JP3926574B2 - Tomography equipment - Google Patents

Description

【００３１】
ただし、ｆ（ｒ）は、再構成される立方体（三次元ボリュームデータ）の位置ｒでの画素データである。Ｙ（ｖｒ），Ｚ（ｖｒ）は、位置ｒの画素がフラットパネル型Ｘ線検出器Ｄの検出面上に投影される点の座標であり、前記の小文字のｖは「ベクトル」を意味しており、以下、適宜に小文字の「ｖ」でもってベクトルを表すものとする。ＰΦ は、投影角度Φでのフラットパネル型Ｘ線検出器Ｄの検出面上の投影データである。ｇ_y は、Filtered Back Projectionのフィルタ関数と呼ばれ、後述する｜ω｜（絶対値オメガ）フィルタ関数である。Ｗ₁ ，Ｗ₂ は、ビームの広がりの影響を補正するための係数であり、Ｗ₁ は、後述する第１の重み付け処理に関する係数であり、Ｗ₂ は、後述する第２の重み付け処理に関する係数である。
【００３２】
画像処理部５１は、図１に示すように、撮影により得られた一群の投影データを個別に所定の第１の重み付け処理を行なう第１の重み付け処理部５３と、この第１の重み付け処理後の各投影データに対して所定のコンボリューション処理を施すコンボリューション処理部５４と、このコンボリューション処理後の各投影データに対して所定の第２の重み付け処理を行なう第２の重み付け処理部５５と、この第２の重み付け処理した後の投影データを個別に所定の逆投影（バックプロジェクション：ＢＰ）処理してＢＰ像（３次元ボリュームデータ）を生成する逆投影処理部５６とを備えている。
【００３３】
第１の重み付け処理部５３は、撮影により得られた一群の投影データを個別に所定の第１の重み付け処理を行なう。具体的には、図７に示すように、走査各位置でフラットパネル型Ｘ線検出器Ｄで検出された各投影データに対して、フラットパネル型Ｘ線検出器Ｄの各画素行ｉごとにビュー方向の画素検出レベル変動を補正する。なお、図７に示すように、被検体Ｍに向けてＸ線管Ｒから照射されるコーンビーム状のＸ線の中心点が、常に、被検体Ｍの特定断層面の中心点Ｏ（回転中心軸Ｃ上の点でもある）を透過して、フラットパネル型Ｘ線検出器Ｄの検出面の中心点に垂直に入射されるようになっている。
【００３４】
続いて、第１の重み付け処理部５３は、図７に示すように、投影データに対して、次に示す式（４）に基づく重み付け処理を行なう。
cos θ＝ＳＩＤ／（ＳＩＤ² ＋Ｙj² ＋Ｚj² ）^1/2 …… （４）
つまり、各画素に式（４）のcos θをかけて重み付け処理を行なう。例えば、画素Ｄijは、Ｙj ・cos θとすることで、重み付け処理後の画素値が求められる。これはビューによらず一定であるので、重み付け処理として予め作られている。このようにして、第１の重み付け処理後の投影データを算出する（図８には、「第１の重み付け処理後投影像：ＳＣ（ｉ，ｊ）」として図示している）。
【００３５】
コンボリューション処理部５４は、第１の重み付け処理後の各投影データ、すなわち、第１の重み付け処理後投影像：ＳＣ（ｉ，ｊ）に対して所定のコンボリューション処理を施す。実空間で行なうコンボリューション処理は、フーリエ空間で行なうフィルタリング処理と同等であるので、ここでは説明の便宜上、上述の所定コンボリューション処理を、フーリエ空間で行なうフィルタリング処理（図８に示す｜ω｜（絶対値オメガ）フィルタリング処理）として説明するものとする。なお、以下にコンボリューション処理部５４での｜ω｜フィルタリング処理について説明する。
【００３６】
コンボリューション処理部５４は、フラットパネル型Ｘ線検出器Ｄのｉ行ごとに横方向に１次元フーリエ変換を行ない、フーリエ面像ＳＣＦ（ｉ，ω）を生成する１次元フーリエ変換部と、１次元フーリエ変換したフーリエ面像ＳＣＦ（ｉ，ω）に対して｜ω｜フィルタリングを施すフィルタリング部と、このフィルタリング部でフィルタリングした後のフーリエ面像ＳＣＦ´（ｉ，ω）を１次元逆フーリエ変換して実空間データに戻す１次元逆フーリエ変換部とを備えている。
【００３７】
フィルタリング部は、図８に示すように、１次元フーリエ変換したフーリエ面像ＳＣＦ（ｉ，ω）のｉ行方向に等方的に高周波領域を低減して高周波ノイズ分を抑制するフィルタとデータ収集走査形態に依存するフィルタとにより構成される｜ω｜フィルタリング部を備えている。なお、上述のデータ収集走査形態に依存するフィルタは、フィルタリング後のフーリエ面像ＳＣＦ´（ｉ，ω）を１次元逆フーリエ変換する際に、直流成分が強調されて生成されるのを抑制しており、直流成分が強調されることに起因する偽像を低減しているのである。
【００３８】
ここで、１次元フーリエ空間でフィルタリング処理を行なうことの意味合いについて説明する。１次元フーリエ空間でフィルタリング処理を行なうことは、数学的には次に示す式（５）で示される。なお、ＳＣＦ´（ｉ，ω）はフィルタリング処理された後の１次元フーリエ面像であり、Ｍ（ωi ）は上述したフィルタリング部のフィルタ特性を示す関数である。
ＳＣＦ´（ｉ，ω）＝ＳＣＦ（ｉ，ω）×Ｍ（ωi ） … （５）
【００３９】
なお、Ｍ（ωi ）は、前述の２個のフィルタ特性を表す関数の積として次に示す式（６）のように表される。
Ｍ（ωi ）＝Ｍｉ（ωi ）・Ｍω（ωｉ） … （６）
式（６）に示した各フィルタ関数系の典型例について、以下に示す。
【００４０】
Ｍｉ（ωi ）は、図９（ａ）に示すようなフィルタ特性を有しており、次に示す式（７）〜（９）で表される。

ただし、高周波成分が図９（ａ）に示すように滑らかに減衰する正弦波状関数型にした。ＣＦＲはカットオフ周波数であり、ＷＦＲはフィルタ強度の遷移全周波数幅である（図９（ａ）参照）。このＭｉ（ωi ）は、１次元フーリエ空間での高周波成分を削除するものである。
【００４１】
Ｍω（ωｉ）は、図９（ｂ）に示すようなフィルタ特性を有しており、次に示す式（10）で表される。
Ｍω（ωｉ）＝｜ωｉ｜ … （10）
【００４２】
なお、図９（ａ），（ｂ）には、横軸のプラス方向の特性のみを図示しているが、横軸のマイナス方向の特性は、縦軸を中心に横軸のプラス方向の特性を線対称させたものと同じであるので、図示省略している。
【００４３】
図８に戻って、１次元逆フーリエ変換部は、｜ω｜フィルタリング部で｜ω｜フィルタリングした後のフーリエ面像ＳＣＦ´（ｉ，ω）を１次元逆フーリエ変換して実空間データに戻して、コンボリューション後の投影像ＳＣ´（ｉ，ｊ）を生成する。
【００４４】
第２の重み付け処理部５５は、走査各位置におけるコンボリューション処理後の投影データＳＣ´（ｉ，ｊ）に対して所定の第２の重み付け処理を行なう。具体的には、次に示す式（11）に従って、被検体固定座標系での３次元画素ポイント：Ｐ（ｌ，ｍ，ｎ）の重み関数Ｗ（ｌ，ｍ，ｎ）を計算する（図１０参照）。ただし、Ｈは、画素ポイントＰ（ｌ，ｍ，ｎ）からＸ軸に下ろした垂線の位置である。
Ｗ（ｌ，ｍ，ｎ）＝ＲＯ² ／（ＲＯ＋ＯＨ）² … （11）
【００４５】
続いて、第２の重み付け処理部５５は、図１１に示すように、３次元画素ポイント：Ｐ（ｌ，ｍ，ｎ）の投影像ＳＣ´（ｉ，ｊ）上での座標（Ｉ，Ｊ）と重み用の仮数（ａ_z ，ａ_y ）とを求める。このようにして、第２の重み付け処理を行なう。
【００４６】
次に、逆投影処理部５６は、第２の重み付け処理後の投影データを個別に所定の逆投影（バックプロジェクション：ＢＰ）処理してＢＰ像（３次元ボリュームデータ）を生成する。具体的には、図１１に示すように、走査各位置で検出された、被検体Ｍの関心領域についての一群の第２の重み付け処理後の投影データを、撮影された被検体Ｍの関心領域に仮想的に設定される３次元格子群Ｋの所定の格子点に逆投影して、関心領域の３次元ボリュームデータを生成する画像再構成を行う。
