔Δｄ毎に断層像対表示を行なうか否かを判断する（ステップ１２７）。今、入力部３ｃから断層像対表示を終了させたい旨を表す指令が入力されている場合、この判断の結果はＮＯとなり、処理を終了する。また、入力部３ｃから次断層像対表示指令が入力されている場合、あるいはシングルコマンドモードの場合、制御用ＣＰＵ３ａの処理はステップ１１８の実行前に移行し、以下、上述した処理が繰り返されるため、所定間隔Δｄ毎に順次断層像対の表示が行なわれる。
【０１２０】
例えば、シングルコマンドモード、つまり動画像表示モードの場合では、各モニタＭ1〜Ｍ6には、略同一の解剖学的位置の断層像がそれぞれ同期して順次表示される。
【０１２１】
そして、入力部３ｃからの断層像対表示終了指令、又は断層像対を形成する断層像データが無くなったときは、ステップ１２７の判断の結果はＮＯとなり、処理を終了する。
【０１２２】
以上述べたように、本実施例によれば、最初に読影者が異なる組の三次元画像の相互間で解剖学的断層位置が略同一の断層像対をいくつか指定すれば、複数の３次元画像に基づき異なる組の三次元画像の相互間で解剖学的断層位置が略同一の断層像対が自動的に導出（設定）され、さらに、その断層像対が所定のタイミングで同期して表示される。したがって、読影者は、比較読影を行なう際の断層像対を導出する、つまり、解剖学的断層位置を合わせるという非常に困難な操作を簡単な操作により行なうことができる。この結果、読影者の負担が大幅に軽減するとともに、比較読影を非常に簡単に行なうことができるようになる。
【０１２３】
また、複数の断層像対を所定間隔毎に順次表示することもできる。つまり、従来、複数の断層像対を順次表示する場合には、その複数の断層像対間の断層間隔の違いによる解剖学的断層位置の位置合わせが必要であったが、本実施例ではその位置合わせを行なう必要がなくなるため、比較読影を非常に簡単に、しかもスピーディーに行なうことができる。
【０１２４】
さらに、同期して表示される断層像対は、すべて同一の表示態様（ＷＷ、ＷＬ、ルックアップテーブル、及びフィルタの種類等が同一）、同一のピクセルサイズ、及び同一の座標系で表示されているため、読影者が比較読影する際に、余計な設定を行なう必要がなくなり、読影者の労力を大幅に低減させることができる。
【０１２５】
さらにまた、複数のモニタを階層構造で配設している場合に、同期して表示される断層像対は、それぞれ隣接した位置のモニタに自動的に表示される。つまり、読影者が余計な操作をすることなしに、断層像対は比較しやすい位置（モニタ）に表示されるため、読影者の負担を軽減させることができる。もちろん、モニタの構造は、階層構造としなくてもよく、例えば、横１列や縦１列に並んでいてもよいし、１つの装置に複数のモニタが含まれていないで、個々のモニタが並んでいてもよい。
【０１２６】
なお、本実施例では、本発明の要旨を変更しない上で種々の変形が可能である。
【０１２７】
・変形例１
図５のステップ１０１の処理おいて複数の三次元画像データを読み込んだが、この読み込むデータは三次元画像データではなく、生データ（投影データ）群であってもよい（ＣＴ装置１により取得された生データ群を直接読み込む）。この場合は、ステップ１０２の処理において、ＣＰＵ４ｂ1〜４ｂkは、送られてきた生データ群を再構成処理し、この処理の結果得られた複数の三次元画像データをモニタＭ1〜Ｍkに表示させることになる。この場合、図６のステップ１２１及びステップ１２６の画像補間処理においては、その再構成処理された複数の三次元画像データを用いてもよい。また、Ｘ線ＣＴ装置１がヘリカルスキャン（スパイラルスキャン、螺旋状スキャンとも呼ばれる）ＣＴやコーンビームＣＴのような三次元ＣＴである場合、つまり、生データを取得したモダリティが三次元的に画像再構成を行なうことができる情報を有している場合には、目的の位置（ｚ座標）の三次元画像をその生データから直接再構成することができる。
【０１２８】
・変形例２
比較読影を行なうこと、及びその比較対象が予めはっきりしていれば、図５のステップ１０６の処理を予め行なった後、図６のステップ１２２、ステップ１２３の処理を最初の断層像を表示する段階で行なってもよい。この場合、読み込まれる複数の三次元画像データは、全て比較対象であると判断されるため、ステップ１２２、ステップ１２３の処理が自動的に施される。
【０１２９】
・変形例３
図１及び図２における表示装置５ａ1〜５ａnのモニタＭ1〜Ｍnは、ディスプレイでもウインドウでもよい。また、表示装置５ａ1〜５ａnは、１つあるいは数個のディスプレイまたはウインドウ上に複数の画像を表示（多コマ表示）するシステムであってもよい。この場合、断層像対を形成する各断層像は、１コマ（１枚の画像）を表示する領域にそれぞれ表示される。さらに、表示装置５ａ1〜５ａnは、多コマ表示（プリント）が可能なハードコピー装置を備えていてもよい。この場合、断層像対を形成する各断層像は、１コマ（１枚の画像）をプリントする領域にそれぞれ表示（プリント）される。
【０１３０】
・変形例４
本実施例において、表示装置５ａ1〜５ａnの数が同期表示される断層像対の数より多ければ、制御装置３ａが各処理装置４ａ1〜４ａmに複数個の表示装置５ａ1〜５ａnを割り当てることにより、１検査画像群（三次元画像）当たり複数の断層像を表示してもよい。
【０１３１】
例えば、断層像対の数が２つ（三次元画像「Ａ，Ｂ」）とすると、図１２に示すように、各表示装置５ａ1〜５ａ6（図１２では６個として説明する）のモニタＭ1〜Ｍ6の内、表示装置５ａ1〜５ａ3（Ｍ1〜Ｍ3）は、処理装置４ａi（三次元画像Ａ）に割り当てられ、表示装置５ａ4〜５ａ6（Ｍ4〜Ｍ6）は、処理装置４ａj（三次元画像Ｂ）に割り当てられる。この例では、図１２からも分かるように、三次元画像中の複数の断層像を同時に比較することができるため、平面的な比較だけでなく立体的な比較が可能となり、比較の精度を向上させることができる。
【０１３２】
・変形例５
変形例３において、各モニタＭ1〜Ｍnの複数の表示領域は、図１３（ａ）及び（ｂ）に示すように、ｍ行ｎ列（ｍ、ｎは２以上の自然数）のマトリクス状に配列させてもよい（図１３では、モニタＭ1及びＭ２の２つが、２行２列の表示領域を有するものとしている）。
【０１３３】
そして、このとき、制御用ＣＰＵ３ａは、各モニタＭ1〜Ｍkに表示させる断層像対の表示位置を、その表示領域の行方向の位置及び列方向の位置の少なくとも一方が各モニタＭ1〜Ｍkの表示領域間で互いに対応するようにしてもよい。例えば、図１３（ａ）に示すように、三次元画像Ａの断層像対を形成する断層像ａがモニタＭ1の表示領域の左上隅（１行１列部分、以下「１−１」とする）に表示された場合、モニタＭ2に表示される三次元画像Ｂの断層像対を形成する断層像ｂは、図１３（ｂ）に示すように、当該モニタＭ2の同じ表示領域「１−１」に表示されるか、あるいは隣接する表示領域（「１行２列部分；「１−２」、あるいは２行１列部分；「２−１」）に表示される。また、図１４（ａ）に示すように、三次元画像Ａのｎ−１スライス番目の断層像「ア」、ｎスライス番目の断層像「イ」、ｎ＋１スライス番目の断層像「ウ」、ｎ＋２スライス番目の断層像「エ」は、それぞれ、モニタＭ1の表示領域「１−１」、「１−２」、「２−１」、「２−２」（２行２列目部分）に表示された場合、各断層像「ア、イ、ウ、エ」とそれぞれ断層像対を形成する三次元画像Ｂの断層像「あ、い、う、え」は、モニタＭ2の表示領域「１−１」、「１−２」、「２−１」、「２−２」に表示される。
【０１３４】
すなわち、本変形例では、多くの断層像対が同時に各モニタＭ1〜Ｍkに表示される。しかも、例えば、各モニタＭ1〜Ｍkの表示領域が２行２列のマトリクス状に形成されている場合、断層像対を形成する各断層像は、当該モニタＭ1〜Ｍk間で対応がとれた表示領域（例えば、行及び列の少なくとも一方が対応している）に表示されることになる。つまり、複数の断層像対をオペレータが比較読影しやすい状態で同時に表示することができ、比較読影の精度をより向上させることができる。
【０１３５】
・変形例６
図５のステップ１０７の処理における断層像対を指定する処理において、その指定処理を三次元画像の特徴量を利用して自動的に行なってもよい。その方法は、三次元画像間の相互相関（cross-correlation）係数を、片方の画像を移動させながら算出すると、三次元画像間のズレ量が、相関係数が最大になったときのシフト量から求められる。なお、相関演算については、文献（３）等で広く紹介されており、それを画像間のズレ量を推定する方法として使用した例は、文献（４）で紹介されている。また、対応をとる必要があるのはｚ軸だけであるので、ｚ軸に平行な投影面に各検査画像をパラレルビーム系で投影する（投影面に垂直な直線上の三次元画像の線積分を各画像毎に求める）ことにより、擬似スキャノグラム像を作成し、このスキャノグラム像に対し二次元相関演算を施すことにより、対応をとってもよい。
【０１３６】
例えば、ステップ１０７の代わりとして、図１５に示す処理を行なう。すなわち、制御用ＣＰＵ３ａからの制御指令に応じて、各処理装置４ａ1〜４ａ6は、画像記憶部４ｅ1〜４ｅ6から三次元画像データ「Ａ〜Ｆ」を読み出し（ステップ１０７ａ）、その三次元画像データ「Ａ〜Ｆ」に基づいて疑似スキャノグラムデータ「Ｓ1〜Ｓ6」を作成し、画像記憶部４ｅ1〜４ｅ6に格納する（ステップ１０７ｂ）。
【０１３７】
続いて制御用ＣＰＵ３ａは、作成された疑似スキャノグラムデータ「Ｓ1〜Ｓ6」に対して２次元相関演算処理を行ない、解剖学的断層位置が略同一の断層像対グループを抽出（指定）する（ステップ１０７ｃ）。以下、ステップ１０８に移行して上述したステップ１０８〜１２７の処理を行なう。
【０１３８】
すなわち、三次元相関演算を用いてズレ量を推定する手法を用いた場合、推定されるズレ量の精度は高いが多くの時間がかかるのに対し、疑似スキャノグラム像に対し二次元相関演算を用いてズレ量を推定する手法は、推定されるズレ量の精度は三次元相関演算を用いてズレ量を推定する手法に比べるとやや落ちるが、短時間で処理できることにある。さらに、投影面（投影角度）を変えて複数回処理する（ステップ１０７ｂ、ステップ１０７ｃの処理を投影面を変えて繰り返す）ことにより、推定されるズレ量の精度を上げることもできる。これらの処理は、全ての比較対象である三次元画像が間接的に関連づけられるような組み合わせを自動的に選択して行なわれる。
【０１３９】
・変形例７
変形例６において、画像の中で推定されるズレ量の精度に大きく貢献する情報を予め強調し、あるいは抽出してもよい。例えば、その情報に大きく貢献する情報の一つは、境界情報であるので、予めハイパスフィルタや微分フィルタ等を用いて、エッジ強調やエッジ抽出処理等を行ってもよい。この処理の特色は、モダリティや収集条件の違いにより画素値が保存されないピクセルが存在した場合でも、精度良くズレ量を推定することができるところにある。
【０１４０】
・変形例８
変形例６において、特徴の少ない領域（他の領域と区別がつきにくい、例えばのっぺりした領域）では、推定されるズレ量の精度はあまり良くないことは、文献（３）等多くの論文で述べられている。そこで、全ての領域で相関演算を行なわないで、特徴量の多い領域（他の領域と区別がつきやすい領域）を１つ以上選択してその相関演算の結果（相関の高いシフト量。このｚ軸方向のシフト量が対応している断層像対を表す）を算出し、この結果をステップ１０７の入力部３ｃから指定された結果と同様に考えることにより、ステップ１０７、ステップ１１１〜ステップ１２１の手順で位置合わせを行なってもよい。
【０１４１】
また、特徴量の多い領域は、オペレータが断層像やＲＯＩを選択することにより指定しても良いし、また周波数解析を行ない高周波成分が多い部位を選ぶ等の手段で、各処理装置若しくは制御装置が自動的に指定しても良い。
【０１４２】
・変形例９
図７のステップ１１４の処理において、三次元画像間の位置合わせのために指定された断層像対が３対以上ある場合には、ｚ座標を関連づけるために全ての断層像対を使用しないで近くにある複数個の断層像対を用いてローカルにｚ座標を関連づけてもよい。例えば、図１６（ａ）のように三次元画像「Ａ，Ｂ」の間に、断層像対「（Ａ1−Ｂ1）、（Ａk−Ｂk）、（Ａp−Ｂp）｛1＜k＜p｝」が指定されていた場合、断層像対（Ａ1−Ｂ1）と（Ａk−Ｂk）を用いて算出したズレ量（例えば、Ｇkとする）に基づいて、断層像対（Ａ1−Ｂ1）と（Ａk−Ｂk）の間またはその近辺に存在する断層像の位置合わせを行う。すなわち、図１６（ｂ）に示すように、断層像「Ａ1〜Ａk」間の断層像「Ａf、Ａg、Ａj」は、断層像「Ｂ1〜Ｂk」間の断層像「Ｂf、Ｂg、Ｂj」の位置と上記ズレ量Ｇkで表せる。同じく断層像対（Ａk−Ｂk）と（Ａp−Ｂp）との間で算出したズレ量（例えばＧpとする）に基づいて、断層像対（Ａk−Ｂk）と（Ａp−Ｂp）との間またはその近辺に存在する断層像の位置合わせを行なう。すなわち、図１６（ｃ）に示すように、断層像「Ａk〜Ａp」間の断層像「Ａm、Ａn、Ａo」は、断層像「Ｂk〜Ｂp」間の断層像「Ｂm、Ｂn、Ｂo」の位置と上記ズレ量Ｇpで表せる。
【０１４３】
これは、体格差のある被検体を撮影した際の画像間で比較する場合や、体格が同じものでも撮影時に多少姿勢が異なっていた等の場合に、ｚ座標のズレを極力抑える効果がある。
【０１４４】
・変形例１０
図６のステップ１２１、及びステップ１２６の画像補間処理において次の断層像対を作成するときに、前後に存在する断層像対への距離の比が同じ位置の断層像を作成してもよい。例えば、図１７のように、三次元画像「Ａ，Ｂ」の間に、断層像対「（Ａ1−Ｂ1）、（Ａk−Ｂk）、（Ａp−Ｂp）｛1＜k＜p｝」が指定されていた場合であり、断層像対（Ａ1−Ｂ1）と（ＡK−ＢK）の間において当該断層像「Ａ1〜Ａk」間を「ｍ1：ｎ1」に内分する位置の断層像を表示する際には、三次元画像Ｂにおいて断層像「Ｂ1〜ＢK」間を「ｍ1：ｎ1」に内分する位置の断層像を表示すればよい。同じく断層像対（Ａk−Ｂk）と（Ａp−Ｂp）の間において当該断層像「Ａk〜Ａp」間を「ｍ2：ｎ2」に内分する位置の断層像を表示する際には、三次元画像Ｂにおいて断層像「Ｂk−Ｂp」間を「ｍ2：ｎ2」に内分する位置の断層像を表示すればよい。
【０１４５】
これは変形例９と同様に、体格差のある被検体を撮影した際の画像間で比較する場合や、体格が同じものでも撮影時に多少姿勢が異なっていた等の場合に、ｚ座標のズレを極力抑える効果があり、特にこの目的に対しては変形例９より効果的である。
【０１４６】
・変形例１１
図６のステップ１２１において、順次表示するための断層像対データを作成するタイミングは、予め全ての対応する位置（ｚ座標）の断層像データを作成しておいてもよいし、次断層像対表示指令（シングルアクション）又は断層像対を順次表示する指令（シングルコマンド）が入力されたときに、対応する位置の断層像対データを作成してもよい。前者は、最初に全ての断層像対を作成してしまうので、作成にまとまった時間が必要になり、断層像対の表示を即座に開始することができない。しかし、断層像対の表示を一度開始してしまえば、後は高速に断層像対の順次表示を行なうことができる。一方、後者は前者の逆で、断層像対の順次表示は前者ほど高速に行なえない代わりに、順次表示をすぐに開始することが出来、また各断層像対データを記憶しておく画像記憶部４ｅ1〜４ｅmの記憶容量が少なくてもよいという長所を有する。さらに、前者では、基準断層像を変更すると観察しなかった画像を作成する時間を無駄に費やしたことになるが、後者は、断層像対を表示する段階で、基準断層像を簡単に変更することができる。
【０１４７】
・変形例１２
図６のステップ１２０〜１２１の処理では、断層像対を次々表示するために、算出したｚ座標の位置に断層像が無い場合は補間処理を用いて新たに算出している。
【０１４８】
しかしながら、表示すべき位置（ｚ座標位置、断層面）に断層像が無い場合は、その断層面からいちばん近い断層像を表示してもよい。例えば、図１８のように、三次元画像「Ａ、Ｂ」の間で位置合わせを行なった結果、「Ａia−Ｂi（i=1,2,…,N)」が断層像対であり、その断層面Ａiaの位置には、断層像が存在しない場合、この断層面Ａiaから一番近い断層像Ａiを断層像Ｂiに対応する断層像として表示する。この方法では、算出されたｚ座標位置の断層像をわざわざ作成する必要がないので、高速な順次断層像対表示が可能となる。
【０１４９】
上述した断層面から一番近い位置に存在する断層像を表示する処理は、例えば比較対象となる部位の構造が細かい場合等においては、図９（ｂ）に示したように補間処理を用いて算出されたｚ座標位置のＣＴ値を求めると、その部位の構造が消滅してしまうような場合に用いられる。
【０１５０】
つまり、このような比較対象部位が細かい場合には、補間処理を用いて新たな断層像を算出せず、常に最も近い位置に存在する断層像を表示するシステムを用いることもできる。また、比較対象部位が細かい場合においては、算出されたｚ座標の前後のスライスを同時に表示することもできる。
【０１５１】
すなわち、図１９及び図２０に示すように、三次元画像「Ａ、Ｂ」の間で位置合わせを行なった結果、「Ａia’−Ｂi'（i=1,2,…,N)」が断層像対であり、例えば、その断層面Ａ1a’の位置には断層像が存在しない場合、図６のステップ１２０の判断の結果（ＮＯ）の後、図２０に示すように、この断層面Ａia’のｚ座標位置に対して前後するｚ座標位置の断層像Ａi'及び断層像Ａ(i+1)'を断層像Ｂi'に対応する断層像として画像記憶部４ｃ1から順次読み出し、所定タイミング毎にモニタに同時表示する（ステップ１２１ａ）。なお、モニタの表示領域が２断層像分しかない場所には、Ｂi'と同時にＡi'もしくはＡ(i+1)’を表示し、一定時間毎にＡi'とＡ(i+1)’を交互に表示させてもよい。また、シングルコマンドにてＡi'とＡ(i+1)'とを表示を切り換えてもよい。この方法においても、算出されたｚ座標位置の断層像を作成する必要がないので、高速な順次断層像対表示が可能となる。
【０１５２】
上述した補間処理を用いることなく、算出されたｚ座標に最も近い位置の断層像（あるいは、ｚ座標前後の位置の断層像）を表示させるシステムでは、演算処理がｚ座標を求める処理のみになり、補間処理を行なう必要がなくなるので、処理が大幅に軽減され、ソフトウエア・ハードウエアともにその構成が単純となり、非常に安価なシステムとなる。
【０１５３】
・変形例１３
本実施形態では、ステップ１１１〜ステップ１２１やステップ１２６の処理において、断層像対を次々に表示するために、第１の断層像対の位置情報に基づいて略同一の各断層像対のｚ座標とそのｚ座標位置に相当（対応）する断層像対を設定（その位置に断層像がある場合には、その位置の断層像をそのまま導出して設定し、無い場合には補間処理を用いた新たに算出（導出））している。言い換えれば、いわゆるＩＦ文で判断し、その位置に断層像がある場合はその断層像をそのまま表示し、無い場合は補間処理を用いて新たに作成している。
【０１５４】
しかしながら、断層像間の間隔は、各病院間によって部位毎に基準値がある。例えば耳は１ｍｍ、肺野は２ｍｍ等である。また近年、集団検診においても三次元画像を撮影する動きが出てきたが、その検査にも基準値が設定されている。例えば、集団検診にＣＴ三次元画像を利用している代表的な例としては、東京から肺癌を無くす会（ＡＬＣＡ；Anti-Lung Cancer Association）によって行なわれている肺癌検診があるが、この検診においては、前記基準値を１０ｍｍとしている。このように多くの三次元検査画像は、基準値（基準断層像間隔）で再構成されている。
【０１５５】
すなわち、各三次元検査画像が同一の基準断層像間隔で作成されている等の理由により、読み込んだままの三次元画像を順次表示しても位置ズレがほとんど生じない場合には、各三次元画像の第１の断層像間の位置情報に基づいて一度同一の断層像対が指定されれば、後は略同一の各断層像のｚ座標とそのｚ座標位置に相当する断層像を設定しなくても（すなわち、お互い断層像位置を更新しても、その更新位置が基準間隔毎で略同一位置であるから、補間処理を用いて新たに断層像を作成する必要はない）、次断層像対表示指令（シングルアクション）又は次断層像対順次表示指令（シングルコマンド）を受ける度毎に各断層像を次の断層像に更新するだけで、常に略同一の断層像対が表示できる。
【０１５６】
例えば、三次元画像Ａと三次元画像Ｂとが略同じ断層間隔（基準断層像間隔）で取得されていた場合において、三次元画像Ａと三次元画像Ｂの断層像対を同期表示させるとする。なお、三次元画像Ａは、処理装置４ａ1の画像記憶部４ｅ1に格納されているとし、三次元画像Ｂは、処理装置４ａ2の画像記憶部４ｅ2に格納されているとする。
【０１５７】
このとき、オペレータは、図５のステップ１０７の処理として、三次元画像Ａの所望のスライス位置（そのスライス位置に対応するｚ座標を「Ｚr」とする）の断層像Ａ（ｚr）を指定するともに、その断層像Ａ（Ｚr）と解剖学的断層位置が略同一の断層像Ｂ（Ｚrb）を三次元画像Ｂの中から指定する（図２１、ステップ３０１）。そして、制御用ＣＰＵ３ａは、その２つの断層像「Ａ（ｚr）、Ｂ（Ｚrb）」をそれぞれ同期させて表示させる指令を処理装置４ａ1及び４ａ2に送る。処理装置４ａ1及び４ａ2のＣＰＵ４ｂ1及び４ｂ2は、ステップ１１９〜１２６と同様の処理を行なう。この結果、断層像Ａ（Ｚr）及び断層像Ｂ（Ｚrb）が例えばモニタＭ1、Ｍ2に同期して表示される（ステップ３０２）。
【０１５８】
このとき、三次元画像データ「Ａ、Ｂ」は、同一の断層間隔で取得されているため、読影者が、入力部３ｃのキーボードＫや次断層像表示指令スイッチＮ等を操作してシングルアクションか、あるいは、シングルコマンドを入力すれば、三次元画像「Ａ、Ｂ」の中からその断層間隔毎に断層像（断層像対）が選択される。そして、その断層像対は、それぞれ同期して順次モニタＭ1、Ｍ2に表示される（ステップ３０３）。以下、ステップ１２７の処理を行なう。
【０１５９】
このように、断層間隔が略同一の場合、最初に解剖学的断層位置が略同一の断層像を指定してしまえば、後はその断層間隔毎に断層像対を順次表示することができる。したがって、処理が大幅に簡略化され、処理速度が大幅に向上する。また、処理が大幅に簡略化されるため、ソフトウエア・ハードウエアともにその構成が単純となり、非常に安価なシステムとなる。
【０１６０】
・変形例１４
図６のステップ１２０〜１２１の処理において、同期して順次表示している際に、例えば、画像指定部Ｉ1を所定の方向に切り換えること、あるいは画像指定部Ｉ2の同期解除スイッチＣを押すことにより、一部又は全ての画像の同期表示を停止させることもできる。これは、個別に読影したくなった場合に有効である。また、一部の表示画像の断層面が他の表示画像の断層面の位置とずれてきた場合、その画像を非同期にして位置を調節した後に、再度同期モードにするという同期位置のズレに対する補正手段としても有効である。
【０１６１】
・変形例１５
図６のステップ１２２の処理でのフィルタの種類を断層像対間で一致させる処理において、画像再構成時にかけられるフィルタを考慮してもよい。これは、再構成フィルタの別の再構成像のＭＴＦ（Modulation Transfer Function：変調伝達関数）を登録しておき、ＭＴＦを一致させるフィルタをかけることにより、達成される。例えば、画像Ａ，ＢのＭＴＦがそれぞれＨA（ｕ，ｖ）、ＨB（ｕ，ｖ）とした場合、画像Ａ，Ｂの対象物体が同じＦ（ｕ，ｖ）だとすると、三次元画像「Ａ，Ｂ」の分布ＦA（ｕ，ｖ）、ＦB（ｕ，ｖ）は、それぞれ次のように得られている。
【０１６２】
【数４】
ＦA（ｕ，ｖ）＝Ｆ（ｕ，ｖ）ＨA（ｕ，ｖ） ……（４）
【数５】
ＦB（ｕ，ｖ）＝Ｆ（ｕ，ｖ）ＨB（ｕ，ｖ） ……（５）
ここで、（ｕ，ｖ）は、フーリエ面の座標であり、Ｆ（ｕ，ｖ）、ＦA（ｕ，ｖ）、ＦB（ｕ，ｖ）は、全て元の物体や画像をフーリエ変換した結果である。また、ここでは簡単のために二次元断層面の処理について説明しているが、三次元での処理も同様に処理できる。ところで、ＦA（ｕ，ｖ）、ＦB（ｕ，ｖ）は本来同じ対象物体なので、ＭＴＦが同じであれば、同じ画像となるはずである。したがって、ＭＴＦを一致させるためには、ＦB（ｕ，ｖ）にフィルタＨ（ｕ，ｖ）をかければよい。
【０１６３】
【数６】
Ｈ（ｕ，ｖ）＝ＨA（ｕ，ｖ）／ＨB（ｕ，ｖ） ……（６）
この画像再構成時にかけるフィルタも考慮する処理の特徴は、画素値が保存されるモダリティにおいて、画素値を直接比較できることにある。