【００４７】
具体的には、次に示す式（12）に従って、線型補間演算とバックプロジェクションとを行なう。なお、バックプロジェクション蓄積量をＩ_n （ｌ，ｍ，ｎ）とし、前回までのバックプロジェクション蓄積量をＩ_n-1 （ｌ，ｍ，ｎ）とする。

【００４８】
なお、投影像の画素間隔を１に規格化して、次に示す式（13）〜（16）のような乗算重み付け方式の場合の重み関数を示す。
Ｗ₁₁＝（１−ａ_z ）・（１−ａ_y ） …（13）
Ｗ₁₂＝（１−ａ_z ）・ａ_y …（14）
Ｗ₂₁＝ａ_z ・（１−ａ_y ） …（15）
Ｗ₂₂＝ａ_z ・ａ_y …（16）
【００４９】
３次元格子群Ｋの残りの所定の格子点についても、前記と同様にして逆投影を行ない、さらに、走査各位置ごと、すなわち、＋θmax （＋３０°）〜−θmax （−３０°）の範囲にわたって、これと同様の逆投影を行なうことで、ＢＰ像（３次元ボリュームデータ）が生成される。
【００５０】
画像情報記憶部５２は、逆投影処理部５６で生成された３次元ボリュームデータを記憶しており、操作部１０から任意の断層面の画像情報が選択されると、その断層面の画像情報をモニタ６０に出力する。
【００５１】
モニタ６０は、画像情報記憶部５２に蓄積された所定の画像情報を出力表示するものである。
【００５２】
以上のように、駆動部３０は、被検体Ｍの周りの円周軌道上に被検体Ｍを挟んで２つの円弧軌道を対向して設定し、両円弧軌道のいずれか一方の円弧軌道上にＸ線管Ｒを移動させるのと同期して、他方の円弧軌道上にフラットパネル型Ｘ線検出器ＤをＸ線管Ｒとの間隔が一定になるように移動させる円弧走査を行う。コンボリューション処理部５４は、走査各位置で検出された投影データに対してコンボリューション処理を施す。逆投影処理部５６は、コンボリューション処理部５４でコンボリューション処理された投影データを、撮影された被検体の関心領域に仮想的に設定される３次元格子群Ｋの所定の格子点に逆投影して、関心領域の３次元ボリュームデータを生成するというコーンビームＣＴ再構成手法であるフェルドカンプ（Feldkamp）法により画像再構成を行えば、各角度からの撮影により得られた複数枚の投影像（走査各位置の投影像）を単一面上に重ね合わせるような検出信号の加算演算処理を行って２次元断層画像データを生成するという従来手法を用いないので、複数回にわたって断層撮影することなく関心領域の３次元ボリュームデータを生成でき、被検体の関心領域についての３次元断層画像を迅速に生成できる。
【００５３】
次に、上記フェルドカンプ（Feldkamp）法による画像再構成を、Ｘ線管Ｒと検出器Ｄとを対向して平行直線走査する場合に適用する場合の本発明実施例を示す。
【００５４】
（１）上述の説明では、駆動部３０は、Ｘ線管Ｒとフラットパネル型Ｘ線検出器Ｄとを円弧移動させて走査（円弧走査）しているが、本願発明では、この円弧走査に替えて、図１２（ａ）に示すように、被検体Ｍを挟んでＸ線管Ｒとフラットパネル型Ｘ線検出器Ｄとが対向配置され、Ｘ線管Ｒを第１方向に直線移動させるのと同期して、フラットパネル型Ｘ線検出器Ｄを前記第１方向とは反対方向である第２方向に平行直線移動させるように連動移動させる、所謂、平行直線走査を行う。以下に、この平行直線走査を行う場合について説明する。
【００５５】
コーンビームＣＴ再構成手法であるフェルドカンプ（Feldkamp）法では、Ｘ線管Ｒから回転中心軸Ｃに下ろした垂線がフラットパネル型Ｘ線検出器Ｄの検出平面と直交しており、ジオメトリの各種情報（Ｘ線管Ｒからフラットパネル型Ｘ線検出器Ｄ間の距離や、回転中心軸Ｃが検出平面上ではどこに現れるか、等の条件）をパラメータとして与え再構成処理を行っている。上述の円弧走査では、Ｘ線管Ｒとフラットパネル型Ｘ線検出器Ｄとの関係が、常に、前述の条件を満たしているので、固定したパラメータで再構成を行っている。
【００５６】
しかし、平行直線走査では、図１２（ａ），（ｂ）に示すように、被検体Ｍに向けてＸ線管Ｒから照射されるコーンビーム状Ｘ線のビーム中心は、被検体Ｍの特定断層面の中心点Ｏを透過し、常に、フラットパネル型Ｘ線検出器Ｄの検出平面の中心点に垂直に入射されるようになっていない、つまり、フラットパネル型Ｘ線検出器Ｄの検出平面の中心点への入射角度が各ビュー毎に異なる角度となっているので、このままでは直接フェルドカンプ（Feldkamp）法を適用できない。
【００５７】
そこで、図１２（ｂ）に示すように、被検体Ｍに向けてＸ線管Ｒから照射されるコーンビーム状Ｘ線のビーム中心がフラットパネル型Ｘ線検出器Ｄの検出平面の中心点に垂直に入射される直線に直交するとともに、Ｘ線管Ｒおよびフラットパネル型Ｘ線検出器Ｄの移動方向に直交するものであって、被検体の関心領域のほぼ中心の断層面内を通過する軸を、仮想回転中心軸ＶＣとし、Ｘ線管Ｒからフラットパネル型Ｘ線検出器Ｄの検出平面の中心点に垂直に下ろされる第１垂線長さＳＩＤとし、Ｘ線管Ｒとフラットパネル型Ｘ線検出器Ｄの検出平面中心とを結ぶ直線と、被検体の関心領域の断層面に直交する断層軸とのなす角度をψとすると、各ビューにおける、フラットパネル型Ｘ線検出器Ｄの検出平面の中心点から第１垂線に下ろした第２垂線の長さＹＡは、ＹＡ＝ＳＩＤtan ψで表される。このように、仮想回転中心軸ＶＣは、回転中心軸Ｃと平行で、回転中心軸ＣとＸ軸座標は等しくなっているが、ビュー毎に異なる位置となっている。つまり、図１２（ｂ）に示すように、Ｂ´では、仮想回転中心軸ＶＣは回転中心軸Ｃと同一点上にあり一致しているが、Ａ´、Ｃ´では、仮想回転中心軸ＶＣは回転中心軸Ｃと同一点上になく、ずれた位置にある。
【００５８】
したがって、各ビューの逆投影を行なう際に、その都度この仮想回転中心軸ＶＣに合わせてパラメータ（前述の式（１）におけるＹ（ｖｒ）をＹＡ＝ＳＩＤ・tan ψに応じて変えて再構成処理を行なうことで、元画像データに手を加えずに（例えば、元画像データを面変換するなどの手を加えないで）フェルドカンプ（Feldkamp）法をそのまま適用できる。なお、前述の式（１）の説明においても定義しておいたように、小文字の「ｖ」はベクトルを表しており、以下においても同様とする。
【００５９】
上述のパラメータＹ（ｖｒ）の変更について具体的に説明する。なお、図１２（ｂ）に示すように、Ｂ´にてｖｒ（ｒベクトルと呼ぶ）で示される点と、Ｃ´にてｖｒｃ（ｒｃベクトルと呼ぶ）で示される点とは、３次元格子群Ｋの同一の点であるとする。仮想回転中心軸ＶＣを基準に考えると、上述の２点のそれぞれのＹ座標の関係は、次に示す式（17）で表される。ただし、図１２（ｂ）のＣ´に示すように、ＳＯＤは、Ｘ線管Ｒから仮想回転中心軸ＶＣまでの距離であり、ｒｃ_y は、仮想回転中心軸ＶＣからｖｒｃで示される点までのＹ軸上の距離であり、図１２（ｂ）のＢ´に示すように、ｒ_y は、仮想回転中心軸ＶＣからｖｒで示される点までのＹ軸上の距離である。なお、被検体Ｍの特定断層面の中心点Ｏは、仮想回転中心軸ＶＣ上にある。
ｒｃ_y ＝ｒ_y ＋ＳＯＤ・tan ψ …（17）
【００６０】
そして、Ｘ線管Ｒからｖｒｃを通ったＸ線がフラットパネル型Ｘ線検出器Ｄの検出平面上に投影される点のＹ座標をＹｃ（ｖｒ）とする。実際のフラットパネル型Ｘ線検出器ＤはＹＡ＝ＳＩＤ・tan ψだけ、シフトして考えているので、実際のフラットパネル型Ｘ線検出器Ｄの検出平面上での座標は、次に示す式（18）で表される。