【０１６４】
・変形例１６
本システムの機能の一部又は全部は、撮影モダリティに設けられていてもよい。
【０１６５】
・変形例１７
図５のステップ１０６の基準データを設定する処理は、基準断層像Ａを制御装置３の入力部３ｃより設定し、その基準断層像Ａの各表示条件をその基準断層像Ａが保持されている画像処理装置４ａ1のメモリ４ｃ1から読み込むことにより、基準データを基準断層像Ａの各表示条件に設定してもよい。この処理の特徴は、非常に多くの基準データの設定を自動化できることにある。
【０１６６】
・変形例１８
断層像対を形成する各断層像の座標系を基準断層像の座標系に合わせる回転処理、特に断層面の回転（断層像対の断層面を平行にする回転）は行なわなくてもよい。つまり、基準断層面（ｘｙ平面）と比較する断層面とのなす角が小さい（平行に近い）場合や、注目領域（ＲＯＩ:Region of interest）が決まっている場合は、比較する領域の一部（例えば中心）の位置を合わせさえすれば、比較対象であるその他の領域の位置ズレ量はそれほど大きくはなく、このような場合には、比較対象の三次元データの断層面を、基準断層面と平行にする処理（断層面の傾きを変えるような回転処理）を省いてもよい。この方法の特長は、三次元補間という時間のかかる処理を省けるので、高速に処理できる点にある。
【０１６７】
すなわち、基準断層像からのズレ量が小さい場合やＲＯＩが決まっている場合に、三次元補間という時間のかかる処理を省いて高速に処理できので、有効な方法であると思われる。後者の場合の処理方法は、任意の三次元画像の断層面上に注目領域を設定した後、その注目領域に対応する領域を比較対象の三次元画像において算出した後、対応する領域を多く含む断層面を表示してもよいし、図５のステップ１０１〜ステップ１０５の処理において、座標の不一致は無視してＲＯＩの位置合わせを行なうことにより、座標を関係づけてもよい。
【０１６８】
・変形例１９
本実施例では、断層像対をフレーム毎に異なるモニタに表示させたが、本発明はこれに限定されるものではなく、例えば、断層像対を構成するある断層像の一部を他の断層像対を構成する断層像の画像に変換することもできる。
【０１６９】
今、ｋ組の三次元画像データがそれぞれ、処理装置４ａ1〜４ａkの画像記憶部４ｅ1〜４ｅkに記憶されているとした場合、最初オペレータの操作に基づく制御用ＣＰＵ３ａの処理により、処理装置４ａ1の画像記憶部４ｅ1に記憶された三次元画像データＡがモニタＭ1に順次表示されているとする。
【０１７０】
このとき、読影者は、表示された三次元画像Ａを見ながら、現在表示された断層像上のある部位を他の三次元画像と比較して読影したいとする。ここで、読影者は、入力部３ｃのキーボードＫ等を操作して現在表示されている断層像Ａ（Ｚａ１）を停止（フリーズ）させる。そして、キーボードＫ等を操作してその比較したい部位をＲＯＩで指定するとともに、比較読影したい三次元画像（その三次元画像が記憶されている処理装置、例えば４ａ2とする）を指定する（図２２、ステップ４０１）。このとき、制御用ＣＰＵ３ａは、その指定されたＲＯＩ及び処理装置４ａ2を読み込み、画像記憶部４ｅ1に格納された三次元画像データの中から三次元画像データＡ（Ｚａ１）を読み出すとともに、画像記憶部４ｅ2に格納された三次元画像データＢの中から、当該ｚ座標（Ｚａ１）と解剖学的断層位置の等しい三次元画像データ（断層像対）Ｂ（Ｚｂ１）を読み出す（ステップ４０２）。そして、制御用ＣＰＵ３ａは、読み出したＡ（Ｚａ１）の内、ＲＯＩに指定された領域のデータ（ピクセル値）を消去し、その領域へ読み出した三次元画像データＢ（Ｚｂ１）の対応する領域のデータ（ピクセル値）を合成する（ステップ４０３）。そして、制御用ＣＰＵ３ａは、この結果得られた合成三次元画像データＡb（Ｚａ１）を表示させる指令を処理装置４ａ1に送る。処理装置４ａ1のＣＰＵ４ｂ1は、その指令に基づいてモニタＭ1に画像データＡb（Ｚａ１）を表示させる。この結果、モニタＭ1には、図２３に示すように、指定されたＲＯＩ内の領域だけ比較読影したい他の三次元画像Ｂ（Ｚｂ１）が合成された三次元画像Ａb（Ｚａ１）が表示される（ステップ４０４）。
【０１７１】
ここでさらに、入力部３ｃのキーボードＫ等を操作して、ＲＯＩを移動若しくは、「合成オン、合成オフ」等のコマンドを送ると、前者は、合成した部分をもとの三次元画像Ａに戻し、指定された移動量だけ平行移動した領域にＲＯＩを設定（ステップ４０１と同様の処理）し、ステップ４０２以下の処理を繰り返して、新たに指定されたＲＯＩに三次元画像Ｂの対応する領域を合成して表示する。また、後者では、オフが押された場合には、合成した部分をもとの三次元画像Ａに戻し、再びオンが押された場合は、指定されていたＲＯＩに再び三次元画像Ｂを合成する。注目領域の近辺でＲＯＩを左右に移動させたり、注目領域で合成のオン・オフを繰り返すことにより、注目領域における三次元画像ＡとＢとの違い（陰影の変化や有無）が明確に判別できるので、正確な読影を行なうことができる。
【０１７２】
なお、比較したい領域は部分的ではなく、例えばモニタ画面半分というような指定の仕方も可能である。すなわち、オペレータが入力部３ｃの例えばマウスｍを操作してモニタ画面上に境界線データ（例えば、画面中央で表示領域を左右２等分するラインマーカー）を指定する。そして、制御用ＣＰＵ３ａは、その境界線データに基づいて、三次元画像データＡ（Ｚａ１）上の例えばその境界線位置に対して向かって左側は元の三次元画像データＡ（Ｚａ１）を残し、向かって右側には三次元画像データＢ（Ｚｂ１）の内の右半分を合成させる。この結果、図２４に示すように、同一の解剖学的断層位置における異なった三次元画像がモニタＭ1の中央を境にして合成された状態で表示される。
【０１７３】
また、この状態でマウスｍ等を操作してラインマーカーを例えば左右に移動させることにより、ラインマーカーを境にして三次元画像ＡとＢとの違い（陰影の変化や有無）が明確に判別できるので、正確な読影を行なうことができる。
【０１７４】
さらに、この考え方の発展した例として、ＲＯＩで指定した領域（あるいは、境界線で仕切られた内の一方の領域）に、他のモダリティで得られた画像を合成することもできる。
【０１７５】
これらの変形例によれば、読影者は視点を移動させることなしに、断層像対を読影することができる。また、比較する位置の対応がつきやすいため、読影効率を大幅に高めることができる。さらに、断層像を一枚しか表示できない表示装置でも比較読影が可能となるという特長を有している。
【０１７６】
・変形例２０
図６のステップ１２２において、各ＣＰＵ４ｂ1〜４ｂ6は、同一のｚ座標の断層像データ（断層像対データ）を単にモニタＭ1〜Ｍ6に同一のタイミングで表示したが、本発明はこれに限定されるものではなく、例えば、断層像対を形成する各断層像データの中から複数（例えば２つ）の断層像データ（１組の断層像対データ）を選択し、その２つの断層像データをサブトラクション処理して得られたデータ（サブトラクションデータ）を表示することもできる。
【０１７７】
このサブトラクション処理を行なう場合には、読影者は、図５のステップ１０６の処理における基準データを入力する際に、サブトラクション処理した画像を表示したい２つの三次元画像（その２つの三次元画像が記憶された処理装置、例えば４ａ1、４ａ2）、及びそのサブトラクション画像を表示するモニタ（例えばＭ1）を入力している。
【０１７８】
一方、図６のステップ１２２の処理を終了した各ＣＰＵ４ｂ1〜４ｂ6の内、ＣＰＵ４ｂ1、４ｂ2は、図６に示す処理ではなく、図２５に示す処理を行なう。すなわち、ＣＰＵ４ｂ1、４ｂ2は、画像記憶部４ｅ1、４ｅ2からそれぞれ断層像データ（断層像対データ）「Ａ（Ｚａ１），Ｂ（Ｚｂ１）」を読み出す。そして、ＣＰＵ４ｂ2は、読み出した断層像データＢ（Ｚｂ１）をＣＰＵ４ｂ1に送る（ステップ５０１）。ＣＰＵ４ｂ1は、断層像データＡ（Ｚａ１）から断層像データＢ（Ｚｂ１）をサブトラクション処理する（ステップ５０２）。そして、得られたサブトラクション画像データ｛（Ａ（Ｚａ１）−Ｂ（Ｚｂ１））；ＳＵ１とする｝を表示装置５ａ1に送る。表示装置５ａ1は、送られたサブトラクション画像データ「ＳＵ１」に対し第２の表示条件であるγ特性変換等の画像処理やブライトネス、コントラスト等の設定を行なった後、そのサブトラクション画像データ「ＳＵ１」をモニタＭ1に表示する（ステップ５０３）。
【０１７９】
このとき、図６のステップ１２５〜ステップ１２７に類した処理が行なわれているため、ＣＰＵ４ｂ1、４ｂ2は、サブトラクション処理された断層像データ「Ａ（Ｚａ１），Ｂ（Ｚｂ１）」から所定間隔Δｄ離れた断層像データＡ「（Ｚａ２），Ｂ（Ｚｂ２）」を画像メモリ４ｅ1、４ｅ2から同一のタイミングで読み出し、ステップ５０１〜５０３に示すサブトラクション処理を行なう。以下、ステップ１２０の処理にしたがってサブトラクション画像｛「ＳＵ２」、「ＳＵ３」、…｝が順次モニタＭ1表示される（ステップ５０４）。
【０１８０】
本変形例によれば、解剖学的断層位置が同一の断層像対を同期して表示させるだけでなく、その断層像対どうしのサブトラクション画像を順次表示することができるため、そのサブトラクション画像を読影することにより、患者の診断対象部位の変化状態等を詳細に検討することができる。
【０１８１】
・変形例２１
上述した変形例１９の合成処理、及び変形例２０のサブトラクション処理では、合成あるいはサブトラクションされる断層像対におけるスライス内の位置関係は略一致していることが望ましい。したがって、断層像対間でスライス位置が若干ずれている場合には、上述した図２２及び図２５の処理に加えて相関演算処理を行なうことにより、スライス内の位置関係のずれによる画像の差を補正し（対応している断層像対において各部位のピクセル位置を合わせる）、より正確な合成処理及びサブトラクション処理を行なうことができる。スライス内の位置関係のズレによる画像の差を補正する方法は、全体の位置ズレを補正する方法、及び体動による部分的なズレまで補正するためにローカルに相関演算を行ない、その結果をもとに対象を部分的に変形する方法等があり、これらの方法は、参考文献（４）により公知となっている。
【０１８２】
すなわち、合成処理（あるいはサブトラクション処理）において、図２２のステップ４０２の処理後（あるいは図２５のステップ５０１の処理後）、制御用ＣＰＵ３ａは、図２６に示すように、断層像データＡ（Ｚａ１）と断層像データＢ（Ｚｂ１）とを相関演算する（ステップ６０１）。そして、制御用ＣＰＵ３ａはその相関演算結果に基づいて断層像データＡ（Ｚａ１）と断層像データＢ（Ｚｂ１）との間のシフト量を演算する（ステップ６０２）。そして、このシフト量、及び断層像データＢ（Ｚｂ１）に基づいて、断層像データＡ（Ｚａ１）との相関が非常に高い（同一のスライス位置とみなされる）断層像データＢa（Ｚｂ１）を生成して再度画像記憶部４ｅ2に記憶する（ステップ６０３）。以下、ステップ４０３以降の処理が行なわれる。ただし、断層像データＢ（Ｚｂ１）として用いられるのは、新たに作成された断層像データＢa（Ｚｂ１）である。
【０１８３】
本変形例によれば、スライス位置にずれがある断層像対を用いて合成処理あるいはサブトラクション処理を行なう場合でも、正確な合成処理あるいはサブトラクション処理を行なうことができる。
【０１８４】
・変形例２２
本実施例の画像処理装置のシステム構成は、図１の示す構成に限定されるものではなく、例えば、図２７に示すように、図１に示す処理装置４ａ1〜４ａmの内の１つ（処理装置１０ａ1）が制御装置３を兼ねる構成（新たに入力部３ｃが加わっている）としてもよいし、図２８に示すように、処理装置４ａ1〜４ａmを無くし、制御装置３が処理装置４ａ1〜４ａmを兼ねてもよい。この場合は、制御装置３が各処理装置４ａ1〜４ａmの行なう全ての処理を変わって行なう。図２５、２６のようなシステムは、図１のシステムに比べて処理速度が遅くなる可能性があるが、比較的安価に実現できるので、実用化しやすいとう特長を有する。
【０１８５】
・変形例２３図５のステップ１０８〜ステップ１０９とステップ１０７の各処理を行なう順番は、最終的には同じとなるので、どのように設定してもよいが、図５のステップ１０８〜ステップ１０９の処理を行ない易いように設定するのが望ましい。断層像対の各表示条件、例えば、ピクセルピッチが大きく違う場合は、位置合わせを行なうことが難しいこともある。このような時は、ステップ１０７の処理を予め行なったほうが位置合わせを行ない易い。
【０１８６】
なお、本実施例及び変形例において、処理装置４ａ1〜４ａmはそれぞれ１つのモニタＭ1〜Ｍmを有する構成でもよい。また、本画像処理装置が種々の画像処理等を高速に行なうことができる高速演算装置（マイクロプロセッサ）を別個に有し、上述した制御用ＣＰＵ３ａ及び処理装置４ａ1〜４ａmで行なわれた処理等を必要に応じて上記高速演算装置で行なう構成にしてもよい。
【０１８７】
さらに、上述した変形例で述べられた合成演算、サブトラクション演算、及び相関演算等は制御用ＣＰＵ３ａあるいは各処理装置４ａ1〜４ａmのＣＰＵ４ｂ1〜４ｂmで行なう構成としたが、本発明はこれに限定されるものではなく、例えば、合成演算を行なうディジタルロジック回路、サブトラクション演算を行なうディジタルロジック回路、及び相関演算を行なうディジタルロジック回路を別個に設けて、上記合成演算、サブトラクション演算、及び相関演算を行なわせてもよい。
【０１８８】
さらにまた、本実施例では、医用画像撮影モダリティとしてＣＴ装置を用いたが、本発明はこれに限定されるものではなく、例えば、ＭＲＩ装置であってもよく、また、ＣＴ装置とＭＲＩ装置とを備え、これらの装置で得られた複数組の三次元画像を比較読影してもよい。
【０１８９】
なお、以下に、実施例で参照した参考文献名を掲載する。
（１） 「テレビジョン画像工学ハンドブック」テレビジョン学会編、オーム社。
（２） 「High Speed Display through Hybrid Processing」Optics Letters, Vol.15, No.10, 565-567(1990)。
（３） 「ディジタル画像処理」Kak Rosenfeld著、長尾真訳、近代科学社。
（４） 「Digital Image Subtraction of Temporally Sequential Chest Image for Detection of Interval Change 」,A. Kano et al., Medical Physics, Vol. 21, No.3, 453-461(1994)。
【０１９０】
【発明の効果】
以上述べたように本発明によれば、複数組の３次元画像の中から、異なる組相互間で解剖学的断層位置が略同一の少なくとも１つの断層像対を設定し、さらに、その断層像対を例えば所定のタイミングで同期させて表示することができるため、異なる３次元画像同士において互いの解剖学的断層位置を合わせるという非常に困難な操作をすることなしに、断層像対を表示することができる。したがって、比較読影の際の読影者の負担が減少し、また、比較読影にかかる時間及びコスト（読影者の人件費等）が減少する。そして、この効果による２次的な効果としては、比較読影を行なう頻度が多くなり、診断の正確さがより向上することが挙げられる。
【０１９１】
また、複数の断層像対間の断層間隔の違いによる解剖学的断層位置の位置合わせを行なうことなしに、複数の断層像対を順次表示することができるため、比較読影にかかる時間及びコストが減少する。
【０１９２】
さらに、断層像対を構成する複数の断層像（表示画像）の明るさ、コントラスト、適用される画像フィルタ等の画像パラメータの少なくとも一部を、読影者の余計な操作なしで当該断層像対を構成する複数の断層像間で同一に設定することができるため、読影者が比較読影する際に、上記画像パラメータを逐一合わせる必要がなくなり、読影者の労力を大幅に低減させながら比較読影の精度を向上させることができる。
【０１９３】
さらにまた、複数のモニタを階層構造で配設している場合に、当該モニタに表示される断層像対を構成する複数の断層像の少なくとも一部は、それぞれ隣接した位置のモニタ、つまり、比較される断層像対が隣接して見られるような位置に自動的に表示される。したがって、一度表示された断層像対の表示位置を変更する等の余計な手間が不要になり、読影者の労力を大幅に低減させながら比較読影の精度を向上させることができる。
【図面の簡単な説明】
【図１】 本発明に係る画像表示システムの概略構成を示すブロック図。
【図２】 図１における表示装置の概略構成を示すブロック図。
【図３】 （Ａ）は、入力部の概略構成を示す斜視図であり、（Ｂ）は、画像指定部Ｉ2の構成の一例を示す図。
【図４】 階層構造に配設された複数のモニタを示す斜視図。
【図５】 解剖学的断層位置が略同一の断層像対を同期させて表示する際の画像表示システム全体の動作の一例を表す概略フローチャート。
【図６】 解剖学的断層位置が略同一の断層像対を同期させて表示する際の画像表示システム全体の動作の一例を表す概略フローチャート。
【図７】 解剖学的断層位置が略同一の断層像対を同期させて表示する際の画像表示システム全体の動作の一例を表す概略フローチャート。
【図８】 断層像対の概念を示す図。
【図９】 算出されたｚ座標の断層像データを生成する際の考え方を説明するための図。
【図１０】 ピクセルサイズを一致させる処理の概念を説明する図。
【図１１】 座標系を一致させる処理の概念を説明する図。
【図１２】 処理装置に複数個の表示装置が割り当てられている場合の、各モニタに表示された断層像対を表す図。
【図１３】 複数のモニタの複数の表示領域内の同一位置に表示された断層像対を示す図。
【図１４】 複数のモニタの複数の表示領域内の同一位置に表示された断層像対を示す図。
【図１５】 スキャノグラム像に対する２次元相関演算処理の一例を説明するための概略フローチャート。
【図１６】 複数個の断層像対を用いてローカルにｚ座標を関連づける処理を説明する概念図。
【図１７】 前後に存在する断層像対への距離の比が同じ位置の断層像を作成する処理を説明する概念図。
【図１８】 表示すべきｚ座標位置から一番近いｚ座標位置に存在する断層像を表示する処理を説明する概念図。
【図１９】 表示すべきｚ座標位置の前後のｚ座標位置に存在する断層像を同時に表示する処理を説明する概念図。
【図２０】 表示すべきｚ座標位置の前後のｚ座標位置に存在する断層像を同時に表示する処理の一例を説明するための概略フローチャート。
【図２１】 略同じ断層間隔で取得された２組の三次元画像を同期表示させる際のシステム全体の処理の一例を説明するための概略フローチャート。
【図２２】 断層像対を構成するある断層像の一部を他の断層像対を構成する断層像の画像に変換する際のシステム全体の処理の一例を説明するための概略フローチャート。
【図２３】 指定ＲＯＩ内だけ他の断層像が合成された断層像を示す図。
【図２４】 境界を挟んで２つの異なる断層像が合成された断層像を示す図。
【図２５】 サブトラクション処理を行なう際のシステム全体の動作の一例を説明するための概略フローチャート。
【図２６】 相関演算処理を行なう際のシステム全体の動作の一例を説明するための概略フローチャート。
【図２７】 本発明の画像表示システムのその他の構成例を示す概略ブロック図。
【図２８】 本発明の画像表示システムのその他の構成例を示す概略ブロック図。
【符号の説明】
１ ＣＴ装置
２ 画像ＤＢ
３ 制御装置
３ａ 制御用ＣＰＵ
３ｂ 主メモリ
３ｃ 入力部
４ａ1〜４ａm 画像処理装置
４ｂ1〜４ｂm ＣＰＵ
４ｃ1〜４ｃm メモリ
４ｅ1〜４ｅm 画像記憶部
５ａ1〜５ａn 表示装置
５ｂ1〜５ｂn ＬＵＴ
５ｃ1〜５ｃn 画像メモリ
５ｄ1〜５ｄn 表示部
Ｍ1〜Ｍn モニタ[0001]
BACKGROUND OF THE INVENTION
  The present invention relates to an image display system that displays a plurality of sets of diagnostic tomograms (three-dimensional images) obtained by a plurality of examinations, and an image display method using the system, and more particularly to a plurality of sets of three-dimensional images. In particular, the present invention relates to an image display system used for interpretation of a group of tomograms (hereinafter referred to as tomogram pairs) having substantially the same anatomical position, and an image display method using the system.
[0002]
[Prior art]
  Comparing and interpreting three-dimensional images (multiple tomographic images) obtained by performing multiple examinations with the same or different medical imaging modalities (for example, CT apparatus, MRI apparatus, etc.) Has contributed greatly. In the present invention, as described above, comparing and interpreting tomographic images obtained with the same or different medical image photographing modalities is referred to as comparative interpretation.
[0003]
  In particular, it is well known that when a newly taken examination image is compared with a past examination image of the same subject, the occurrence of an abnormal site is recognized as a change between the images, so that the accuracy of diagnosis is improved.
[0004]
  For example, in the case of comparative interpretation of a plurality of tomographic images obtained by two different examinations (first examination and second examination), two image display systems are prepared in advance. A plurality of tomographic images (first three-dimensional images) obtained by the first examination are displayed on the monitor, and a plurality of tomographic images obtained by the second examination are displayed on the monitor of the other image display system. (Second three-dimensional image) is displayed.
[0005]
  Then, the first three-dimensional image and the second three-dimensional image are sequentially displayed while operating the two image display systems separately, and the tomographic position that can be judged to be almost the same anatomically (considered) in the subject. A pair of tomographic images (hereinafter referred to as anatomical tomographic positions) were selected in each system for comparative interpretation.
[0006]
[Problems to be solved by the invention]
  As described above, when a plurality of tomographic images obtained by two different examinations are comparatively interpreted, a pair of tomographic images having substantially the same anatomical tomographic position must be selected from the plurality of tomographic images. However, the selection of the pair of tomographic images has the following problems, for example, in comparative interpretation of the entire three-dimensional image. Therefore, comparative interpretation is currently used only for a part of differential diagnosis in which a part of tomographic images in a three-dimensional image may be compared.
[0007]
  Problem (1)-When the tomographic interval (slice interval) of the three-dimensional images to be compared is different, the next anatomical tomographic position is compared each time a pair of tomographic images having substantially the same anatomical tomographic position is compared. However, the reader has spent a great deal of effort on the manipulation of the tomographic image pairs. In addition, even if the tomographic interval happens to be the same, when updating the tomographic images sequentially to compare the entire three-dimensional image, it is necessary to update the tomographic images one after another by the number of tomographic images to be compared. Therefore, this is also a troublesome process for the reader, and the burden has increased.
[0008]
  Problem (2)-When the ratio of the spatial distance between the display image and the real object is different between the inspection images to be compared, the size of an arbitrary part in the tomographic image is different between the inspection images. Correspondence must be performed for each site of interest using anatomical knowledge. Therefore, the burden on the reader was great.
[0009]
  Problem (3)-If the tomographic image intervals obtained by the inspection device are different, once a tomographic image with the correct anatomical tomographic position is found, comparative interpretation is always performed with the anatomical tomographic position aligned. As a result, the interpreter required a great deal of effort.
[0010]
  Problem (4)-Some of display conditions such as monitor characteristics such as WW, WL, brightness, contrast, look-up table, and γ characteristics, and the type of filter (hereinafter simply referred to as filter) applied to the image It was very difficult to make everything consistent with each image display system. For example, in order to match the types of WW, WL, lookup table, and filter, it is necessary to set several times for each image to be compared, increasing the burden on the reader. In addition, when comparing comparative readings between different monitors, it is generally difficult to adjust the brightness and contrast accurately because it is difficult to grasp the exact level of brightness and contrast, which is very inconvenient for comparative reading. It was.
[0011]
  Further, even if the exact level of brightness and contrast can be grasped as numerical values, it is very troublesome because it is necessary to set several times for each image to be compared in the same way as WW or the like. Furthermore, since it is impossible to match the characteristics of the monitor itself such as the γ characteristic at the time of product design or shipment, it is very inconvenient for comparative interpretation.
[0012]
  Problem (5)-Multiple monitors are arranged in a hierarchical structure so that the monitor screens can be easily compared with each other.Window tableWhen it is possible to display images and when film output capable of recording many images is possible, multiple monitors and multiple windowsIndo,Alternatively, it is necessary to bring the comparative image to a position on the film where it is easy to compare, for example, the adjacent upper and lower positions in the corresponding display area. For this purpose, it is necessary to display the comparison image again at the target position, which is a heavy burden on the reader.
[0013]
  The present invention has been made to solve all of the above-mentioned problems. After specifying a tomographic image pair having substantially the same anatomical tomographic position, the tomographic image having other anatomical tomographic positions having substantially the same position is designated. By providing an image display system capable of displaying image pairs automatically, for example, sequentially in a display mode that is easy for a reader to compare and interpret, the burden on the reader related to comparative interpretation, interpretation time, and cost for interpretation The purpose of this is to greatly reduce.
[0014]
[Means for Solving the Problems]
  In order to achieve the object, the image display system according to claim 1 is obtained by a plurality of examinations based on at least one medical image photographing modality.And each set consists of multiple tomographic imagesIn an image display system configured to display a plurality of sets of three-dimensional images on an output device, from among the plurality of sets of three-dimensional images,At the time of taking the examinationAnatomical fault locationBetween different sets of 3D imagesAlmost identicalA group of tomographic imagesFirst tomogram pairAsSpecifying means for specifying at least one and the plurality of sets of three-dimensional imagesOut ofat leastA set of3D imageOf multiple tomographic imagesFault spacing andSpecified by the specifying meansBased on position information between the first tomographic image pairBeforeAnatomic fault locationBetween different sets of 3D imagesAlmost identicalA group of tomographic images as a tomographic image pairAt least one tomographic image pairUsing the plurality of sets of three-dimensional imagesTomogram pair setting means to be set;By this tomographic image pair setting meansDisplay control means for displaying at least one set of tomographic image pairs on the output deviceWhen,It has.
[0015]
  In particular, according to the image display system of claim 2, the tomographic image pair setting unit automatically derives a tomographic image pair having substantially the same anatomical tomographic position.Configured toMeans.
[0016]
  In particular, according to the image display system described in claim 3, the display control means includes:EachSynchronize multiple tomograms that make up a tomogram pairLet meAs displayed on the output deviceConfigureing.
[0017]
  In particular, according to the image display system of claim 4,The tomographic image pair setting means is a means configured to set a plurality of the tomographic image pairs,The display control means sequentially orders the plurality of tomographic image pairs at predetermined intervals.Next,As displayed on the output deviceConfigureing.
[0018]
  In particular, according to the image display system according to claim 5, the designation unit serves as a reference from the plurality of sets of three-dimensional images.A set ofThe first designation means for designating a three-dimensional image, and this standardBecomeAt least one tomographic image in the three-dimensional image and the rest of the anatomical tomographic positions are substantially identicalPair ofTomographic images in 3D imagesAllSecond specifying means for specifying that all three-dimensional images are indirectly associated with each other.
[0019]
  Furthermore, according to the image display system according to claim 6, the specifying means automatically selects the first tomographic image pair from the three-dimensional image.Have.
[0020]
  Furthermore, according to the image display system of claim 7, the selection unit includes the plurality of sets of three-dimensional images.Using the amount of shift obtained by cross-correlation betweenThe first tomographic image pair is selected.
[0021]
  In the image display system according to the eighth aspect, the position information is position coordinate information based on coordinate axes set in a direction perpendicular to the tomographic plane of the three-dimensional image.
[0022]
  According to the image display system of claim 9, the tomographic image pair setting unit is configured to calculate a shift amount of position coordinate information between the plurality of sets of three-dimensional images, and to determine a shift amount of the position coordinate information. Based on the means for obtaining a relative position coordinate between the three-dimensional images, a designation means for designating an arbitrary position on the relative position coordinates between the three-dimensional images, and corresponding to the designated position Tomographic image to be determined from each 3D imageAcquisition meansThe display control means displays tomographic image groups respectively obtained from the respective three-dimensional images on the output device as tomographic image pairs.
[0023]
  In particular, according to the image display system of claim 10, theAcquisition meansUses at least one tomographic image at a position near the designated position when there is no tomographic image corresponding to the designated position.To create new tomographic image dataIt is means for obtaining a tomographic image corresponding to the designated position by interpolation processing.
[0024]
  In particular, according to the image display system of claim 11, theAcquisition meansIs means for selecting a tomographic image at a position closest to the designated position from among the three-dimensional images as a tomographic image corresponding to the designated position when there is no tomographic image corresponding to the designated position. .
[0025]
  Furthermore, according to the image display system of claim 12, theAcquisition meansWhen there is no tomographic image corresponding to the designated position, a tomographic image of a pair of positions adjacent to the designated position on the coordinate axis is used as a tomographic image corresponding to the designated position in each three-dimensional image. The display control unit is configured to select the pair of tomographic images corresponding to the designated position so as to form a group of tomographic image pairs as a whole.PairwiseThe information is displayed on the output device.