Ｙ（ｖｒ）＝Ｙｃ（ｖｒ）−ＳＩＤ・tan ψ …（18）
ただし、Ｙｃ（ｖｒ）＝ＳＩＤ・ｒｃ_y ／ＳＯＤである。
【００６１】
このように、各ビューの逆投影を行なう際に、前述の式（１）のパラメータＹ（ｖｒ）を前述の式（18）に基づいて変えて（座標補正を行なって）各画素についての再構成処理を行なうだけでよく、元画像データに手を加えずに（例えば、元画像データを面変換するなどの手を加えないで）フェルドカンプ（Feldkamp）法をそのまま適用できる。
【００６２】
このように、Ｘ線管Ｒとフラットパネル型Ｘ線検出器Ｄを平行直線移動させた場合には、Ｘ線管Ｒとフラットパネル型Ｘ線検出器Ｄとを被検体Ｍを挟んで平行直線走査して、被検体Ｍの関心領域の３次元ボリュームデータを生成する画像再構成を行うための非ＣＴタイプの断層撮影を行うことができる。
【００６５】
また、Ｘ線管Ｒおよびフラットパネル型Ｘ線検出器Ｄを移動させるようにして走査しているが、例えば、Ｘ線管Ｒを固定としフラットパネル型Ｘ線検出器Ｄと被検体Ｍとを移動させて走査したり、フラットパネル型Ｘ線検出器Ｄを固定としＸ線管Ｒと被検体Ｍとを移動させて走査したりするなど、Ｘ線管Ｒとフラットパネル型Ｘ線検出器Ｄと被検体Ｍとのうちのいずれか２つを移動させるようにして走査してもよい。
【００６６】
（２）上述の各実施例では、面検出器としてフラットパネル型Ｘ線検出器Ｄを採用しているが、Ｘ線ＣＣＤカメラやＩ．Ｉ管やイメージングプレートなど各種の２次元面検出器を採用することもできる。
【００６７】
（３）上述の各実施例の断層撮影装置は、被検体Ｍを人体などとして医療用に用いることもできるし、被検体ＭをＢＧＡ（Ball Grid Array）基板やプリント配線基板など各種の電子部品などとして非破壊検査用に用いることもできる。
【００６８】
（４）上述の各実施例では、Ｘ線管ＲによってＸ線を被検体Ｍに照射しているが、Ｘ線に限らず、被検体Ｍに対して透過性を有する例えば、ガンマ線、光などの電磁波を用いた場合であっても、同様の効果を有する。したがって、この発明の断層撮影装置は、Ｘ線断層撮影装置に限定されるものではなく、Ｘ線以外で被検体Ｍに対して透過性を有する電磁波を用いて断層撮影を行う断層撮影装置にも適用可能である。
【００６９】
【発明の効果】
以上の説明からも明らかなように、請求項１に記載の平行直線走査型の断層撮影装置によれば、検出器を被検体の周囲に周回させない、いわゆる非ＣＴタイプの断層撮影装置であっても、ＣＴタイプの再構成手法であるフェルドカンプ（Feldkamp）法による画像再構成手法を適用することができ、従来のように複数回にわたって断層撮影することもなく被検体の関心領域についての３次元断層画像を迅速に生成することができる。
【００７０】
そして、照射源と面検出器とを平行直線走査した場合であっても、投影データ（元画像データ）を面変換するなどの手を加える必要がなく、パラメータを変化させるだけでこの投影データ（元画像データ）を直接に再構成処理を行うことができる。
【００７１】
また、請求項２に記載の発明によれば、面検出器としてフラットパネル型検出器またはイメージインテンシファイアを用いた場合であっても、被検体の関心領域についての３次元断層画像を迅速に生成できる。
【図面の簡単な説明】
【図１】この発明の実施例に係るＸ線ＣＴ装置のブロック図である。
【図２】（ａ）は非ＣＴタイプのＸ線断層撮影装置におけるＸ線管とフラットパネル型Ｘ線検出器との一走査形態を示す概略側面図であり、（ｂ）は（ａ）の概略斜視図である。
【図３】フラットパネル型Ｘ線検出器の構成図である。
【図４】フラットパネル型Ｘ線検出器の概略構成を示す斜視図である。
【図５】（ａ）、（ｂ）はフラットパネル型Ｘ線検出器の層構造を示す断面図である。
【図６】実施例に係るフェルドカンプ（Feldkamp）アルゴリズムを説明するための模式図である。
【図７】実施例の第１の重み付け処理部による余弦補正を説明するための模式図である。
【図８】実施例のコンボリューション部での一連の処理を説明するための模式図である。
【図９】（ａ）、（ｂ）は実施例のフィルタリング部の各フィルタ関数を示す特性図である。
【図１０】コンボリューション処理後の投影データを仮想の３次元格子群に逆投影処理することを説明するための模式図である。
【図１１】コンボリューション処理後の投影データを仮想の３次元格子群に逆投影処理することを説明するための模式図である。
【図１２】（ａ）、（ｂ）は非ＣＴタイプのＸ線断層撮影装置で平行直線走査する場合におけるフェルドカンプ（Feldkamp）適用方法を説明するための様式図である。
【符号の説明】
３０ … 駆動部
５１ … 画像処理部
５４ … コンボリューション処理部
５６ … 逆投影処理部
Ｄ … フラットパネル型Ｘ線検出器
Ｍ … 被検体
Ｒ … Ｘ線管[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a non-CT type tomographic apparatus for performing tomographic imaging of a subject used in the medical field, industrial field, and the like, and in particular, a technique for quickly generating a three-dimensional tomographic image of a region of interest of a subject. About.
[0002]
[Prior art]
As a conventional tomography apparatus, for example, there is an X-ray tomography apparatus. In this X-ray tomography apparatus, an X-ray tube and an image intensifier (hereinafter referred to as “I.I tube”) are opposed to each other with a subject interposed therebetween, and the X-ray tube is linearly moved in a first direction. In synchronization with The I tube is moved in an interlocking manner so as to move in a straight line parallel to a second direction opposite to the first direction, and an arbitrary point on the specific tomographic plane of the subject is I.I. Intermittent imaging is performed while changing the X-ray irradiation angle to the subject of the X-ray tube so as to always be at the same position on the detection surface of the I-tube. Then, the image information of the specific tomographic plane of the subject and the front and back of the specific tomographic plane can be obtained only by performing the addition calculation processing of the detection signal so as to superimpose a plurality of projection images obtained by photographing from each angle. The image information of the nearby tomographic plane is calculated.
[0003]
As described above, the X-ray tomography apparatus described above is of a non-CT type tomography system that is not an X-ray CT (X-ray computer tomography) type tomography system of remarkable progress in recent years. In other words, the X-ray CT type tomography method uses an X-ray tube and an I.D. By capturing the I tube with one rotation (at least half rotation) around the body axis of the subject, a transmission image of one rotation (at least half rotation) around the body axis of the subject is acquired and transmitted through these images. An image reconstruction is performed based on the image, and a tomographic image viewed from the body axis direction of the subject can be obtained. However, the non-CT type tomography method is an X-ray tomography apparatus as described above. Tube and I. A tomographic image in the body axis direction of the subject can be obtained without causing the I tube to rotate more than half a rotation around the body axis of the subject.
[0004]
[Problems to be solved by the invention]
However, the conventional example having such a configuration has the following problems. That is, the non-CT type tomography apparatus has a problem in that it cannot quickly generate a three-dimensional tomographic image of the region of interest of the subject.
[0005]
Specifically, the non-CT type tomography apparatus of the conventional example can perform a specific tomographic image of a region of interest of a subject and neighboring tomographic planes before and after this specific tomographic plane in one tomographic imaging. Only an image can be obtained, and only a two-dimensional tomographic image of the region of interest of the subject can be obtained. Therefore, when it is desired to generate three-dimensional volume data (three-dimensional tomographic image) for the region of interest of the subject, the tomographic imaging is repeated a plurality of times, and the two-dimensional tomographic images obtained for each imaging are sequentially stacked. Since it is necessary to generate a three-dimensional tomographic image, the number of exposures to the subject increases, which not only burdens the subject but also takes a very long time to generate the three-dimensional volume data. It will take.
[0006]
For example, I.I. Even when it is desired to obtain a tomographic image that is not parallel to the detection surface of the I tube (tilted tomographic image), two-dimensional tomographic images obtained for each imaging are sequentially stacked by performing tomographic imaging multiple times. As described above, the three-dimensional tomographic image is generated, and a desired inclined tomographic image is calculated from the three-dimensional tomographic image by calculation.
[0007]
The present invention has been made in view of such circumstances, and an object thereof is to provide a tomographic apparatus capable of quickly generating a three-dimensional tomographic image of a region of interest of a subject.