[0026]
  On the other hand, according to the image display system according to claim 13, each tomographic image of the plurality of sets of three-dimensional images is acquired at substantially equal tomographic intervals, and the tomographic image pair setting means includes the substantially equal tomographic interval and Based on position information between first tomogram pairsBeforeMultiple sets of 3D imagesUsingAt least one tomographic image pair having substantially the same anatomical tomographic position is set.
[0027]
  In the image display system according to claim 14, the output device includes a plurality of monitors arranged in a hierarchical structure, and the display control unit includes a plurality of tomographic images forming the tomographic image pair. Each of the plurality of monitorsHitAt least a part of the plurality of tomographic images is displayed adjacent to each other.
[0028]
  In particular, according to the image display system according to claim 15, the display control means includes brightness, contrast of a plurality of tomographic images constituting the tomographic image pair in each monitor,as well as,Applied image filterincludingMake at least some of the image parameters the sameToteThe plurality of tomographic images are displayed on the respective monitors.
[0029]
  In particular, according to the image display system according to claim 16, the output device includes at least one monitor having a plurality of display areas, and the display control means includes a plurality of tomographic image pairs. Each of the tomographic images is displayed in the plurality of display areas.
[0030]
  Furthermore, according to the image display system of claim 17, the display area is arranged in a matrix of m rows and n columns (m and n are natural numbers of 2 or more), and the display control means Multiple tomographic images that make up the image pairBeforeEach display is displayed in a row or a column adjacent to each other in the display area.
[0031]
  According to the image display system of claim 18, the display control means includes subtraction processing means for performing subtraction processing on at least a pair of tomographic images among a plurality of tomographic images forming the tomographic image pair, The subtraction image obtained by the subtraction process is displayed on the output device.I have to.
[0032]
  Furthermore, in order to achieve the objectClaim 19According to the described image display system, it is obtained by multiple examinations based on at least one medical imaging modality.And each set consists of multiple tomographic imagesIn an image display system configured to display a plurality of sets of three-dimensional images on an output device, from among the plurality of sets of three-dimensional images,At the time of taking the examinationAnatomical fault locationBetween different sets of 3D imagesAlmost identicalA group of tomographic imagesFirst tomogram pairAsSpecifying means for specifying at least one and the plurality of sets of three-dimensional imagesOut ofat leastA set of3D imageOf multiple tomographic imagesFault spacing andSpecified by the specifying meansBased on position information between the first tomographic image pairBeforeAnatomic fault locationBetween different sets of 3D imagesAlmost identicalA group of tomographic images as a tomographic image pairAt least one tomographic image pairUsing the plurality of sets of three-dimensional imagesTomogram pair setting means to be set;By this tomographic image pair setting meansDisplay control means for displaying at least one set of tomographic image pairs on the output device, and one arbitrary first tomogram among a plurality of tomographic images constituting the tomographic image pair displayed on the output device. Means for setting a region of interest on an image; means for designating an arbitrary second tomographic image from the plurality of tomographic images other than the first tomographic image in which the region of interest is set; Conversion means for converting an image of the region of interest in the first tomogram into an image of a portion corresponding to the region of interest on the second tomogramAndI have.
[0033]
  Furthermore, in order to achieve the objectClaim 20According to the image display method using the image display system described in the above, it is obtained by a plurality of examinations based on at least one medical image photographing modality.And each set consists of multiple tomographic imagesIn an image display method using an image display system configured to display a plurality of sets of three-dimensional images on an output device, from among the plurality of sets of three-dimensional images,At the time of taking the examinationAnatomical fault locationBetween different sets of 3D imagesAlmost identicalA group of tomographic imagesFirst tomogram pairAsSpecify at least oneFirstAnd a plurality of sets of three-dimensional imagesOut ofat leastA set of3D imageOf multiple tomographic imagesBased on the tomographic interval and positional information between the first tomographic image pairBeforeAnatomic fault locationBetween different sets of 3D imagesAlmost identicalA group of tomographic images as a tomographic image pairAt least one tomographic pairUsing the plurality of sets of three-dimensional imagesSetSecondSteps,This second stepThe set at least one tomographic image pair is sequentially displayed on the output device.ThirdStepAnd have.
[0034]
  Claims 1 to 18,19 or 20According to the image display system and the image display method using the system described in the above, it is acquired by a plurality of examinations based on at least one medical image photographing modality.Every pairFrom multiple sets of 3D images consisting of multiple tomographic images, via a designation meansBetween different pairsAt least one first tomographic image pair having substantially the same anatomical tomographic position is designated. And at least in the multiple sets of 3D imagesA set ofA 3D imageInterval between multiple tomographic imagesAnd the position information between the first tomogram pairs (for example, position coordinates based on coordinate axes set in a direction perpendicular to the tomographic plane of the three-dimensional image) by the tomogram pair setting means by the plurality of tertiary images The original imageBetween different pairsA tomographic image pair having substantially the same anatomical tomographic position is set. This tomographic image pair is displayed, for example, by the display control means.EachMultiple tomographic images that make up a layer image pairSynchronizeAnd sequentially displayed on the output device at predetermined intervals. That is, firstBetween different pairsBy specifying a pair of tomographic images with the same anatomical tomographic position,Between different pairsA pair of tomographic images having substantially the same anatomical tomographic position is automatically displayed on the output device.
[0035]
  In particular, according to the image display system according to any one of claims 6 to 7, the plurality of sets of three-dimensional images are selected by, for example, a selection unit.Using the amount of shift obtained by cross-correlation betweenSince the first tomogram pair is automatically selected and displayed via the output device, the tomogram pairs having substantially the same anatomical tomographic position are sequentially passed through the output device without the need for the operation of the reader. Displayed automatically.
[0036]
  In particular, according to the image display system of claim 9, as the tomographic image pair setting unit, a shift amount of positional coordinate information between the plurality of sets of three-dimensional images is calculated and calculated by a shift amount calculation unit. Based on the shift amount of the position coordinate information, the relative position coordinates between the three-dimensional images are obtained. An arbitrary position is designated on the relative position coordinates between the three-dimensional images by the designation means, and a tomographic image corresponding to the designated position is obtained from each three-dimensional image. At this time, the display control means displays the tomographic image groups obtained from the respective three-dimensional images on the output device as tomographic image pairs.
[0037]
  In particular, according to the image display system of claim 10, when there is no tomographic image corresponding to the designated position, at least one tomographic image at a position near the designated position is used.To create new tomographic image dataA tomographic image corresponding to the designated position is obtained by the interpolation process, and according to the image display system according to claim 11, when there is no tomographic image corresponding to the designated position, the closest to the designated position is obtained. A tomographic image of a position is selected from each three-dimensional image as a tomographic image corresponding to the designated position. Further, when there is no tomographic image corresponding to the designated position, a tomographic image of a pair of positions adjacent to the designated position on the coordinate axis is selected from each three-dimensional image as a tomographic image corresponding to the designated position. The display control means is selected so that the tomographic image of the pair corresponding to the designated position forms a group of tomographic image pairs as a whole.For each pairDisplayed on the output device.
[0038]
  According to the image display system of the thirteenth aspect, the tomographic images of the plurality of sets of three-dimensional images are acquired at substantially equal tomographic intervals. At this time, the tomogram pair setting means is based on the substantially equal tomographic interval and the positional information between the first tomogram pairs.BeforeMultiple sets of 3D imagesUsingAt least one tomographic image pair having substantially the same anatomical tomographic position is set.
[0039]
  In particular, according to the image display system according to claim 14, the output device includes a plurality of monitors arranged in a hierarchical structure, and the plurality of tomographic images forming the tomographic image pair are displayed and controlled. Each of the tomographic images is displayed on the plurality of monitors by the means so as to be adjacent to each other.
[0040]
  Furthermore, according to the image display system of claim 15, at least some of the image parameters such as the brightness, contrast, and applied image filter of the tomographic images constituting a plurality of tomographic image pairs in each monitor are the same. Is displayed on each monitor.
[0041]
  According to the image display system described in claim 16 or claim 17, the output device includes a plurality of displays arranged in a matrix of m rows and n columns (m and n are natural numbers of 2 or more), for example. At least one monitor having a region is provided, and a plurality of tomographic images forming a tomographic image pair are respectively displayed in adjacent rows or columns of the display region by the display control means.
[0042]
  In particular, according to the image display system of claim 18, for example, a pair of tomographic images out of a plurality of tomographic images forming a tomographic image pair is subjected to subtraction processing by the subtraction processing means, and obtained by the subtraction processing. The subtraction image is displayed on the output device under the control of the display control means.
[0043]
  Meanwhile, claimsRecorded in 19According to the mounted image display system, a region of interest is set on an arbitrary first tomographic image from a plurality of tomographic images (for example, tomographic image pairs) displayed on the output device, and the region of interest One second tomographic image having substantially the same anatomical tomographic position as the first tomographic image from among a plurality of sets of three-dimensional images other than the three-dimensional image including the first tomographic image set with An image is specified. Then, the image of the region of interest in the first tomographic image is converted into an image of a portion corresponding to the region of interest on the second tomographic image by the converting means.
[0044]
DETAILED DESCRIPTION OF THE INVENTION
  Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings, particularly an image display system using a CT apparatus as a medical image photographing modality.
[0045]
  FIG. 1 is a schematic block diagram showing an example of an image display system using a CT apparatus.
[0046]
  The image display system includes a CT apparatus 1, an image database 2 (hereinafter referred to as an image DB) 2 that holds image data acquired by imaging of the CT apparatus 1, a control apparatus 3 that controls the entire apparatus, The image processing devices 4a1 to 4am (m in this embodiment) and a plurality (n in the present embodiment) of display devices 5a1 to 5an are provided.
[0047]
  The CT apparatus 1 includes a gantry, a bed, and the like (not shown), and performs a CT scan on a subject (referred to as a patient in the present embodiment) conveyed in the gantry to thereby move the patient in the body axis direction. For example, three-dimensional image data composed of orthogonal tomographic image data, for example, is obtained. Furthermore, in this embodiment, in order to compare and interpret the system time change of the patient's diagnosis site, a plurality of CT scans are performed on the same patient at intervals. Then, a plurality of three-dimensional image data (A, B, C,...; Obtained as a result, the three-dimensional image data A is the most recently acquired three-dimensional image data. , 3D image data C, and so on are stored in the image DB 2 as N tomographic image data of slice numbers 1 to N, respectively.
[0048]
  The control device 3 includes a central processing unit (hereinafter referred to as a control CPU) 3a for controlling the entire display system, a main memory 3b for storing program data and processing data necessary for the processing of the control CPU 3a, and interpretation. For example, a keyboard or the like for inputting data and commands from an operator (operator). The control CPU 3a, the memory 3b, and the input unit 3c are connected to each other via a bus and exchange data with each other via the bus B1.
[0049]
  The image processing devices 4a1 to 4am are CPUs 4b1 to 4bm for performing image processing arithmetic processing and control of the entire processing device, and memories 4c1 to 4cm for storing program data and processing data necessary for the processing of the CPUs 4b1 to 4bm. And image storage units 4e1 to 4em each having a plurality of frame memories capable of storing image data at the time of image processing by the CPUs 4b1 to 4bm or image data after image processing. The processing CPU 4b1, the memory 4c1, the input unit 4d1, and the image storage unit 4e1 of the image processing device 4a1 (the same applies to the other image processing devices 4a2 to 4am) are connected to each other via a bus Ba1 (others). In the image processing apparatuses 4a2 to 4am, data is exchanged with each other via the bus Ba2 to the bus Bam).
[0050]
  As shown in FIG. 2, each of the display devices 5a1 to 5an includes a data conversion look-up table (hereinafter referred to as LUT) 5b1 to 5bn, image data holding image memories 5c1 to 5cn, a display control unit, D / A converter and display units 5d1 to 5dn including monitors M1 to Mn and the like. The display devices 5a1 to 5an are connected to each other via a bus B2. The bus B2 is connected to the connection buses Ba1 to Bam of the image processing apparatuses 4a1 to 4am.
[0051]
  The CT device 1, the image DB 2, the control CPU 3a of the control device 3, and the CPUs 4b1 to 4bm of the image processing devices 4a1 to 4am are connected to each other via a bus, and the image data and raw data ( Processing data such as projection data) and command data are transferred.
[0052]
  The control device 3 reads a plurality of tomographic image data groups from the image DB 2 by the processing of the control CPU 3a, and the image processing devices 4a1 to 4am and the display devices 5a1 to 5a corresponding to the read three-dimensional image data. In addition to performing control such as assignment of 5 an, alignment between the plurality of tomographic image data groups is performed. Note that the assignment of the processing devices 4a1 to 4am and the display devices 5a1 to 5an when a plurality of three-dimensional image data is read out is determined in advance, and the assigned data is stored in the main memory 3b. In this embodiment, the processing devices 4a1, 4a2,... Are sequentially assigned to process the plurality of 3D image data in the order in which the plurality of 3D image data is read from the image DB2. In the present embodiment, the processing devices 4a1 to 4am are assigned so as to correspond to the display devices 5a1 to 5am, respectively.
[0053]
  Further, as shown in FIG. 3A, for example, as shown in FIG. 3A, the input unit 3c of the control device 3 is a mouse capable of inputting image display commands such as an image advance command, a stop command, and a fast forward command, necessary data, and ROI. m, a keyboard K, and image designating sections I1 and I2. This image designating part I1 is a kind of selector used when two pieces of three-dimensional image data are displayed in synchronization. A lever that can be tilted in two directions (for example, left and right directions), and a tilt direction of the lever It is comprised from the detector etc. which are not shown in figure which detect this. The lever can be fixed when a predetermined inclination angle is reached. Note that a force for returning to the initial state (upright state) is applied until a predetermined inclination angle is reached.
[0054]
  Then, an image displayed on a predetermined display device (for example, 5a1, 5a2) which is a control target of the control CPU 3a of the control device 3 can be designated (selected) by the inclination direction of the lever. The state where the lever is upright represents the initial state or the synchronous display state.
[0055]
  The image designation unit I2 is used when there are more than two sets of three-dimensional images to be compared. As specific examples of the image designation unit I2, the following two are conceivable.
[0056]
  (1) It is configured like a gear of a manual car, and the operation of the 3D image A is performed at the low position, and the operation of the 3D images B, C, D, and E is similarly performed at the second, third, top and overtop respectively. Yes, all 3D images are synchronized when in the neutral position.
[0057]
  (2) In the image designation section I2 (selector) of (1), designation of chained form (including indirect cases) described later and a part of the displayed three-dimensional image can be synchronized. Since it is not possible, a selector as shown in FIG. 3B is used. This selector includes switches SM1 to SMk (corresponding to the number of monitors, which has six switches in FIG. 3B), a SHIFT key (SHIFT), a synchronization release switch C, a next tomographic image display command switch (P, N) and a trackball T, and the on / off states of these switches are sent to the control CPU 3a of the control device 3.
[0058]
  The switches SM1 to SM6 are used to select the three-dimensional images displayed on the monitors M1 to M6. When the switches SM1 to SM6 are pressed, the switches SM1 to SM6 are lit by the backlight and selected. Can be recognized by the reader.
[0059]
  The SHIFT key is a key used when designating synchronous display.
[0060]
  The synchronization cancellation key C is a key for canceling the synchronization display state.
[0061]
  The next tomogram display command switch is a switch for instructing to display the tomogram of the next frame. When “N” is pressed, the tomogram of the next frame is pressed. When “P” is pressed, the tomogram of the next frame is displayed. Display command. The trackball T is also a tomographic image display command switch, and rotates only in the front-rear direction. In the case of forward rotation, the tomographic images are sequentially turned, and in the case of reverse rotation, the tomographic images are sequentially returned.
[0062]
  The image processing apparatuses 4a1 to 4am are based on the control of each CPU 4b1 to 4bm according to a control command from the control CPU 3a, and display pixel (pixel) size, interval between tomographic images to be displayed, WW (window width), WL ( (Window level), the first display conditions such as the rotation angle of the subject with respect to the absolute coordinate system of the coordinate system, the type of filter, and the like can be set. Note that the display pixel size (example of unit: [mm / pixel]) represents a spatial distance on the real object at an interval of 1 pixel of the reconstructed image. The rotation angle of the subject relative to the absolute coordinate system is an orthogonal coordinate system (subject) determined by the horizontal right-right direction (x axis), the horizontal front direction (y axis), and the body axis direction (z axis) of the subject. It represents the tilt angle (rotation angle) of the coordinate system of the obtained tomographic image (absolute coordinate system of the specimen). The first display condition is usually predetermined in each processing apparatus 4a1 to 4am.
[0063]
  Then, the image processing apparatuses 4a1 to 4am perform image processing according to the first display condition on the respective three-dimensional image data sent according to the assignment of the control CPU 3a by the processing of the CPUs 4b1 to 4bm. It is like that. Each three-dimensional image data subjected to image processing is sent to the display devices 5a1 to 5an that are also assigned.
[0064]
  The display devices 5a1 to 5an perform image processing such as density gradation conversion (γ characteristic conversion) on input three-dimensional image data by means of LUTs 5b1 to 5bn. The contents (data) of the addresses of the LUTs 5b1 to 5bn can be rewritten in accordance with control signals from the CPUs 4b1 to 4bn of the processing devices 4a1 to 4an. The display control units of the display units 5d1 to 5dn can change various characteristics of the monitors M1 to Mn such as brightness and contrast in accordance with control signals from the CPUs 4b1 to 4bn. In the present embodiment, setting conditions such as γ characteristic conversion, brightness, and contrast set in the display devices 5a1 to 5an are referred to as second display conditions. Further, the monitors M1 to Mn of the display devices 5d1 to 5dn are integrally provided in a hierarchical structure so that their display areas (display screens) are arranged in a matrix (for example, m rows and n columns).Be taken(Refer to FIG. 4. Note that FIG. 4 shows the case where the number of monitors is six (M1 to M6) and is arranged in 2 rows and 3 columns).
[0065]
  Next, the overall operation of the image display system will be described.
[0066]
  FIG. 5 to FIG. 6 show the entire system when a pair of tomographic images having substantially the same anatomical tomographic position are synchronously displayed from a plurality of three-dimensional image data “A, B,... It is a schematic flowchart showing operation | movement.
[0067]
  Now, when comparatively interpreting tomographic images in which the anatomical tomographic positions are considered to be substantially the same among a plurality of three-dimensional image data “A, B,...” Stored in the image DB 2, according to a command from the input unit 3 c. The activated control CPU 3a reads the allocation data with reference to the main memory 3b, and also includes a plurality (for example, k (<m, n) sets) of three-dimensional image data “A (first), B,... , K (k-th) ”are sequentially read out, and the assigned data (first read 3D image data A → processing device 4a1, next read 3D image data B → processing device 4a2,...) In response, the data is sent to the corresponding processing devices 4a1 to 4ak (step 101). The CPUs 4b1 to 4bk of the processing devices 4a1 to 4ak that have received the three-dimensional image data “A to K” store the three-dimensional image data “A to K” in the image storage units 4e1 to 4ek, and each of the three-dimensional images. Processing for displaying the first tomographic image data “A1 to K1” (first slice position; that is, the first slice number) of the image data “A to K” on the corresponding display devices 5a1 to 5ak is performed. . As a result, the tomograms “A1 to K1” at the first slice positions in the three-dimensional image data “A to K” are displayed on the monitors M1 to Mk of the display units 5d1 to 5dk in the first and second displays. Displayed in a display mode based on conditions. The allocation data of the display devices 5a1 to 5ak of the respective processing devices 4a1 to 4ak are stored in the memories 4c1 to 4kk (step 102).