[0008]
[Means for Solving the Problems]
  In order to achieve such an object, the present invention has the following configuration. That is, the tomography apparatus according to claim 1 includes: (a) an irradiation source that irradiates a subject with an electromagnetic wave having permeability to the subject in a divergent beam shape; and (b) the subject across the subject. A surface detector that is disposed opposite to the irradiation source and detects the electromagnetic wave transmitted through the subject, and (c) in synchronization with linear movement of one of the irradiation source and the surface detector in the first direction, (D) a projection detected at each position of the scanning; and (d) scanning means for performing linear scanning that moves the other side in parallel in a second direction opposite to the first direction across a tomographic plane to be imaged of the subject. An image processing unit that performs image reconstruction for generating a three-dimensional tomographic image of the region of interest of the subject from the data by calculation.
[0009]
  And in particular, (e) the image processing unit corrects the influence of the first weighting process for correcting the beam spread from the irradiation source, the convolution process, and the beam spread from the irradiation source. The process is performed by the Feldkamp method including the second weighting process and the backprojection process, and (f) the backprojection process in the Feldkamp method passes through the substantially center of the region of interest of the subject and takes an image. An axis orthogonal to the tomographic plane to be used is a virtual rotation center axis, and each position is determined according to an angle formed by a straight line connecting the irradiation source and the center of the surface detector at each scanning position and the virtual rotation center axis. This projection data is subjected to coordinate correction and then back-projected onto a predetermined lattice point of the three-dimensional lattice group.
[0010]
  A tomographic apparatus according to claim 2 is the tomographic apparatus according to claim 1, wherein the surface detector is a flat panel detector or an image intensifier.
[0011]
[Action]
  The operation of the present invention is as follows. That is, according to the first aspect of the invention, the irradiation source R irradiates the subject M with an electromagnetic wave having permeability to the subject M in a divergent beam shape. The surface detector D is disposed opposite to the irradiation source R with the subject M interposed therebetween, and detects an electromagnetic wave transmitted through the subject M. The scanning unit 30 synchronizes with the linear movement of one of the irradiation source R and the surface detector D in the first direction, and the other in the second direction that is opposite to the first direction. Straight line scanning is performed by moving parallel straight lines across the tomographic plane to be imaged. The projection data detected at each scanning position is subjected to arithmetic processing in the image processing unit 51 to obtain a three-dimensional tomographic image of the region of interest of the subject M, and the image is reconstructed. Since the Feldkamp method, which is a reconstruction method for cone beam CT, is used for the calculation processing of the above, a plurality of projected images (projected images at each scanning position) obtained by photographing from various angles as in the conventional method 3D tomographic images of the region of interest of the subject M can be quickly generated without using detection signal addition calculation processing that superimposes the images on a single surface.
[0012]
  In general, when the irradiation source R and the surface detector D are scanned in parallel straight lines, the Feldkamp method cannot be used because the angle of incidence on the detection plane differs for each scanning position. In the present invention, during back projection processing in the Feldkamp method, an axis that passes through substantially the center of the region of interest of the subject and is orthogonal to the tomographic plane to be imaged is a virtual rotation center axis, and the irradiation source and the surface at each scanning position The Feldkamp method can be used in parallel straight line scanning by correcting the coordinates of the projection data at each position according to the angle between the straight line connecting the center of the detector and the virtual rotation center axis. It is a feature. With this coordinate correction, even when the irradiation source R and the surface detector D are scanned in parallel straight lines, it is not necessary to perform a process such as plane conversion of the projection data (original image data). Image data) can be directly reconstructed.
[0013]
  According to the second aspect of the present invention, even when a flat panel detector or an image intensifier is used as the surface detector, a three-dimensional tomographic image of the region of interest of the subject can be quickly obtained. Generated.
[0014]
DETAILED DESCRIPTION OF THE INVENTION
A non-CT type X-ray tomography apparatus as an embodiment according to the tomography apparatus of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram according to an embodiment of the X-ray tomography apparatus of the present invention.
[0015]
The X-ray CT apparatus of this embodiment includes an operation unit 10 for inputting various information and commands, an imaging control unit 20 for controlling X-ray imaging based on the input information and commands, and the imaging control unit 20. The driving unit 30 that operates the imaging unit 40 while being controlled by the imaging unit, the imaging unit 40 that captures the region of interest of the subject M, and the region of interest 3 of the subject M based on the image information detected from the imaging unit 40. A data processing unit 50 that performs image reconstruction for generating three-dimensional volume data, stores the generated three-dimensional volume data, and a monitor 60 that outputs and displays image information stored in the data processing unit 50 is provided. .
[0016]
  Hereinafter, the configuration and function of each unit will be described in detail. The X-ray tomography apparatus according to the present invention performs linear parallel scanning. First, the case of taking an arc orbit will be described. As shown in FIG. 2, two circular arc trajectories are set opposite to each other with the subject M sandwiched on a circumferential trajectory around the subject M, and an X-ray tube is placed on one of the circular arc trajectories. Synchronously with the movement of R, arc scanning is performed to move the flat panel X-ray detector D on the other arc orbit so that the distance from the X-ray tube R is constant, and the X-ray tube R Intermittent imaging is performed while changing the X-ray irradiation angle θ of the subject M. The X-ray irradiation angle θ can be set within an arbitrary angle range, but here, for example, it is set within the range of + θmax (+ 30 °) to −θmax (−30 °). Then, from the operation unit 10, before imaging the region of interest of the subject M, the distance from the X-ray tube R to the flat panel X-ray detector D shown in FIG. The number of views (however, it can be set to a desired number, for example, 50) is set and input in advance, at what pitch the X-ray detector D is moved while moving the arc. As the operation unit 10, an input device such as a keyboard, a mouse, or a touch panel is used. The X-ray tube R described above corresponds to an irradiation source.
[0017]
An operation unit 10, a drive unit 30, and a data processing unit 50 are connected to the imaging control unit 20. The imaging control unit 20 controls the drive unit 30 and the data processing unit 50 based on each information set and input from the operation unit 10. Details of the control will be described later in each section.
[0018]
  As shown in FIG. 2, the drive unit 30 sets two circular arc tracks facing each other across the subject M on a circumferential trajectory around the subject M, and either one of the circular arc trajectories. Synchronous with the movement of the X-ray tube R upward, the arc scanning is performed to move the flat panel X-ray detector D on the other arc trajectory so that the distance from the X-ray tube R is constant. It is. At this time, the center of the cone-beam X-ray irradiated from the X-ray tube R toward the subject M is always a specific tomographic plane of the subject M, which is also a specific point on the rotation center axis C. The X-ray tube R and the flat panel X-ray detection are transmitted through the central point O (see FIG. 2B) and perpendicularly incident on the central point of the detection surface of the flat panel X-ray detector D. The container D is opposed. The driving unit 30 described above corresponds to a scanning unit.
[0019]
As shown in FIG. 1, the imaging unit 40 transmits the subject M, a top plate T on which the subject M is placed, an X-ray tube R that irradiates the subject M with cone-beam-shaped X-rays, and the subject M. And a flat panel X-ray detector D for detecting the detected X-rays.
[0020]
The flat panel X-ray detector D detects an X-ray fluoroscopic image of the subject M generated by X-ray irradiation by the X-ray tube R, converts it into an electric signal as an X-ray detection signal, and outputs it. As shown in FIG. 3, the line detector is a so-called two-dimensional matrix X-ray detector in which a large number of detection elements Du are arranged vertically and horizontally. The arrangement of the detection elements Du in the flat panel X-ray detector D of the embodiment is, for example, a square matrix in the horizontal (i row) direction 1024 and the vertical (j column) direction 1024. For convenience of explanation, the horizontal (i) It is assumed that the matrix is a square matrix of 1000 in the direction 1000 and 1000 in the vertical (j) direction, and FIG. 3 shows only 9 matrix configurations in a 3 × 3 matrix configuration. Unlike the image intensifier, which has a rectangular planar shape, the flat panel X-ray detector D enables a rectangular detection surface suitable for imaging large areas such as the chest and abdomen. In particular, it is a useful X-ray detector.
[0021]
As shown in FIG. 4, the flat panel X-ray detector D includes an X-ray conversion layer 12 that converts incident X-rays into charges or light, and an element that detects charges or light generated in the X-ray conversion layer 12. Has a laminated structure with the detection array layer 13 arranged in a matrix form vertically and horizontally. As a planar dimension of the X-ray conversion layer 12 of the flat panel X-ray detector D, for example, about 30 cm in length and width can be mentioned.
[0022]
The flat panel X-ray detector D includes a direct conversion type shown in FIG. 5A and an indirect conversion type shown in FIG. 5B. In the case of the former direct conversion type, the X-ray conversion layer 12 is composed of a selenium layer, a CdZnTe layer, or the like that converts incident X-rays directly into charges, and is opposed to the surface electrode 15 as the charge detection element 14 on the surface of the detection array layer 13 The formed charge collecting electrode group detects charges and stores the charges in a capacitor Cs. One charge detecting element Du is formed by each charge detecting element 14 and a part of the X-ray conversion layer 12 thereon. Will be formed. In the case of the latter indirect conversion type, the X-ray conversion layer 12 is composed of a scintillator layer that converts incident X-rays into light, and light is detected by a photodiode group formed as a light detection element 16 on the surface of the detection array layer 13. Thus, a single detection element Du is formed by each photodetection element 16 and a part of the X-ray conversion layer 12 thereon.
[0023]
As shown in FIG. 3, the flat panel X-ray detector D includes an X-ray detection substrate 41 on which an X-ray conversion layer 12 and a detection array layer 13 are formed, and carrier collection electrodes (charges) of the X-ray detection substrate 41. Capacitor Cs that collects collected carriers (collected charges) via a collecting electrode), and a thin film transistor (TFT) that is a charge extraction switch element 42 for normally off (blocked) charges for extracting the charges accumulated in the capacitor Cs A multiplexer 45 of a readout circuit in the X and Y directions, and a gate driver 47.
[0024]
Further, as shown in FIG. 3, the flat panel X-ray detector D is connected to a vertical readout wiring 43 in which the source of the thin film transistor for the switch element 42 of the detection element Du is arranged in the i direction, and the gate is in the j direction. Are connected to horizontal readout wirings 46 arranged in the horizontal direction. The read wiring 43 is connected to the multiplexer 45 via the charge-voltage converter group (preamplifier group) 44, and the read wiring 46 is connected to the gate driver 47. In the charge-voltage converter group 44, one charge-voltage converter group 44 is connected to one readout wiring 43, although not shown.
[0025]
In the case of the flat panel X-ray detector D, a scanning signal for signal extraction is sent to the multiplexer 45 and the gate driver 47. Identification of each detection element Du of the detection unit 10 is performed by sequentially assigning addresses to each detection element Du along the arrangement in the i direction and the j direction (the number of detection elements Du is 1000, so 0 to 999, detection elements When Du is 1024, the scanning is performed on the basis of 0 to 1023), so that the scanning signal for extraction is a signal designating the i-direction address or the j-direction address, respectively.
[0026]
As the extraction voltage is applied from the gate driver 47 to the read wiring 46 in the j direction according to the scan signal in the j direction, each detection element Du is selected in units of columns. Then, the multiplexer 45 is switched according to the scanning signal in the i direction, so that the charge accumulated in the capacitor Cs of the detection element Du in the selected column is sent to the outside through the charge-voltage changer group 44 and the multiplexer 45 in this order. Will be sent. As described above, the detection signal detected by the flat panel X-ray detector D is sequentially output to the data processing unit 50 in real time. The flat panel X-ray detector D described above corresponds to the surface detector in the present invention.
[0027]
Next, the configuration and function of the data processing unit 50 will be described. As shown in FIG. 1, the data processing unit 50 performs image reconstruction that generates three-dimensional volume data of a region of interest based on projection data (detection signal) detected at each scanning position in the imaging unit 40. A processing unit 51 and an image information storage unit 52 that stores the three-dimensional volume data of the region of interest reconstructed by the image processing unit 51 are provided. Hereinafter, specific functions of the image processing unit 51 and the image information storage unit 52 will be described. The image processing unit 51 performs surface detection of the Feldkamp method, which is a cone beam CT reconstruction method, that is, an X-ray tube R that irradiates cone beam X-rays and X-rays transmitted through the subject. The flat panel X-ray detector D is rotated once (at least half rotation) around the body axis of the subject, and a transmission image of the subject is photographed to obtain a transmission image for one rotation around the body axis of the subject. Based on this, a CT-type image reconstruction method of performing image reconstruction for generating three-dimensional volume data of a region of interest is performed by using an X-ray tube R and a flat panel X-ray detector D around the body axis of the subject. This method can be applied to a so-called non-CT type image reconstruction method that does not rotate half a turn or more.