[0068]
  Then, the control CPU 3a determines whether or not to change the first display condition and the second display condition (step 103). If data indicating that the first display condition and the second display condition are to be changed (display condition change data) is sent from the input unit 3c, the result of the determination in step 103 is YES, and the control CPU 3a The process proceeds to step 104.
[0069]
  In the process of step 104, the control CPU 3a reads the input display condition change data and changes the display condition to each of the corresponding processing devices 4a1 to 4ak according to the allocation data stored in the main memory 3b. Send a display condition change command including the contents. The CPUs 4b1 to 4bk of the processing devices 4a1 to 4ak store the display condition change command data in the memories 4c1 to 4kk, and at the same time change the settings of the assigned monitors M1 to Mk with respect to the change of the second display condition. As for the first display condition, after the image processing corresponding to the changed first display condition is performed on the three-dimensional image data “A to K” read from the image storage units 4e1 to 4ek. Then, the images are displayed on the monitors M1 to Mk of the display devices 5a1 to 5ak (step 105).
[0070]
  If the display condition change data is not input at the time of the determination at step 103, the determination result at step 103 is NO and the process proceeds to step 106.
[0071]
  On the other hand, the radiogram interpreter sets a slice interval (display interval) and a display mode reference (collectively referred to as reference data) for synchronous display while viewing the tomographic images displayed on the monitors M1 to Mk. In this setting, a reference tomographic image (reference tomographic image) may be input (specified) by, for example, the image specifying units I1 and I2 of the input unit 3c. In the case of inputting by the image designating part I1, it can be designated by inclining to a position where the lever is not fixed in the tomographic image direction to be used as a reference.
[0072]
  In addition, the tomographic image serving as the reference for the slice interval and the tomographic image serving as the reference for the display mode (first and second display conditions) may be specified separately. Furthermore, if the display mode is different from the display mode of all tomographic images, the display conditions can be input again. In the present embodiment, the tomographic image “A1”, that is, the three-dimensional image “A” is used as the reference tomographic image.
[0073]
  The control CPU 3a reads the reference data sent in this way into the main memory 3b (step 106).
[0074]
  Further, the radiogram interpreter inputs (specifies) a plurality of three-dimensional images that are to be displayed synchronously among the tomographic images displayed on the monitors M1 to Mk, for example, by the image designating unit I2 of the input unit 3c. Comparative interpretation of tomograms with the same anatomical tomographic position as the tomogram at the desired slice position of the reference tomographic image (a group of tomographic images with the same anatomical tomographic position is called a tomographic image pair) Specify from the multiple 3D images you want.
[0075]
  If there are two sets of three-dimensional images (a three-dimensional image A displayed on the monitor M1 of the display device 5a1 and a three-dimensional image B displayed on the monitor M2 of the display device 5a2), the interpreter performs the tertiary reading for comparative interpretation. There is no need to select the original image. First, the lever of the image designating part I1 is tilted to the right (corresponding to the display device 5a1). At this time, the control CPU 3a reads the tilt direction, and recognizes the three-dimensional image A that is the subject of comparative interpretation. In this state, when the image interpreter inputs an image feed command using the return key or the like of the keyboard K, the three-dimensional image A displayed on the monitor M1 is displayed by the processing of the control CPU 3a, the processing device 4a1, and the display device 5a1. The tomographic image of 1Frame by frameIt is forwarded.
[0076]
  Then, the radiogram interpreter designates a tomographic image ALA having an arbitrary slice position (slice number), for example, a slice number LA (1 ≦ LA ≦ N) while viewing the tomographic images sequentially displayed.
[0077]
  Subsequently, the radiogram interpreter selects the three-dimensional image B (the displayed tomographic image) by tilting the lever to the left. When the lever is tilted to the left, the control CPU 3a controls the three-dimensional image B. Therefore, as in the case of the three-dimensional image A, the tomogram of the three-dimensional image B displayed on the monitor M2 is 1 when the image reader inputs an image feed command using the return key of the keyboard K or the like.Frame by frameIt is forwarded. The image interpreter designates a tomographic image having substantially the same anatomical tomographic position as that of the tomographic image ALA while viewing the tomographic images of the sequentially displayed three-dimensional image B (see FIG. 8; however, FIG. 8 shows the three-dimensional image). The relationship between “A, B” is shown). In this way, the radiogram interpreter can designate a tomographic image (that is, a tomographic image pair) substantially the same as the display tomographic image of the three-dimensional image A.
[0078]
  On the other hand, when comparing a plurality of sets of three-dimensional images, first, the radiogram interpreter selects the three-dimensional image A of the monitor M1 by pressing the switch SM1 corresponding to the reference tomographic image. Then, when the radiogram interpreter presses the next tomographic image display command switch N, the tomographic image of the three-dimensional image A displayed on the monitor M1 is processed by the control CPU 3a, the processing device 4a1, and the display device 5a1. 1In orderSent.
[0079]
  The radiogram interpreter designates a tomographic image ALA having an arbitrary slice position (slice number), for example, a slice number LA (1 ≦ LA ≦ N), while viewing the tomographic images sequentially displayed.
[0080]
  Next, the radiogram interpreter designates a three-dimensional image to be displayed synchronously. Assume that the image interpreter wants to compare and interpret the 3D image C displayed on the monitor M3 and the 3D image F displayed on the monitor M6 together with the 3D image A. At this time, the radiogram reader first selects the switch SM3 and operates the next tomographic image display command switch N and the like, and the tomographic image of the anatomical tomographic position substantially the same as the reference tomographic image ALA displayed on the monitor M1. Select (ie, tomographic image pair). When SM1 and SM3 are selected at the same time after the designation of the first tomographic image pair is finished, SM1 and SM3 are turned on (lighted) at the same time, and when SM1 is pressed, SM3 is also turned on at the same time. (That is, switch SM1 and switch SM3 are synchronized). If it is difficult to turn on simultaneously, SM1 and SM3 may be sequentially pressed while pressing the SHIFT key. In this case, the control CPU 3a determines that the keys are simultaneously pressed while the SHIFT key is pressed. Such use of the SHIFT key has a feature that three or more keys can be easily selected as compared to pressing the switch at the same time. However, as will be described later, for the purpose of comparison, when the tomographic image displayed at the position of the monitor M3 is moved to the monitor M2, M3 is simultaneously turned off and M2 is turned on.
[0081]
  Next, the switch SM6 is selected, and the tomographic image on the monitor M6 is sequentially displayed by operating the next tomographic image display command switch N and the like, and a tomographic image that allows easy alignment of an arbitrary slice position is selected.
[0082]
  Then, the switch SM3 is selected, and the tomogram on the monitor M3 is sequentially displayed by operating the next tomogram display command switch N and the like, and the tomogram substantially the same as the tomogram on the monitor M6 is selected (at this time the switch SM1 And SM3 are turned on at the same time, and the tomographic image of the monitor M1 is forwarded in synchronization with the tomographic image of the monitor M3 being forwarded, so that substantially the same tomographic image is displayed). By this processing, the three-dimensional images A, C, and F can be indirectly associated. To cancel the synchronization between the switches, the synchronization cancellation switch C is used. For example, when the switch C is pressed, if the switch SM1 is pressed, the synchronization between the switch SM1 and the other switches is released, and if no switch is pressed, all the synchronization is released.
[0083]
  In the above example, the three-dimensional image displayed on the monitor M1, the monitor M3, and the monitor M6 is selected as a target for comparative interpretation and a tomographic image pair is designated. The procedure is as follows: image C, 3D image F → 3D image C (and synchronized 3D image A). ShiBut thisThe procedure is not limited in this way. In short, any procedure may be used as long as the 3D image A, the 3D image C, and the 3D image F are indirectly associated with each other. The term “indirectly related” means that there is at least one tomographic image pair between the three-dimensional image A and the three-dimensional image C, and at least one tomographic image between the three-dimensional image C and the three-dimensional image F. This means that there is a pair (the 3D image A and the 3D image F are indirectly related via the 3D image C).
[0084]
  Therefore, for example, when the three-dimensional image data “A to F” displayed on the monitors M1 to M6 is selected as the target of comparative interpretation and the tomographic image pair is designated, the tomographic image pair with the reference tomographic image A is selected. When the method of selecting from among the three-dimensional images B to F is used, the radiogram interpreter determines that the tomographic image ALA at the first slice position LA is substantially the same as the anatomical tomographic position.Layered image, tertiaryThe original images B to F are designated while being sequentially fed. Then, when the designation of the tomographic image pair is completed or there is no tomographic image whose anatomical tomographic position is substantially the same as the first tomographic image ALA, the next desired three-dimensional image A is selected from the reference three-dimensional image A. A tomographic image ALA1 at the slice position LA1 is designated, and a tomographic image having substantially the same anatomical tomographic position as the tomographic image ALA1 is designated in the same procedure from the three-dimensional images “BF”.
[0085]
  In the following, all the three-dimensional images “B to F” required by the image interpreter and compared are tomographic images whose anatomical tomographic positions are substantially the same as the tomographic images in the designated three-dimensional image A. The tomographic images in the three-dimensional images A to F are designated so as to have at least one image (the three-dimensional images “A to F” are indirectly associated).
[0086]
  When a tomographic image pair is designated, the slice position of each tomographic image forming the first tomographic image pair (the tomographic image ALA (slice position LA) and the designated tomographic image in each of the three-dimensional images “BF”). The slice position (slice number) of the image and the slice position of each tomographic image forming another tomographic image pair are associated with each other and stored in the main memory 3b by the processing of the control CPU 3a.
[0087]
  When there are two sets of three-dimensional images to be comparatively interpreted (three-dimensional image A and three-dimensional image B) (that is, when a tomographic image pair is designated using the image designation unit I1), After the designation is finished, the image designation section I1 is brought upright (the lever is brought in the middle). Note that, as a tomographic image pair simultaneous display command to be described later, for example, when a command is input from the keyboard K, the tomographic images forming the tomographic image pair between the 3D image A and the 3D image B are synchronized. Represents the mode (synchronous mode) to be displayed (step 107).
[0088]
  Subsequently, the control CPU 3a assigns the positional relationship of the monitors M1 to M6 on which the tomographic images in the three-dimensional image data “A to F” forming the tomographic image pair are displayed, stored in the main memory 3b. Whether or not the tomographic images in the designated three-dimensional image data “A to F” are displayed vertically or horizontally adjacent to each other, or the number of adjacent tomographic images is the largest. Check whether it is. Then, it is determined whether or not the assignment of the display devices 5a1 to 5ak to the processing devices 4a1 to 4ak is to be changed (step 108). As a result, if the tomographic images in the three-dimensional image data “A to F” are not adjacent to each other or the number of adjacent tomographic images is small, the determination in step 108 is YES, and the step At 109, the assignment of the display devices 5a1 to 5ak to the processing devices 4a1 to 4ak is changed, and the result is stored in the main memory 3b. Then, an assignment change command is sent to the processing devices 4a1 to 4ak.
[0089]
  For example, if there are two designated 3D image data and each tomographic image forming a tomographic image pair in the two 3D image data is displayed on the monitors M1 and M3 (that is, two When assigned to the three-dimensional image data processing devices 4a1 and 4a3), the two tomographic images are not adjacent (see FIG. 4). Therefore, the control CPU 3a sends a command for changing the allocation of the processing devices (4a1, 4a3) to the two three-dimensional image data to, for example, (4a1, 4a2) to the processing devices 4a1 to 4a3.
[0090]
  When an assignment change command is sent, the processing devices 4a1 to 4a6 read the change command into the memories 4c1 to 4c6, and based on the changed assignment, the tomographic image data sent to the monitors M1 to M6 corresponding to the assignment. Processing for display is performed (step 109).
[0091]
  On the other hand, if the tomographic images in the three-dimensional image data “A to F” are adjacent to each other or the number of adjacent tomographic images is the maximum, the result of the determination in step 108 is NO. Proceed to step 110.
[0092]
  In the present embodiment, the three-dimensional image data “A to F” are displayed on the upper and lower sides or the left and right sides adjacent to each other, and therefore, the changing process is not performed.
[0093]
  Then, the radiogram interpreter simultaneously selects the switches SM1 to SM6 in which the tomographic image pair is designated. At this time, the switches SM1 to SM6 are simultaneously turned on (lighted), and until the switch SM1 is released, for example, if the switch SM6 is pressed, all the remaining switches SM1 to SM5 are simultaneously turned on. In this way, the three-dimensional images “A to F” can be indirectly associated, that is, synchronized. Note that the synchronization cancellation switch C is used to cancel this synchronization. For example, when the synchronization release switch C is pressed, if SM1 is pressed, the synchronization between SM1 and the other switches is released. If nothing is pressed, all synchronization is released.
[0094]
  On the other hand, the control CPU 3a sends a notification of entering the synchronous mode and reference data to the processing devices 4a1 to 4a6 (step 110).
[0095]
  Subsequently, in step 111 and subsequent steps, the control CPU 3a performs alignment processing between the tomographic directions of the three-dimensional images “A to F”.
[0096]
  That is, the control CPU 3a has two three-dimensional images (in this embodiment, a reference three-dimensional image A and an arbitrary three-dimensional image B out of the three-dimensional images “A to F” including the tomographic image pairs. The coordinates of the interval direction (that is, the body axis (z direction)) of both three-dimensional images are set as follows using the slice number and the tomographic interval (step 111). ).
[0097]
[Expression 1]
    z = (slice number) × (fault interval) (1)
[0098]
  Then, the control CPU 3a determines whether or not the number of tomographic image pairs between the 3D image A and the 3D image B is a pair (step 112). If the result is YES, in step 112, the position of the specified pair of tomographic images is calculated at the respective coordinates, and the amount of deviation between the coordinates is obtained. For example, the coordinates of the tomographic image pair of the three-dimensional image “A, B” are z_A ⁰And z_B ⁰Then, the coordinates of the three-dimensional image B are (z_B ⁰-Z_A ⁰) And the arbitrary coordinate z of the 3D image B_BIs z_A+ (Z_B ⁰-Z_A ⁰) By z_A(Step 113).
[0099]
  If there are more than one pair of tomographic image pairs, the determination in step 112 is NO, and in step 114, z determined by the above equation (1) based on the position data of a plurality of tomographic image pairs._ACoordinates and z_BPerform processing that minimizes inconsistency between coordinates, z_ACoordinates and z_BCalculate the relative displacement between coordinates. Now, the relative shift amount of these coordinates is z_eThen, the error E between the tomographic image pairs can be expressed as the following equation.
[0100]
[Expression 2]

  Here, the designated tomographic image pair is s pair, and z_A ⁱAnd z_B ⁱIs the z-coordinate of the i-th (i = 1, 2,..., S) tomographic image pair. In order to minimize the error E, “ΔE / Δz_e= 0 ”to solve equation (2), the following equation is obtained.
[0101]
[Equation 3]

[0102]
  Thus, as with the pair, z_BAny coordinate of is z_A+ Z_eBy z_AConverted to the coordinates of. This relative shift amount z_eAll the three-dimensional images (tomographic image A and tomographic image B) to be processed are stored in the main memory 3b in the control device 3 (step 114). Then, the process of the control CPU 3a proceeds to step 115.
[0103]
  The control CPU 3a repeats the above-described processing between the three-dimensional images “B, C”, the three-dimensional images “C, D”,..., And the relative coordinates of all the three-dimensional images “A to F”. Associate relationships indirectly. The “indirectly” means, for example, the z coordinate of the three-dimensional image A and the z coordinate of the three-dimensional image B, the z coordinate of the three-dimensional image B, the z coordinate of the three-dimensional image C, and the z coordinate of the three-dimensional image C. The coordinates and the z-coordinate of the three-dimensional image D,... Are finally associated with the z-coordinate of the three-dimensional image F and the z-coordinate of the three-dimensional image A, or the z-coordinate and the tertiary of the three-dimensional image A. The z-coordinate of the original image B, the z-coordinate of the three-dimensional image A and the z-coordinate of the three-dimensional image C, the z-coordinate of the three-dimensional image C and the z-coordinate of the three-dimensional image D,. This means that the original images F are associated (relatively) with each other. At this time, the control CPU 3a selects a pair of three-dimensional images that are easily associated among the three-dimensional images “A to F” based on the respective z coordinates of the three-dimensional images “A to F”. (Step 115).
[0104]
  Then, the control CPU 3a determines whether or not the relative relationships of the coordinates of all the three-dimensional images “A to F” are indirectly associated (step 116).
[0105]
  If the coordinates of the three-dimensional images “A to F” are not relatively associated, the determination in step 116 is NO and the process returns to step 115 described above to repeat the above-described processing.
[0106]
  If the determination result in step 116 is YES, the z-coordinates of all the three-dimensional images “A to F” are linked in a chain, and the process proceeds to step 117.
[0107]
  In the process of step 117, the control CPU 3a uses the amount of deviation between the z coordinates of each of the three-dimensional images “A to F” stored in the main memory 3b based on the reference three-dimensional image A. Recalculate the amount of misalignment. For example, the deviation amount of the three-dimensional image B with respect to the reference three-dimensional image A is gAB, and the deviation amount of the three-dimensional image C with respect to the three-dimensional image B is g._BC, The amount of deviation of the three-dimensional image D with respect to the three-dimensional image C is g_CD, The amount of deviation of the three-dimensional image E with respect to the three-dimensional image D is g_DEWhen the amount of deviation of the three-dimensional image F with respect to the three-dimensional image E is gEF, an equation for obtaining a corresponding position of the three-dimensional image B from an arbitrary position (z coordinate) of the reference three-dimensional image A is “z_A+ (G_AB) ”. Similarly, equations for obtaining corresponding positions of the three-dimensional images C, D, E, and F from arbitrary positions of the reference three-dimensional image A are “z”._A+ (G_AB+ G_BC) "," Z_A+ (G_AB+ G_BC+ G_CD) "," Z_A+ (G_AB+ G_BC+ G_CD+ G_DE) "," Z_A+ (G_AB+ G_BC+ G_CD+ G_DE+ G_EF) ”. It is widely and generally performed to obtain the overall relationship from each of these relative relationships. Many algorithms that can be realized on a computer have also been proposed.
[0108]
  The amount of deviation of the coordinates of the respective three-dimensional images “B to F” with respect to the coordinates of the reference three-dimensional image A thus obtained is stored in the memories 4c1 to 4c6 of the corresponding processing devices 4a1 to 4a6. The shift amount of each coordinate corresponding to the reference coordinate is in parentheses () in the above example. In other words, by adding the amount of deviation to the reference coordinates of the reference 3D image, it is possible to perform alignment between the three-dimensional images “A to F” (step 117).