[0028]
Here, a series of processing steps of image reconstruction for generating three-dimensional volume data of a region of interest by the Feldkamp method that can be applied to a non-CT type image reconstruction method is shown in FIGS. Outline with reference. First, as shown in FIG. 2, two circular arc trajectories are set opposite to each other with the subject M sandwiched on a circumferential trajectory around the rotation center axis C set almost at the center of the region of interest of the subject M, In synchronization with the movement of the X-ray tube R on one of the two circular arc tracks, the distance between the flat panel X-ray detector D and the X-ray tube R is kept constant on the other circular arc track. By performing an arc scan that moves in such a manner, the region of interest of the subject M is imaged, and a group of projection data for the region of interest of the subject M detected at each position of the scan is acquired. Next, a predetermined first weighting process to be described later is performed on the group of projection data individually. Next, a predetermined convolution process to be described later is performed on each projection data after the first weighting process. Next, a predetermined second weighting process described later is performed on each projection data after the convolution process. Next, the projection data after the second weighting process is individually processed by a predetermined back projection (back projection: BP) described later to generate a BP image (three-dimensional volume data). In this way, image reconstruction for generating three-dimensional volume data of the region of interest is performed. The operator selects an image of an arbitrary tomographic plane from the three-dimensional volume data, so that the selected tomographic image (tomographic image viewed from the X-axis direction) can be seen.
[0029]
Further, the above-described Feldkamp algorithm is expressed as the following equations (1) to (3). Multiple projection data PΦ from different angles The cube f (r) is reconstructed based on (see FIG. 6).
[0030]
[Expression 1]

[0031]
Here, f (r) is pixel data at the position r of the reconstructed cube (three-dimensional volume data). Y (vr) and Z (vr) are the coordinates of the point at which the pixel at the position r is projected on the detection surface of the flat panel X-ray detector D, and the lowercase letter v means “vector”. In the following, it is assumed that a vector is appropriately represented by a lowercase letter “v”. PΦ Is projection data on the detection surface of the flat panel X-ray detector D at the projection angle Φ. g_yIs a filter function of Filtered Back Projection, and is a | ω | (absolute value omega) filter function described later. W₁ , W₂ Is a coefficient for correcting the influence of beam spread, and W₁Is a coefficient relating to a first weighting process described later, and W₂ Is a coefficient relating to a second weighting process to be described later.
[0032]
As shown in FIG. 1, the image processing unit 51 includes a first weighting processing unit 53 that individually performs a predetermined first weighting process on a group of projection data obtained by photographing, and a post-first weighting process. A convolution processing unit 54 that performs a predetermined convolution process on each projection data, and a second weighting processing unit 55 that performs a predetermined second weighting process on each projection data after the convolution process, The projection data after the second weighting process is individually provided with a back projection processing unit 56 that performs a predetermined back projection (back projection: BP) process to generate a BP image (three-dimensional volume data).
[0033]
The first weighting processing unit 53 performs a predetermined first weighting process on each group of projection data obtained by photographing. Specifically, as shown in FIG. 7, for each projection data detected by the flat panel X-ray detector D at each scanning position, for each pixel row i of the flat panel X-ray detector D. The pixel detection level fluctuation in the view direction is corrected. As shown in FIG. 7, the center point of the cone-beam X-ray irradiated from the X-ray tube R toward the subject M is always the center point O (rotation center) of the specific tomographic plane of the subject M. (Is also a point on the axis C) and is incident perpendicularly to the center point of the detection surface of the flat panel X-ray detector D.
[0034]
Subsequently, as shown in FIG. 7, the first weighting processing unit 53 performs weighting processing based on the following expression (4) on the projection data.
cos θ = SID / (SID² + Yj² + Zj² )^1/2  (4)
That is, the weighting process is performed by applying cos θ of Expression (4) to each pixel. For example, the pixel Dij is Yj · cos θ, and the pixel value after the weighting process is obtained. Since this is constant regardless of the view, it is created in advance as a weighting process. In this way, the projection data after the first weighting process is calculated (shown as “first weighted projection image: SC (i, j)” in FIG. 8).
[0035]
The convolution processing unit 54 performs a predetermined convolution process on each projection data after the first weighting process, that is, the first weighted projection image: SC (i, j). Since the convolution processing performed in the real space is equivalent to the filtering processing performed in the Fourier space, here, for convenience of explanation, the above-described predetermined convolution processing is performed in the Fourier space (| ω | ( It will be described as (Absolute Value Omega) Filtering Process). Hereinafter, the | ω | filtering process in the convolution processing unit 54 will be described.
[0036]
The convolution processing unit 54 performs a one-dimensional Fourier transform in the horizontal direction for each i row of the flat panel X-ray detector D, and generates a Fourier plane image SCF (i, ω). A filtering unit that applies | ω | filtering to a Fourier plane image SCF (i, ω) subjected to two-dimensional Fourier transformation, and a one-dimensional inverse Fourier transformation of the Fourier plane image SCF ′ (i, ω) after filtering by the filtering unit And a one-dimensional inverse Fourier transform unit for returning to real space data.
[0037]
As shown in FIG. 8, the filtering unit is a filter and a data collection that suppresses a high-frequency noise component by isotropically reducing a high-frequency region in the i-row direction of the Fourier plane image SCF (i, ω) subjected to the one-dimensional Fourier transform. | Ω | filtering section configured by a filter depending on the scanning form. Note that the above-described filter that depends on the data acquisition scanning mode suppresses generation of a DC component that is emphasized when the filtered Fourier plane image SCF ′ (i, ω) is subjected to a one-dimensional inverse Fourier transform. Therefore, the false image resulting from the enhancement of the DC component is reduced.
[0038]
Here, the meaning of performing the filtering process in the one-dimensional Fourier space will be described. Performing the filtering process in the one-dimensional Fourier space is mathematically expressed by the following equation (5). Note that SCF ′ (i, ω) is a one-dimensional Fourier plane image after the filtering process, and M (ωi) is a function indicating the filter characteristics of the filtering unit described above.
SCF ′ (i, ω) = SCF (i, ω) × M (ωi) (5)
[0039]
M (ωi) is expressed as the following equation (6) as a product of the functions representing the two filter characteristics described above.
M (ωi) = Mi (ωi) · Mω (ωi) (6)
A typical example of each filter function system shown in Expression (6) is shown below.
[0040]
Mi (ωi) has a filter characteristic as shown in FIG. 9A and is expressed by the following equations (7) to (9).

However, the high-frequency component is a sinusoidal function type that attenuates smoothly as shown in FIG. CFR is a cutoff frequency, and WFR is a transition full frequency width of the filter strength (see FIG. 9A). This Mi (ωi) deletes high frequency components in the one-dimensional Fourier space.
[0041]
Mω (ωi) has a filter characteristic as shown in FIG. 9B and is expressed by the following equation (10).
Mω (ωi) = | ωi | (10)
[0042]
9A and 9B show only the characteristics in the positive direction of the horizontal axis, the characteristics in the negative direction of the horizontal axis are characteristics in the positive direction of the horizontal axis centering on the vertical axis. Are the same as those obtained by making the lines symmetrical with each other, and are not shown.