[0109]
  Subsequently, the processing shifts to the processing of step 118 in FIG. 6, and the control CPU 3a indicates that an arbitrary reference slice position, that is, a tomographic image pair of z coordinates corresponding to the slice position is displayed in synchronization. The display command to represent is sent to each processing apparatus 4a1-4a6. The slice position may be determined in advance or may be specified separately from the input unit 3c. Alternatively, it may be the first slice position of the reference tomographic image A (step 118). The CPUs 4b1 to 4b6 of the processing devices 4a1 to 4a6 perform alignment between the three-dimensional images on the basis of the input slice positions and the shift amounts stored in the memories 4c1 to 4c6, respectively, and reference z coordinates (Za1). , Zb1,..., Zf1) are calculated (step 119).
[0110]
  Each of the CPUs 4b1 to 4bk determines whether tomographic image data exists at the calculated z coordinate position (Za1, Zb1,..., Zf1) (step 120). If the result of this determination is NO, that is, if tomographic image data does not exist, tomographic image data at the calculated z position (Za1, Zb1,..., Zf1) is obtained.
[0111]
  The concept in this case will be described with reference to FIG. As shown in FIGS. 9A and 9B, in the three-dimensional image, when pixels are fixed and the pixel values are plotted in the z direction, the values are recorded at regular intervals zd. That is, it becomes a problem of estimating values between sample points in a sampled state. Therefore, the value of a certain pixel position can be obtained by performing the interpolation processing shown in the literature (1) described later. By performing this process on all the pixels, tomographic image data can be newly created at the z coordinate position where there is no tomographic image data. In the example of FIG. 9B, a case where primary (linear) interpolation is used as an interpolation method is introduced.
[0112]
  That is, in step 114, the CPUs 4b1 to 4b6 of the respective processing devices have tomographic image data before and after the calculated z coordinate position (Za1, Zb1,..., Zf1) (in FIG. 9B, z0, z0 + zd, z0 + 2zd, .., Z0 + (n) zd are interpolated to create new tomographic image data corresponding to the calculated z coordinates (Za1, Zb1,..., Zf1), and the image storage unit 4c1. To 4c6 (step 121).
[0113]
  On the other hand, if the determination result in step 120 is YES, the process proceeds to step 122.
[0114]
  Then, the CPUs 4b1 to 4b6 of the respective processing devices 4a1 to 4a6 have tomographic image data “A (Za1), corresponding to the reference z coordinates (Za1, Zb1,..., Zf1) stored in the image storage units 4c1 to 4c6. B (Zb1),..., F (Zf1) ”, that is, tomogram data pairs“ A (Za1), B (Zb1),. With reference to 4c1 to 4c6, the display pixel (pixel) size, which is the first display condition, the interval between tomographic images to be displayed, WW (window width), WL (window level), and the coordinate system with respect to the absolute coordinate system of the subject Image processing is performed in accordance with set values such as the rotation angle and filter type. Note that matching the pixel sizes corresponds to making the same size object the same size on the image as shown in FIG. That is, by matching all the three-dimensional images “BF” with the three-dimensional image “A” serving as a reference for a certain display mode, all the generated three-dimensional images “A to F” have the same pixel size. Become. Further, matching the coordinate system with the reference tomographic image A corresponds to observing the same structure from the same angle as shown in FIG. In other words, data rotation processing or the like is performed so that the coordinate system of all the three-dimensional images “BF” is aligned with the coordinate system of a certain reference three-dimensional image A. However, in FIG. 11, only rotation within the tomographic plane is performed, but three-dimensional rotation that changes the inclination of the tomographic plane can be processed in the same manner. When these processes are performed together, even pixel-by-pixel comparison is possible. The method of matching the pixel size and the coordinate system with the reference data can be realized by combining a coordinate conversion technique (rotation / enlargement / reduction) and an interpolation technique widely known from the literature (1) and the like. An example of a combination of these is mentioned in document (2). Further, correction for geometric distortion in which the same processing is performed is shown in Document (3) and is widely used in general.
[0115]
  In addition, when only the rotation angle information with respect to the reference data of the coordinate system is insufficient to align the angle of the coordinate system, or when there is no information about the rotation angle with respect to the absolute coordinate system of the subject in the target three-dimensional image, The tilt angle of the coordinate system can also be estimated using information (recording, storage, etc.) at the time of data collection and anatomical knowledge (step 122).
[0116]
  Each of the CPUs 4b1 to 4b6 sends the tomographic image data pair “A (Za1), B (Zb1),..., F (Zf1)” subjected to the image processing to the assigned display devices 5a1 to 5a6 (step 123). ). Each of the display devices 5a1 to 5a6 performs image processing such as γ characteristic conversion or brightness, which is a second display condition, on the sent tomographic image data “A (Za1), B (Zb1),..., F (Zf1)”. After setting the contrast, the tomographic images “A (Za1), B (Zb1),..., F (Zf1)” are displayed on the monitors M1 to M6 at the same timing (step 124).
[0117]
  At this time, the operator operates the keyboard K of the input unit 3c or the next tomogram display command switch N (or the trackball T) of the image designating unit I2 to input a command for displaying the next tomogram pair ( Single action), or the next tomographic image pair separated by a predetermined interval Δd from the displayed tomographic image pair is synchronized from the keyboard K of the input unit 3c or the like based on a predetermined timing like a moving image. Input a command (single command) to be displayed sequentially. In response to the input single action or single command, the control CPU 3a determines the next tomogram that is a predetermined distance away from the displayed tomogram pair “A (Za1), B (Zb1),..., F (Zf1)”. A command to display the image pairs in synchronization is sent to the CPUs 4b1 to 4b6 of the processing devices 4a1 to 4a6 (step 125). In this case, the predetermined interval is the tomographic interval Δd of the reference tomographic image A. The predetermined interval Δd can be changed during interpretation. This is because it is possible to make an accurate diagnosis by closely observing the region of interest, and at the same time, roughly observing regions that are not of interest (for example, sites that are unlikely to cause disease). Thus, the interpretation time can be saved.
[0118]
  Upon receiving the next tomogram pair display command, each of the CPUs 4b1 to 4bk receives the tomogram data “A (Za2), based on the z coordinates (Za2, Zb2,..., Zf2) at positions separated from the current display image by a predetermined interval Δd. B (Zb2),..., F (Zf2) ”are read from the image memories 4e1 to 4ek at the same timing. If there is no tomographic image data at that position, the same processing as in steps 121 and 122 is performed to generate tomographic image data). The read tomographic image data “A (Za2), B (Zb2),..., F (Zf2)” is subjected to the same processing as Steps 123 and 124, and then has the same timing on the monitors M1 to M6. Is displayed (step 126).
[0119]
  Then, the control CPU 3a determines whether or not to perform tomographic image pair display at every predetermined interval Δd as it is (step 127). If a command indicating that the tomographic image pair display is to be ended is input from the input unit 3c, the result of this determination is NO, and the process ends. Further, when the next tomogram pair display command is input from the input unit 3c or in the single command mode, the process of the control CPU 3a shifts before the execution of step 118, and the process described above is repeated thereafter. The tomographic image pairs are sequentially displayed at predetermined intervals Δd.
[0120]
  For example, in the single command mode, that is, the moving image display mode, tomographic images at substantially the same anatomical positions are sequentially displayed in synchronization on the monitors M1 to M6.
[0121]
  Then, when there is no tomographic image pair display end command from the input unit 3c or tomographic image data forming the tomographic image pair, the result of determination in step 127 is NO, and the process ends.
[0122]
  As described above, according to the present embodiment, an interpreter firstBetween different sets of 3D imagesIf several pairs of tomographic images with the same anatomical tomographic position are specified, multiple 3D imagesBetween different sets of 3D images based onA tomographic image pair having substantially the same anatomical tomographic position is automatically derived (set), and the tomographic image pair is displayed synchronously at a predetermined timing. Therefore, the radiogram interpreter can perform a very difficult operation of deriving a tomographic image pair when performing comparative interpretation, that is, aligning the anatomical tomographic position with a simple operation. As a result, the burden on the interpreter is greatly reduced, and comparative interpretation can be performed very easily.
[0123]
  In addition, a plurality of tomographic image pairs can be sequentially displayed at predetermined intervals. In other words, conventionally, when sequentially displaying a plurality of tomographic image pairs,pluralAlthough it is necessary to align the anatomical tomographic position due to the difference in the tomographic interval between the tomographic image pairs, in this embodiment, it is not necessary to perform the alignment, so comparative interpretation is very easy and speedy. Can be done.
[0124]
  Further, the tomographic image pairs displayed in synchronism are all displayed in the same display mode (WW, WL, lookup table, filter type, etc.), the same pixel size, and the same coordinate system. Therefore, there is no need to make extra settings when the image interpreter performs comparative image interpretation, and the labor of the image interpreter can be greatly reduced.
[0125]
  Furthermore, when a plurality of monitors are arranged in a hierarchical structure, the tomographic image pairs displayed in synchronization are automatically displayed on the monitors at adjacent positions. In other words, the tomographic image pair is displayed at a position (monitor) that is easy to compare without any extra operation by the interpreter, so that the burden on the interpreter can be reduced. Of course, the structure of the monitor does not have to be a hierarchical structure. For example, the monitor may be arranged in a horizontal row or a vertical row, and a plurality of monitors are not included in one device, and each monitor is You may line up.
[0126]
  In the present embodiment, various modifications can be made without changing the gist of the present invention.
[0127]
  ・ Modification 1
  Although a plurality of 3D image data is read in the process of step 101 in FIG. 5, the read data may be a raw data (projection data) group instead of the 3D image data (obtained by the CT apparatus 1). Read raw data directly). In this case, in the processing of step 102, the CPUs 4b1 to 4bk reconstruct the transmitted raw data group, and display a plurality of three-dimensional image data obtained as a result of this processing on the monitors M1 to Mk. become. In this case, in the image interpolation processing in step 121 and step 126 in FIG. 6, a plurality of reconstructed three-dimensional image data may be used. Further, when the X-ray CT apparatus 1 is a three-dimensional CT such as a helical scan (also called a spiral scan or a helical scan) CT or a cone beam CT, that is, the modality that has acquired the raw data is image-reproduced three-dimensionally. In the case of having information that can be configured, a three-dimensional image of the target position (z coordinate) can be directly reconstructed from the raw data.
[0128]
  ・ Modification 2
  If comparative interpretation is performed and if the comparison target is clear in advance, the process of step 106 in FIG. 5 is performed in advance, and then the process in steps 122 and 123 in FIG. 6 is performed to display the first tomographic image. You may do it. In this case, since it is determined that all of the plurality of 3D image data to be read are to be compared, the processes of step 122 and step 123 are automatically performed.
[0129]
  ・ Modification 3
  The monitors M1 to Mn of the display devices 5a1 to 5an in FIGS. 1 and 2 may be displays or windows. The display devices 5a1 to 5an may be a system that displays a plurality of images (multi-frame display) on one or several displays or windows. In this case, each tomographic image forming the tomographic image pair is displayed in an area for displaying one frame (one image). Further, the display devices 5a1 to 5an may include a hard copy device capable of multi-frame display (printing). In this case, each tomographic image forming the tomographic image pair is displayed (printed) in an area where one frame (one image) is printed.
[0130]
  Modification 4
  In this embodiment, if the number of display devices 5a1 to 5an is larger than the number of tomographic image pairs to be synchronously displayed, the control device 3a assigns a plurality of display devices 5a1 to 5an to the respective processing devices 4a1 to 4am. A plurality of tomographic images may be displayed per examination image group (three-dimensional image).
[0131]
  For example, if the number of tomographic image pairs is two (three-dimensional image “A, B”), as shown in FIG. 12, each display device 5a1 to 5a6 (explained as six in FIG. 12) monitors M1 to M1. Among M6, the display devices 5a1 to 5a3 (M1 to M3) are assigned to the processing device 4ai (three-dimensional image A), and the display devices 5a4 to 5a6 (M4 to M6) are assigned to the processing device 4aj (three-dimensional image B). Assigned to. In this example, as can be seen from FIG. 12, since a plurality of tomographic images in a three-dimensional image can be compared at the same time, not only a two-dimensional comparison but also a three-dimensional comparison is possible, thereby improving the accuracy of comparison. Can be made.
[0132]
  ・ Modification 5
  In the third modification, the plurality of display areas of the monitors M1 to Mn are arranged in a matrix of m rows and n columns (m and n are natural numbers of 2 or more) as shown in FIGS. 13 (a) and 13 (b). (In FIG. 13, two monitors M1 and M2 have a display area of 2 rows and 2 columns).
[0133]
  At this time, the control CPU 3a displays the display positions of the tomographic image pairs displayed on the monitors M1 to Mk, and at least one of the position in the row direction and the column direction of the display area is displayed on each monitor M1 to Mk. The areas may correspond to each other. For example, as shown in FIG. 13A, a tomographic image a forming a tomographic image pair of the three-dimensional image A is displayed on the monitor M1.Left of regionWhen displayed in the upper corner (one row and one column portion, hereinafter referred to as “1-1”), the tomographic image b forming the tomographic image pair of the three-dimensional image B displayed on the monitor M2 is shown in FIG. ) As shown in the same display area “1-1” of the monitor M2 or adjacent display areas (“1 row 2 column portion;“ 1-2 ”or 2 row 1 column portion; “2-1”). 14A, the (n-1) th slice tomogram “A”, the nth slice tomogram “I”, the (n + 1) th slice tomogram “U”, and n + 2 of the three-dimensional image A. The slice-th tomogram “d” is displayed in the display areas “1-1”, “1-2”, “2-1”, “2-2” (2nd row, 2nd column portion) of the monitor M1, respectively. In this case, the tomogram “A, I, U, E” of the three-dimensional image B that forms a tomogram pair with each tomogram “A, I, U, D” is displayed on the display area “1- 1 ”,“ 1-2 ”,“ 2-1 ”, and“ 2-2 ”.
[0134]
  That is, in this modification, many tomographic image pairs are simultaneously displayed on the monitors M1 to Mk. In addition, for example, when the display areas of the monitors M1 to Mk are formed in a matrix of 2 rows and 2 columns, the tomographic images forming the tomographic image pairs are displayed in correspondence between the monitors M1 to Mk. It is displayed in an area (for example, at least one of a row and a column corresponds). That is, a plurality of tomographic image pairs can be simultaneously displayed in a state where the operator can easily perform comparative interpretation, and the accuracy of comparative interpretation can be further improved.
[0135]
  Modification 6
  In the process of designating the tomographic image pair in the process of step 107 in FIG. 5, the designation process may be automatically performed using the feature amount of the three-dimensional image. The method calculates the cross-correlation coefficient between three-dimensional images while moving one image, and the amount of shift between the three-dimensional images is the shift amount when the correlation coefficient is maximized. It is requested from. Note that the correlation calculation is widely introduced in the literature (3) and the like, and an example in which it is used as a method for estimating the shift amount between images is introduced in the literature (4). Since only the z-axis needs to be dealt with, each inspection image is projected onto a projection plane parallel to the z-axis with a parallel beam system (line integration of a three-dimensional image on a straight line perpendicular to the projection plane). For each image), a pseudo scanogram image may be created, and a two-dimensional correlation operation may be performed on the scanogram image.
[0136]
  For example, instead of step 107, the processing shown in FIG. That is, in response to a control command from the control CPU 3a, each of the processing devices 4a1 to 4a6 reads out the three-dimensional image data “A to F” from the image storage units 4e1 to 4e6 (step 107a), and the three-dimensional image data “ The pseudo scanogram data “S1 to S6” are created based on “A to F” and stored in the image storage units 4e1 to 4e6 (step 107b).
[0137]
  Subsequently, the control CPU 3a performs two-dimensional correlation calculation processing on the created pseudo-scanogram data “S1 to S6”, and extracts (specifies) a tomographic image pair group having substantially the same anatomical tomographic position. (Step 107c). Thereafter, the process proceeds to step 108 and the above-described steps 108 to 127 are performed.
[0138]
  In other words, when using the method of estimating the amount of deviation using three-dimensional correlation calculation, the accuracy of the estimated deviation amount is high, but it takes a lot of time, whereas two-dimensional correlation calculation is used for the pseudoscanogram image. The method for estimating the amount of misalignment is that the accuracy of the estimated amount of misalignment is slightly lower than that of the method for estimating the amount of misalignment using a three-dimensional correlation calculation, but it can be processed in a short time. Furthermore, the accuracy of the estimated deviation amount can be improved by changing the projection plane (projection angle) and performing the process a plurality of times (repeating the processes of steps 107b and 107c while changing the projection plane). These processes are performed by automatically selecting a combination that indirectly associates all three-dimensional images to be compared.
[0139]
  ・ Modification 7
  In Modification 6, information that greatly contributes to the accuracy of the amount of deviation estimated in the image may be emphasized or extracted in advance. For example, since one piece of information that greatly contributes to the information is boundary information, edge enhancement, edge extraction processing, or the like may be performed using a high-pass filter, a differential filter, or the like in advance. The special feature of this process is that the amount of deviation can be accurately estimated even when there is a pixel whose pixel value is not stored due to a difference in modality or collection conditions.
[0140]
  ・ Modification 8
  In the modified example 6, it is stated in many papers such as the literature (3) that the accuracy of the estimated deviation amount is not so good in a region with few features (for example, a region that is difficult to distinguish from other regions, for example, a superb region). It has been. Therefore, the correlation calculation is not performed on all the regions, and one or more regions having a large amount of features (regions that can be easily distinguished from other regions) are selected, and the result of the correlation calculation (shift amount with high correlation. This z (Representing a tomographic image pair corresponding to the shift amount in the axial direction), and considering the result in the same manner as the result designated from the input unit 3c in step 107, the steps 107, 111 to 121 are performed. You may align by a procedure.
[0141]
  The region having a large amount of feature may be designated by the operator selecting a tomographic image or ROI, or by performing a frequency analysis and selecting a region having a large amount of high-frequency components. May be specified automatically.
[0142]
  ・ Modification 9
  In the process of step 114 in FIG. 7, when there are three or more tomographic image pairs designated for alignment between the three-dimensional images, all the tomographic image pairs are not used for associating the z coordinates. The z coordinate may be associated locally using a plurality of tomographic image pairs in FIG. For example, as shown in FIG. 16A, a tomogram pair “(A1−B1), (Ak−Bk), (Ap−Bp) {1 <k <p} between three-dimensional images“ A, B ”. ”Is specified, the tomographic image pair (A1−B1) and (Ak−Bk) and the tomographic image pair (A1−B1) and ( A tomogram existing between or near Ak-Bk) is aligned. That is, as shown in FIG. 16B, the tomographic images “Af, Ag, Aj” between the tomographic images “A1 to Ak” are the tomographic images “Bf, Bg, Bj” between the tomographic images “B1 to Bk”. And the above-mentioned deviation amount Gk. Similarly, between the tomogram pair (Ak−Bk) and (Ap−Bp) based on the amount of deviation (for example, Gp) calculated between the tomogram pair (Ak−Bk) and (Ap−Bp). Alternatively, alignment of tomographic images existing in the vicinity thereof is performed. That is, as shown in FIG. 16C, the tomographic image “Am, An, Ao” between the tomographic images “Ak to Ap” is the tomographic image “Bm, Bn, Bo” between the tomographic images “Bk to Bp”. And the above-mentioned deviation amount Gp.
[0143]
  This is effective in reducing the z-coordinate deviation as much as possible when comparing images obtained by photographing subjects with different physiques, or when the posture is slightly different even when the physique is the same. .
[0144]
  Modification 10
  When creating the next tomographic image pair in the image interpolation processing in step 121 and step 126 in FIG. 6, a tomographic image at a position where the ratio of the distance to the tomographic image pair existing before and after is the same may be created. For example, as shown in FIG. 17, a tomogram pair “(A1−B1), (Ak−Bk), (Ap−Bp) {1 <k <p}” is present between the three-dimensional images “A, B”. In this case, a tomographic image at a position that internally divides the tomographic image “A1 to Ak” into “m1: n1” between the tomographic image pairs (A1−B1) and (AK−BK) is displayed. In this case, the tomographic image at a position that internally divides the tomographic images “B1 to BK” into “m1: n1” in the three-dimensional image B may be displayed. Similarly, when displaying a tomographic image at a position that internally divides the tomographic image “Ak to Ap” into “m2: n2” between the tomographic image pairs (Ak−Bk) and (Ap−Bp), three-dimensional In the image B, a tomographic image at a position that internally divides the tomographic image “Bk−Bp” into “m2: n2” may be displayed.
[0145]
  As in the case of the modification 9, the z-coordinate misalignment occurs when comparing images obtained by photographing subjects having a physique difference, or when the posture is slightly different at the time of photographing even with the same physique. Is more effective than Modification 9 especially for this purpose.
[0146]
  Modification 11
  In step 121 in FIG. 6, the tomographic image pair data to be sequentially displayed may be created in advance by generating tomographic image data at all corresponding positions (z coordinates) or the next tomographic image pair. When a display command (single action) or a command to sequentially display tomographic image pairs (single command) is input, tomographic image pair data at corresponding positions may be created. In the former case, since all tomographic image pairs are created first, it takes time to create them, and display of the tomographic image pairs cannot be started immediately. However, once the display of tomographic image pairs is started, the tomographic image pairs can be sequentially displayed at a high speed thereafter. On the other hand, the latter is the reverse of the former, and the sequential display of tomographic image pairs cannot be performed as fast as the former, but the sequential display can be started immediately, and each tomographic image pair data is stored. It has an advantage that the storage capacity of 4e1 to 4em may be small. Furthermore, in the former case, if the reference tomographic image is changed, it takes a lot of time to create an image that was not observed. In the latter case, the reference tomographic image is easily changed at the stage of displaying the tomographic image pair. be able to.
[0147]
  Modification 12
  In the processing of steps 120 to 121 in FIG. 6, in order to display tomographic image pairs one after another, when there is no tomographic image at the calculated z coordinate position, new calculation is performed using interpolation processing.
[0148]
  However, when there is no tomographic image at the position to be displayed (z coordinate position, tomographic plane), the tomographic image closest to the tomographic plane may be displayed. For example, as shown in FIG. 18, as a result of the alignment between the three-dimensional images “A, B”, “Aia-Bi (i = 1, 2,..., N)” is a tomographic image pair. When no tomographic image exists at the position of the tomographic plane Aia, the tomographic image Ai closest to the tomographic plane Aia is displayed as a tomographic image corresponding to the tomographic image Bi. In this method, it is not necessary to create a tomographic image at the calculated z-coordinate position, so that a high-speed sequential tomographic image pair display can be performed.
[0149]
  The process for displaying the tomographic image present at the closest position from the tomographic plane described above uses an interpolation process as shown in FIG. 9B, for example, when the structure of the part to be compared is fine. When the computed CT value of the z-coordinate position is obtained, it is used when the structure of that part disappears.
[0150]
  That is, when such a comparison target part is fine, a system that always displays a tomographic image that is present at the closest position without calculating a new tomographic image using interpolation processing can be used. In addition, when the comparison target portion is fine, slices before and after the calculated z coordinate can be displayed simultaneously.
[0151]
  That is, as shown in FIGS. 19 and 20, as a result of alignment between the three-dimensional images “A, B”, “Aia′−Bi ′ (i = 1, 2,..., N)” For example, when there is no tomographic image at the position of the tomographic plane A1a ′, after the result of determination in step 120 of FIG. 6 (NO), as shown in FIG. The tomographic image Ai ′ and the tomographic image A (i + 1) ′ at the z-coordinate position before and after the z-coordinate position are sequentially read out from the image storage unit 4c1 as the tomographic image corresponding to the tomographic image Bi ′, and at every predetermined timing. Simultaneous display on the monitor (step 121a). Note that Ai 'or A (i + 1)' is displayed at the same time as Bi 'in places where the display area of the monitor has only two tomographic images, and Ai' and A (i + 1) 'are displayed at regular intervals. You may display alternately. In addition, Ai 'and A (i + 1)' with a single commandAnd displayMay be switched. Also in this method, it is not necessary to create a tomographic image at the calculated z-coordinate position, so that a high-speed sequential tomographic image pair display becomes possible.
[0152]
  In a system that displays a tomographic image at a position closest to the calculated z coordinate (or a tomographic image at a position before and after the z coordinate) without using the interpolation processing described above, the arithmetic processing is only a process for obtaining the z coordinate. Since there is no need to perform interpolation processing, the processing is greatly reduced, the configuration of both software and hardware is simplified, and the system becomes very inexpensive.
[0153]
  Modification 13
  In the present embodiment, in order to display the tomographic image pairs one after another in the processing of step 111 to step 121 and step 126, the z coordinate of each of the substantially identical tomographic image pairs based on the positional information of the first tomographic image pair. And a tomographic image pair corresponding to (corresponding to) the z-coordinate position (when there is a tomographic image at that position, the tomographic image at that position is derived and set as it is, and when there is no tomographic image, interpolation processing is used. Newly calculated (derived). In other words, the determination is made using a so-called IF sentence. If a tomographic image is present at that position, the tomographic image is displayed as it is, and if there is no tomographic image, it is newly created using interpolation processing.
[0154]
  However, the interval between tomographic images has a reference value for each part between hospitals. For example, the ear is 1 mm and the lung field is 2 mm. In recent years, there has been a movement to take a three-dimensional image in a group examination, but a reference value is also set for the examination. For example, a representative example of using CT 3D images for mass screening is the lung cancer screening conducted by the Anti-Lung Cancer Association (ALCA) from Tokyo. The reference value is 10 mm. In this way, many three-dimensional inspection images are reconstructed with reference values (reference tomographic image intervals).
[0155]
  In other words, if there is almost no misalignment even if the read 3D images are displayed sequentially, for example, because each 3D inspection image is created at the same reference tomographic image interval, each 3D Once the same tomographic image pair is designated once based on the positional information between the first tomographic images of the image, the z-coordinate of each tomographic image that is substantially the same and the tomographic image corresponding to the z-coordinate position are set. Even if it is not (that is, even if the tomographic image positions are updated with each other, since the updated positions are substantially the same at every reference interval, it is not necessary to create a new tomographic image using interpolation processing) Each time an image pair display command (single action) or a next tomogram pair sequential display command (single command) is received, only the same tomographic image pair can always be displayed by updating each tomographic image to the next tomographic image.
[0156]
  For example, when the 3D image A and the 3D image B are acquired at substantially the same tomographic interval (reference tomographic image interval), the tomographic image pairs of the 3D image A and the 3D image B are displayed synchronously. . It is assumed that the 3D image A is stored in the image storage unit 4e1 of the processing device 4a1, and the 3D image B is stored in the image storage unit 4e2 of the processing device 4a2.
[0157]
  At this time, the operator designates a tomographic image A (zr) of a desired slice position of the three-dimensional image A (the z coordinate corresponding to the slice position is “Zr”) as the process of step 107 in FIG. In both cases, a tomographic image B (Zrb) having substantially the same anatomical tomographic position as that of the tomographic image A (Zr) is designated from the three-dimensional image B (FIG. 21, step 301). Then, the control CPU 3a sends instructions to display the two tomographic images “A (zr), B (Zrb)” in synchronization with each other to the processing devices 4a1 and 4a2. The CPUs 4b1 and 4b2 of the processing devices 4a1 and 4a2 perform the same processing as steps 119 to 126. As a result, the tomographic image A (Zr) and the tomographic image B (Zrb) are displayed in synchronization with the monitors M1 and M2, for example (step 302).
[0158]
  At this time, since the three-dimensional image data “A, B” is acquired at the same tomographic interval, the radiogram interpreter operates the keyboard K of the input unit 3c, the next tomographic image display command switch N, etc. to perform a single action. Alternatively, if a single command is input, a tomographic image (tomographic image pair) is selected for each tomographic interval from the three-dimensional image “A, B”. The tomographic image pairs are sequentially displayed on the monitors M1 and M2 in synchronization with each other (step 303). Thereafter, the process of step 127 is performed.
[0159]
  As described above, when the tomographic intervals are substantially the same, if a tomographic image having substantially the same anatomical tomographic position is first designated, then the tomographic image pairs can be sequentially displayed for each tomographic interval. Therefore, the processing is greatly simplified and the processing speed is greatly improved. In addition, since the processing is greatly simplified, the configuration of both software and hardware becomes simple, resulting in a very inexpensive system.
[0160]
  Modification 14
  In the processing of steps 120 to 121 in FIG. 6, when the images are sequentially displayed in synchronization, for example, the image designation unit I1 is switched in a predetermined direction or the synchronization release switch C of the image designation unit I2 is pressed. The synchronous display of some or all images can also be stopped. This is effective when it becomes necessary to individually interpret images. In addition, the tomographic plane of some displayed images may be displayed in other tables.Display imageWhen the position deviates from the position of the tomographic plane, it is also effective as a correction means for the shift of the synchronous position in which the image is made asynchronous and the position is adjusted and then the synchronous mode is set again.
[0161]
  Modification 15
  In the process of matching the type of filter in the process of step 122 in FIG. 6 between the tomographic image pairs, a filter applied at the time of image reconstruction may be considered. This is achieved by registering an MTF (Modulation Transfer Function) of another reconstruction image of the reconstruction filter and applying a filter that matches the MTF. For example, if the MTFs of the images A and B are HA (u, v) and HB (u, v), respectively, and the target objects of the images A and B are the same F (u, v), the three-dimensional image “A, Distributions FA (u, v) and FB (u, v) of “B” are obtained as follows.
[0162]
[Expression 4]
    FA (u, v) = F (u, v) HA (u, v) (4)
[Equation 5]
    FB (u, v) = F (u, v) HB (u, v) (5)
  Here, (u, v) are the coordinates of the Fourier plane, and F (u, v), FA (u, v), and FB (u, v) are all the results of Fourier transform of the original object or image. It is. In addition, for the sake of simplicity, the processing of the two-dimensional tomographic plane is described here, but the processing in three dimensions can be performed in the same manner. By the way, since FA (u, v) and FB (u, v) are originally the same target object, if the MTF is the same, the same image should be obtained. Therefore, in order to make the MTFs coincide with each other, the filter H (u, v) may be applied to FB (u, v).
[0163]
[Formula 6]
    H (u, v) = HA (u, v) / HB (u, v) (6)
  A feature of the processing that also considers the filter applied at the time of image reconstruction is that the pixel values can be directly compared in the modality in which the pixel values are stored.
[0164]
  Modification 16
  Some or all of the functions of this system may be provided in the imaging modality.
[0165]
  Modification 17
  In the process of setting the reference data in step 106 in FIG. 5, the reference tomographic image A is set from the input unit 3c of the control device 3, and the display conditions of the reference tomographic image A are held in the reference tomographic image A. The reference data may be set for each display condition of the reference tomographic image A by reading from the memory 4c1 of the image processing apparatus 4a1. A feature of this processing is that it is possible to automate the setting of a very large amount of reference data.
[0166]
  Modification 18
  Rotation processing that matches the coordinate system of each tomographic image forming the tomographic image pair with the coordinate system of the reference tomographic image, particularly rotation of the tomographic plane (rotation that makes the tomographic plane of the tomographic image pair parallel) does not have to be performed. That is, when the angle formed between the reference tomographic plane (xy plane) and the tomographic plane to be compared is small (close to parallel), or the region of interest (ROI: Region of inteIf the rest is fixed, just align the part of the area to be compared (for example, the center), and the misalignment of the other areas to be compared is not so large. The processing for making the tomographic plane of the comparison target three-dimensional data parallel to the reference tomographic plane (rotation processing to change the inclination of the tomographic plane) may be omitted. The feature of this method is that the time-consuming process of three-dimensional interpolation can be omitted, and the process can be performed at high speed.
[0167]
  That is, when the amount of deviation from the reference tomographic image is small or the ROI is determined, it can be processed at high speed without the time-consuming process of three-dimensional interpolation. In the latter case, the processing method includes setting a region of interest on a tomographic plane of an arbitrary three-dimensional image, calculating a region corresponding to the region of interest in the three-dimensional image to be compared, and then including many corresponding regions. The tomographic plane may be displayed, or the coordinates may be related by aligning the ROI while ignoring the mismatch of the coordinates in the processing of step 101 to step 105 in FIG.
[0168]
  ・ Modification 19
  In this embodiment, the tomographic image pairs are displayed on different monitors for each frame. However, the present invention is not limited to this. For example, a part of a tomographic image constituting the tomographic image pair is transferred to another tomographic image. It is also possible to convert to a tomographic image constituting the image pair.
[0169]
  Now, assuming that k sets of three-dimensional image data are respectively stored in the image storage units 4e1 to 4ek of the processing devices 4a1 to 4ak, the processing of the processing device 4a1 is performed by the processing of the control CPU 3a based on the operation of the operator first. It is assumed that the three-dimensional image data A stored in the image storage unit 4e1 is sequentially displayed on the monitor M1.
[0170]
  At this time, it is assumed that the image interpreter wants to interpret a certain part on the currently displayed tomographic image by comparing it with another three-dimensional image while viewing the displayed three-dimensional image A. Here, the radiogram interpreter operates the keyboard K of the input unit 3c to stop (freeze) the currently displayed tomographic image A (Za1). Then, by operating the keyboard K or the like, the region to be compared is designated by ROI, and the three-dimensional image to be compared and interpreted (the processing device storing the three-dimensional image, for example, 4a2) is designated (FIG. 22). Step 401). At this time, the control CPU 3a reads the designated ROI and the processing device 4a2, reads the 3D image data A (Za1) from the 3D image data stored in the image storage unit 4e1, and also stores the image storage unit. The three-dimensional image data (tomographic image pair) B (Zb1) having the same z coordinate (Za1) and anatomical tomographic position is read out from the three-dimensional image data B stored in 4e2 (step 402). Then, the control CPU 3a erases the data (pixel value) in the area designated by the ROI in the read A (Za1), and the corresponding area of the 3D image data B (Zb1) read out to that area. Data (pixel value) is synthesized (step 403). Then, the control CPU 3a sends a command to display the synthesized 3D image data Ab (Za1) obtained as a result to the processing device 4a1. The CPU 4b1 of the processing device 4a1 displays the image data Ab (Za1) on the monitor M1 based on the command. As a result, as shown in FIG. 23, the monitor M1 displays a three-dimensional image Ab (Za1) obtained by synthesizing another three-dimensional image B (Zb1) to be compared and interpreted only in the region within the designated ROI. (Step 404).
[0171]
  If the keyboard K of the input unit 3c is further operated to move the ROI or send a command such as “composite on / composite off”, the former displays the synthesized portion in the original 3D image A. Return, set the ROI in the region translated by the designated movement amount (the same processing as in step 401), repeat the processing from step 402 onward, and the region corresponding to the newly designated ROI in the 3D image B Are combined and displayed. In the latter case, when OFF is pressed, the combined portion is returned to the original 3D image A. When ON is pressed again, the 3D image B is again combined with the designated ROI. To do. By moving the ROI to the left or right in the vicinity of the region of interest or repeatedly turning on / off the composition in the region of interest, the difference between the three-dimensional images A and B in the region of interest (changes in shade or presence) can be clearly identified. Therefore, accurate interpretation can be performed.
[0172]
  Note that the areas to be compared are not partial, and for example, a method of designation such as half the monitor screen is possible. That is, the operator operates, for example, the mouse m of the input unit 3c to specify boundary line data (for example, a line marker that equally divides the display area into left and right at the center of the screen). Then, based on the boundary line data, the control CPU 3a leaves the original three-dimensional image data A (Za1) on the left side of the boundary line position on the three-dimensional image data A (Za1), for example, On the right side, the right half of the three-dimensional image data B (Zb1) is synthesized. As a result, as shown in FIG. 24, different three-dimensional images at the same anatomical tomographic position are displayed in a synthesized state with the center of the monitor M1 as a boundary.
[0173]
  Further, by operating the mouse m or the like in this state and moving the line marker to the left or right, for example, the difference between the three-dimensional images A and B (changes in shade or presence) can be clearly discriminated with the line marker as a boundary. Therefore, accurate interpretation can be performed.
[0174]
  Furthermore, as an example of the development of this concept, an image obtained by another modality can be combined with an area designated by ROI (or one area divided by a boundary line).
[0175]
  According to these modified examples, the radiogram interpreter can interpret the tomographic image pair without moving the viewpoint. In addition, since the positions to be compared can be easily matched, the interpretation efficiency can be greatly increased. Furthermore, it has a feature that comparative interpretation is possible even with a display device that can display only one tomographic image.
[0176]
  Modification 20
  In step 122 of FIG. 6, each of the CPUs 4b1 to 4b6 simply displays the same z-coordinate tomographic image data (tomographic image pair data) on the monitors M1 to M6 at the same timing, but the present invention is limited to this. For example, a plurality of (for example, two) tomographic image data (one set of tomographic image pair data) is selected from each tomographic image data forming a tomographic image pair, and the two tomographic image data are subtracted. Data (subtraction data) obtained by processing can also be displayed.
[0177]
  When performing this subtraction processing, the radiogram interpreter inputs two reference images in step 106 of FIG. 5 to display two three-dimensional images that the subtraction processing images are to be displayed (the two three-dimensional images are stored). For example, 4a1, 4a2) and a monitor (for example, M1) for displaying the subtraction image.
[0178]
  On the other hand, of the CPUs 4b1 to 4b6 that have finished the process of step 122 in FIG. 6, the CPUs 4b1 and 4b2 perform the process shown in FIG. 25 instead of the process shown in FIG. That is, the CPUs 4b1, 4b2 read tomographic image data (tomographic image pair data) “A (Za1), B (Zb1)” from the image storage units 4e1, 4e2, respectively. Then, the CPU 4b2 sends the read tomographic image data B (Zb1) to the CPU 4b1 (step 501). The CPU 4b1 performs subtraction processing on the tomographic image data B (Zb1) from the tomographic image data A (Za1) (step 502). Then, the obtained subtraction image data {(A (Za1) -B (Zb1)); referred to as SU1} is sent to the display device 5a1. The display device 5a1 performs image processing such as γ characteristic conversion, which is the second display condition, and brightness, contrast, and the like on the sent subtraction image data “SU1”, and then sets the subtraction image data “SU1”. The information is displayed on the monitor M1 (step 503).
[0179]
  At this time, since processing similar to step 125 to step 127 in FIG. 6 is performed, the CPUs 4b1, 4b2 are separated from the tomographic image data “A (Za1), B (Zb1)” subjected to subtraction by a predetermined interval Δd. The tomographic image data A “(Za2), B (Zb2)” is read from the image memories 4e1 and 4e2 at the same timing, and the subtraction process shown in steps 501 to 503 is performed. Thereafter, the subtraction images {“SU2”, “SU3”,...} Are sequentially displayed on the monitor M1 in accordance with the processing of step 120 (step 504).
[0180]
  According to this modification, not only the tomographic image pairs having the same anatomical tomographic position can be displayed synchronously, but also the subtraction images of the tomographic image pairs can be sequentially displayed. By doing so, it is possible to examine in detail the change state of the diagnosis target site of the patient.
[0181]
  Modification 21
  In the above-described combining process of the modification 19 and the subtraction process of the modification 20, it is desirable that the positional relationship in the slice in the tomographic image pair to be combined or subtracted is substantially the same. Therefore, when the slice position is slightly deviated between the tomographic image pairs, the correlation calculation process is performed in addition to the processes of FIGS. 22 and 25 described above, thereby reducing the image difference due to the positional relationship deviation within the slice. Correction is performed (the pixel position of each part is matched in the corresponding tomographic image pair), and more accurate synthesis processing and subtraction processing can be performed. The method of correcting the image difference due to the positional shift in the slice is the method of correcting the overall positional shift and the local correlation calculation to correct the partial shift due to body movement, and the result is also obtained. In addition, there are methods for partially deforming the object, and these methods are known from the reference (4).
[0182]
  That is, in the synthesis process (or subtraction process), after the process in step 402 in FIG. 22 (or after the process in step 501 in FIG. 25), the control CPU 3a performs tomographic image data A (Za1) as shown in FIG. And tomogram data B (Zb1) are subjected to a correlation calculation (step 601). Then, the control CPU 3a calculates the shift amount between the tomographic image data A (Za1) and the tomographic image data B (Zb1) based on the correlation calculation result (step 602). Then, based on the shift amount and the tomographic image data B (Zb1), the tomographic image data Ba (Zb1) having a very high correlation with the tomographic image data A (Za1) (considered as the same slice position) is generated. Then, it is stored again in the image storage unit 4e2 (step 603). Thereafter, the processing after step 403 is performed. However, the newly created tomographic image data Ba (Zb1) is used as the tomographic image data B (Zb1).
[0183]
  According to the present modification, even when the synthesizing process or the subtraction process is performed using the tomographic image pairs whose slice positions are deviated, an accurate synthesizing process or subtraction process can be performed.
[0184]
  Modification 22
  The system configuration of the image processing apparatus according to the present embodiment is not limited to the configuration illustrated in FIG. 1. For example, as illustrated in FIG. 27, one of the processing apparatuses 4 a 1 to 4 am illustrated in FIG. The apparatus 10a1) may be configured to also serve as the control apparatus 3 (a new input unit 3c is added), or, as shown in FIG. 28, the processing apparatuses 4a1 to 4am are eliminated, and the control apparatus 3 is processed by the processing apparatuses 4a1 to 4am. It may also serve as. In this case, the control device 3 changes all the processing performed by the processing devices 4a1 to 4am. The systems as shown in FIGS. 25 and 26 may be slower in processing speed than the system shown in FIG. 1, but can be realized at a relatively low cost, and thus have a feature that is easy to put into practical use.
[0185]
  Modification 23 Since the order in which the processes of Step 108 to Step 109 and Step 107 of FIG. 5 are performed is finally the same, it may be set in any way, but Step 108 to Step 109 of FIG. It is desirable to set so that the above process can be easily performed. When the display conditions of the tomographic image pairs, for example, the pixel pitch is greatly different, it may be difficult to perform alignment. In such a case, it is easier to perform alignment if the processing in step 107 is performed in advance.
[0186]
  In the present embodiment and the modification, the processing devices 4a1 to 4am may each have one monitor M1 to Mm. In addition, the image processing apparatus has a separate high-speed arithmetic device (microprocessor) capable of performing various image processing at high speed, and performs the processing performed by the control CPU 3a and the processing devices 4a1 to 4am described above. If necessary, the high speed arithmetic device may be used.
[0187]
  Furthermore, the composition calculation, subtraction calculation, correlation calculation, and the like described in the above-described modification are performed by the control CPU 3a or the CPUs 4b1 to 4bm of the processing devices 4a1 to 4am, but the present invention is limited to this. For example, a digital logic circuit that performs a composite operation, a digital logic circuit that performs a subtraction operation, and a digital logic circuit that performs a correlation operation are separately provided to perform the above-described composite operation, subtraction operation, and correlation operation. Also good.
[0188]
  Furthermore, in the present embodiment, a CT apparatus is used as a medical image capturing modality, but the present invention is not limited to this, and may be, for example, an MRI apparatus, or a CT apparatus and an MRI apparatus. And a plurality of sets of three-dimensional images obtained by these apparatuses may be comparatively interpreted.
[0189]
  In addition, the reference literature name referred in the Example is published below.
(1) “Television Image Engineering Handbook” edited by the Television Society, Ohmsha.
(2) “High Speed Display through Hybrid Processing” Optics Letters, Vol. 15, No. 10, 565-567 (1990).
(3) “Digital Image Processing” by Kak Rosenfeld, translated by Nagao Shin, Modern Science.
(4) “Digital Image Subtraction of Temporally Sequential Chest Image for Detection of Interval Change”, A. Kano et al., Medical Physics, Vol. 21, No. 3, 453-461 (1994).
[0190]
【The invention's effect】
  As described above, according to the present invention, from a plurality of sets of three-dimensional images,Between different groupsSince at least one tomographic image pair having substantially the same anatomical tomographic position is set, and further, the tomographic image pair can be displayed in synchronization with a predetermined timing, for example.In different 3D imagesThe tomographic image pair can be displayed without performing a very difficult operation of aligning the anatomical tomographic position. Therefore, the burden on the interpreter at the time of comparative interpretation is reduced, and the time and cost (such as personnel costs of the interpreter) for comparative interpretation are reduced. A secondary effect of this effect is that the frequency of comparative interpretation increases and the accuracy of diagnosis is further improved.
[0191]
  In addition, since a plurality of tomographic image pairs can be sequentially displayed without aligning the anatomical tomographic position due to the difference in tomographic interval between the plurality of tomographic image pairs, the time and cost for comparative interpretation can be reduced. Decrease.
[0192]
  Further, at least a part of image parameters such as brightness, contrast, and applied image filter of a plurality of tomographic images (display images) constituting the tomographic image pair can be converted without extra operation of the reader. Since it is possible to set the same among multiple tomographic images, it is not necessary for the reader to perform the comparative reading, and it is not necessary to adjust the above image parameters one by one, and the accuracy of the comparative reading is reduced while greatly reducing the labor of the reader. Can be improved.
[0193]
  Furthermore, when a plurality of monitors are arranged in a hierarchical structure, at least a part of the plurality of tomographic images constituting the tomographic image pair displayed on the monitors is a monitor at an adjacent position, that is, a comparison. The tomographic image pairs are automatically displayed at positions where they can be seen adjacent to each other. Therefore, an extra effort such as changing the display position of the tomographic image pair once displayed becomes unnecessary, and the accuracy of comparative interpretation can be improved while greatly reducing the labor of the interpreter.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a schematic configuration of an image display system according to the present invention.
2 is a block diagram showing a schematic configuration of the display device in FIG. 1;
3A is a perspective view illustrating a schematic configuration of an input unit, and FIG. 3B is a diagram illustrating an example of a configuration of an image specifying unit I2.
FIG. 4 is a perspective view showing a plurality of monitors arranged in a hierarchical structure.
FIG. 5 is a schematic flowchart showing an example of the operation of the entire image display system when displaying a pair of tomographic images having substantially the same anatomical tomographic position in synchronization.
FIG. 6 is a schematic flowchart showing an example of the operation of the entire image display system when displaying tomographic image pairs having substantially the same anatomical tomographic position in synchronization.
FIG. 7 is a schematic flowchart showing an example of the operation of the entire image display system when displaying tomographic image pairs having substantially the same anatomical tomographic position in synchronization.
FIG. 8 is a diagram showing the concept of tomographic image pairs.
FIG. 9 is a diagram for explaining a concept when generating tomographic image data of calculated z coordinates.
FIG. 10 is a diagram for explaining a concept of processing for matching pixel sizes.
FIG. 11 is a diagram for explaining the concept of processing for matching coordinate systems.
FIG. 12 is a diagram illustrating a tomographic image pair displayed on each monitor when a plurality of display devices are assigned to the processing device.
FIG. 13 is a diagram showing a tomographic image pair displayed at the same position in a plurality of display areas of a plurality of monitors;
FIG. 14 is a view showing tomographic image pairs displayed at the same position in a plurality of display areas of a plurality of monitors;
FIG. 15 is a schematic flowchart for explaining an example of a two-dimensional correlation calculation process for a scanogram image.
FIG. 16 is a conceptual diagram illustrating processing for associating z-coordinates locally using a plurality of tomographic image pairs.
FIG. 17 is a conceptual diagram illustrating a process of creating tomographic images at positions where the ratio of the distance to a pair of tomographic images existing before and after is the same.
FIG. 18 shows the z coordinate position to be displayed.ClosestThe conceptual diagram explaining the process which displays the tomographic image which exists in z coordinate position.
FIG. 19 is a conceptual diagram illustrating processing for simultaneously displaying tomographic images existing at z-coordinate positions before and after the z-coordinate position to be displayed.
FIG. 20 is a schematic flowchart for explaining an example of processing for simultaneously displaying tomographic images existing at z-coordinate positions before and after the z-coordinate position to be displayed.
FIG. 21 is a schematic flowchart for explaining an example of processing of the entire system when two sets of three-dimensional images acquired at substantially the same tomographic interval are displayed synchronously.
FIG. 22 is a schematic flowchart for explaining an example of processing of the entire system when a part of a tomographic image constituting a tomographic image pair is converted into an image of a tomographic image constituting another tomographic image pair;
FIG. 23 is a view showing a tomographic image obtained by synthesizing other tomographic images only within the designated ROI.
FIG. 24 is a diagram showing a tomographic image in which two different tomographic images are combined with a boundary interposed therebetween.
FIG. 25 is a schematic flowchart for explaining an example of the operation of the entire system when performing subtraction processing;
FIG. 26 is a schematic flowchart for explaining an example of the operation of the entire system when performing correlation calculation processing;
FIG. 27 is a schematic block diagram showing another configuration example of the image display system of the present invention.
FIG. 28 is a schematic block diagram showing another configuration example of the image display system of the present invention.
[Explanation of symbols]
1 CT device
2 Image DB
3 Control device
3a CPU for control
3b Main memory
3c input section
4a1-4am image processing device
4b1 ~ 4bm CPU
4c1-4cm memory
4e1-4em image storage
5a1-5an display device
5b1 ~ 5bn LUT
5c1-5cn image memory
5d1-5dn display section
M1-Mn monitor

Claims (20)

Image display obtained by displaying a plurality of sets of three-dimensional images obtained by a plurality of examinations based on at least one medical image photographing modality and each set including a plurality of tomographic images on an output device In the system,
From among the plurality of sets of three-dimensional image, at least a group of tomographic images anatomical tomographic position at the time of photographing of the test is substantially identical between the different sets of the three-dimensional image as the first tomographic image partner A designation means to designate one;
Based on the position information between at least one pair of said specified by a plurality of tomographic interval and said specifying means of the tomographic images constituting the three-dimensional image a first tomographic image pair of the plurality of sets of three-dimensional images, before A tomographic image in which at least one tomographic image pair is set using the plurality of sets of three-dimensional images, with a group of tomographic images that are substantially the same between groups of three-dimensional images having different anatomical tomographic positions. Pair setting means;
Image display system characterized by comprising a display control means for displaying at least one tomographic image partner set in the output device by the tomographic image partner setting means. The image display system according to claim 1, wherein the tomographic image pair setting unit is configured to automatically derive a tomographic image pair having substantially the same anatomical tomographic position. Wherein the display control unit, the image display system according to claim 2, wherein configured so as to display a plurality of the output device to synchronize the tomographic images constituting each tomographic image partner. The tomographic image pair setting unit is configured means to set a plurality of the tomographic image partner, wherein the display control unit, sequential said plurality of tomographic image partner at predetermined intervals, to be displayed on the output device The image display system according to claim 3 configured as described above. The designation means includes first specifying means for specifying a set of three-dimensional image to be a reference among the plurality of sets of three-dimensional images, and at least one tomographic image in the three-dimensional image becomes the reference claims the anatomical tomographic position with the rest of the set of three-dimensional three-dimensional image of tomographic images all the hand in the image and a second designating means for designating to be associated indirectly are substantially the same 3. The image display system according to 3. The image display system according to claim 3, wherein the designation unit includes a selection unit that automatically selects the first tomographic image pair from the three-dimensional image. The image display system according to claim 6, wherein the selection unit selects the first tomographic image pair using a shift amount obtained by cross-correlation between the plurality of sets of three-dimensional images as a feature amount. .   The image display system according to claim 3, wherein the position information is position coordinate information based on a coordinate axis set in a direction perpendicular to a tomographic plane of the three-dimensional image. The tomographic image pair setting means includes a means for calculating a shift amount of position coordinate information between the plurality of sets of three-dimensional images, and a relative position between the three-dimensional images based on the shift amount of the position coordinate information. acquiring obtaining means for obtaining the coordinates, and designating means for designating an arbitrary position in a relative position on the coordinates between the respective three-dimensional image, a tomographic image corresponding to the specified position from the respective three-dimensional image 9. The image display system according to claim 8, wherein the display control means displays tomographic image groups respectively obtained from the three-dimensional images on the output device as tomographic image pairs. When there is no tomographic image corresponding to the designated position, the acquisition unit obtains the designated position by interpolation processing for newly creating tomographic image data using at least one tomographic image near the designated position. The image display system according to claim 9, which is means for obtaining a corresponding tomographic image. When there is no tomographic image corresponding to the designated position, the acquisition unit selects a tomographic image at a position closest to the designated position from among the three-dimensional images as a tomographic image corresponding to the designated position. The image display system according to claim 9, which is a means. When there is no tomographic image corresponding to the designated position, the acquiring means uses a pair of tomographic images adjacent to the designated position on the coordinate axis as each tomographic image corresponding to the designated position. The display control means displays the pair of tomograms corresponding to the designated position on the output device for each pair so as to form a group of tomogram pairs as a whole. The image display system according to claim 9. Wherein the tomographic images of the plurality of sets of three-dimensional image is acquired in approximately equal fault distance, the tomographic image pair setting means, before on the basis of the positional information between the substantially equal tomographic interval and the first tomographic image partner Symbol The image display system according to claim 1, wherein at least one tomographic image pair having substantially the same anatomical tomographic position is set using a plurality of sets of three-dimensional images. Wherein the output device comprises a plurality of monitors arranged in a hierarchical structure, wherein the display control unit, at least in those plurality of tomographic images plurality of tomographic images to the plurality of monitors that form the tomographic image pair The image display system according to claim 3, wherein a part of the images is displayed adjacent to each other. Wherein the display control unit, the brightness of the plurality of tomographic images constituting the tomographic image pairs in each monitor, contrast, and, in at least a portion of the image parameters, including image filters applied to the same person plural The image display system according to claim 14, wherein a tomographic image is displayed on each monitor.   The output device includes at least one monitor having a plurality of display areas, and the display control unit displays a plurality of tomographic images forming the tomographic image pairs on the plurality of display areas, respectively. Item 4. The image display system according to Item 3. The display area, m rows and n columns (m, n is a natural number of 2 or more) are arranged in a matrix of, the display control unit, a plurality of tomographic images constituting the tomographic image pairs before Symbol display area The image display system according to claim 16, wherein each image is displayed in a row or a column adjacent to each other.   The display control means includes subtraction processing means for performing subtraction processing on at least a pair of tomographic images among a plurality of tomographic images forming the tomographic image pair, and the output device displays a subtraction image obtained by the subtraction processing. The image display system according to claim 3, wherein Image display obtained by displaying a plurality of sets of three-dimensional images obtained by a plurality of examinations based on at least one medical image photographing modality and each set including a plurality of tomographic images on an output device In the system,
From among the plurality of sets of three-dimensional image, at least a group of tomographic images anatomical tomographic position at the time of photographing of the test is substantially identical between the different sets of the three-dimensional image as the first tomographic image partner A designation means to designate one;
Based on the position information between at least one pair of said specified by a plurality of tomographic interval and said specifying means of the tomographic images constituting the three-dimensional image a first tomographic image pair of the plurality of sets of three-dimensional images, before A tomographic image in which at least one tomographic image pair is set using the plurality of sets of three-dimensional images, with a group of tomographic images that are substantially the same between groups of three-dimensional images having different anatomical tomographic positions. Pair setting means;
Display control means for displaying on the output device at least one tomographic image pair set by the tomographic image pair setting means ;
Means for setting a region of interest on an arbitrary first tomographic image from among a plurality of tomographic images constituting a tomographic image pair displayed on the output device;
Means for designating an arbitrary second tomographic image from the plurality of tomographic images other than the first tomographic image in which the region of interest is set;
The image display system, characterized in that the image of the region of interest portion of the first tomographic image, comprising a converting means for converting an image of a portion corresponding to the region of interest on the second tomographic image . Image display obtained by displaying a plurality of sets of three-dimensional images obtained by a plurality of examinations based on at least one medical image photographing modality and each set including a plurality of tomographic images on an output device In an image display method using a system,
From among the plurality of sets of three-dimensional image, at least a group of tomographic images anatomical tomographic position at the time of photographing of the test is substantially identical between the different sets of the three-dimensional image as the first tomographic image partner A first step that specifies one;
Based on the position information between the plurality of sets of the plurality of tomographic interval and the first tomographic image partner of the tomographic images constituting the at least one set of three-dimensional image of the three-dimensional image, the previous SL anatomical tomographic position A second step of setting at least one tomographic image pair using the plurality of sets of three-dimensional images, with a group of tomographic images that are substantially identical between different sets of three-dimensional images as a tomographic image pair;
At least one third step and an image display method using the image display system characterized by comprising a a tomographic image pair is sequentially displayed on the output device that is set by the second step.

Images