[0043]
Referring back to FIG. 8, the one-dimensional inverse Fourier transform unit performs one-dimensional inverse Fourier transform on the Fourier plane image SCF ′ (i, ω) after | ω | filtering by the | ω | Thus, the projection image SC ′ (i, j) after the convolution is generated.
[0044]
The second weighting processing unit 55 performs a predetermined second weighting process on the projection data SC ′ (i, j) after the convolution process at each scanning position. Specifically, the weight function W (l, m, n) of the three-dimensional pixel point: P (l, m, n) in the subject fixed coordinate system is calculated according to the following equation (11) (FIG. 10). However, H is the position of a perpendicular line drawn from the pixel point P (l, m, n) to the X axis.
W (l, m, n) = RO² / (RO + OH)²    (11)
[0045]
Subsequently, the second weighting processing unit 55, as shown in FIG. 11, coordinates (I, J) on the projection image SC ′ (i, j) of the three-dimensional pixel point: P (l, m, n). ) And the mantissa for weight (a_z, A_y) And ask. In this way, the second weighting process is performed.
[0046]
  Next, the back projection processing unit 56 individually performs predetermined back projection (back projection: BP) processing on the projection data after the second weighting processing to generate a BP image (three-dimensional volume data). Specifically, as shown in FIG. 11, a group of second weighted projection data for the region of interest of the subject M detected at each scanning position is used as the captured region of interest of the subject M. The image is reconstructed by projecting back to a predetermined lattice point of the three-dimensional lattice group K virtually set to 1 to generate three-dimensional volume data of the region of interest.
[0047]
Specifically, linear interpolation calculation and back projection are performed according to the following equation (12). Note that the back projection accumulation amount is I_n (L, m, n) and the amount of back projection accumulated until the previous time is I_n-1 Let (l, m, n).

[0048]
In addition, the weighting function in the case of the multiplication weighting system like the following formulas (13) to (16) is shown by normalizing the pixel interval of the projected image to 1.
W₁₁= (1-a_z) ・ (1-a_y) …(13)
W₁₂= (1-a_z) ・ A_y          …(14)
W_{twenty one}= A_z・ (1-a_y... (15)
W_{twenty two}= A_z・ A_y                  … (16)
[0049]
Back projection is performed on the remaining predetermined lattice points of the three-dimensional lattice group K in the same manner as described above, and further, for each scanning position, that is, over the range of + θmax (+ 30 °) to −θmax (−30 °). BP image (three-dimensional volume data) is generated by performing back projection similar to this.
[0050]
The image information storage unit 52 stores the three-dimensional volume data generated by the backprojection processing unit 56. When image information on an arbitrary tomographic plane is selected from the operation unit 10, the image information on the tomographic plane is displayed. Output to the monitor 60.
[0051]
The monitor 60 outputs and displays predetermined image information stored in the image information storage unit 52.
[0052]
  As described above, the drive unit 30 sets two circular arc tracks facing each other with the subject M sandwiched on the circumferential trajectory around the subject M, and on either one of the circular arc trajectories. In synchronization with the movement of the X-ray tube R, arc scanning is performed in which the flat panel X-ray detector D is moved on the other arc orbit so that the distance from the X-ray tube R is constant. The convolution processing unit 54 performs a convolution process on the projection data detected at each scanning position. The backprojection processing unit 56 backprojects the projection data subjected to the convolution processing by the convolution processing unit 54 onto predetermined lattice points of the three-dimensional lattice group K virtually set in the region of interest of the imaged subject. If image reconstruction is performed by the Feldkamp method, which is a cone beam CT reconstruction method for generating three-dimensional volume data of a region of interest, a plurality of projection images obtained by photographing from various angles are obtained. The conventional method of generating the two-dimensional tomographic image data by performing the addition calculation processing of the detection signals so as to superimpose (projected images at the respective scanning positions) on a single plane is not used, so that the tomographic imaging is not performed multiple times. Three-dimensional volume data of the region of interest can be generated, and a three-dimensional tomographic image of the region of interest of the subject can be generated quickly.
[0053]
  Next, an embodiment of the present invention in the case where the image reconstruction by the Feldkamp method is applied to parallel X-ray scanning with the X-ray tube R and the detector D facing each other will be described.
[0054]
  (1) In the above description, the drive unit 30 performs scanning (arc scanning) by moving the X-ray tube R and the flat panel X-ray detector D in an arc, but in the present invention, the arc scanning is performed. Instead, as shown in FIG. 12A, the X-ray tube R and the flat panel X-ray detector D are arranged opposite to each other with the subject M interposed therebetween, and the X-ray tube R is linearly moved in the first direction. In synchronism with this, so-called parallel linear scanning is performed in which the flat panel X-ray detector D is interlocked and moved so as to move in parallel to a second direction opposite to the first direction. The case where this parallel linear scanning is performed will be described below.
[0055]
  In the Feldkamp method, which is a cone beam CT reconstruction method, the perpendicular drawn from the X-ray tube R to the rotation center axis C is orthogonal to the detection plane of the flat panel X-ray detector D, and various geometries are obtained. Information (conditions such as the distance between the X-ray tube R and the flat panel X-ray detector D and where the rotation center axis C appears on the detection plane) is given as parameters, and the reconstruction process is performed. In the above-described arc scanning, since the relationship between the X-ray tube R and the flat panel X-ray detector D always satisfies the above-described conditions, reconstruction is performed with fixed parameters.
[0056]
However, in the parallel linear scanning, as shown in FIGS. 12A and 12B, the beam center of cone-beam X-rays emitted from the X-ray tube R toward the subject M is specified by the subject M. It passes through the center point O of the tomographic plane and is not always incident perpendicularly to the center point of the detection plane of the flat panel X-ray detector D, that is, the detection of the flat panel X-ray detector D. Since the incident angle to the center point of the plane is different for each view, the Feldkamp method cannot be applied directly.
[0057]
Therefore, as shown in FIG. 12B, the beam center of the cone-beam X-ray irradiated from the X-ray tube R toward the subject M is the center point of the detection plane of the flat panel X-ray detector D. It is perpendicular to the perpendicularly incident straight line and perpendicular to the moving direction of the X-ray tube R and the flat panel X-ray detector D, and passes through the tomographic plane at the center of the region of interest of the subject. The axis is a virtual rotation center axis VC, the first perpendicular length SID is taken down perpendicularly from the X-ray tube R to the center point of the detection plane of the flat panel X-ray detector D, and the X-ray tube R and the flat panel type Assuming that an angle between a straight line connecting the detection plane center of the X-ray detector D and a tomographic axis perpendicular to the tomographic plane of the region of interest of the subject is ψ, the flat panel X-ray detector D in each view First down from the center point of the detection plane to the first perpendicular The length of the perpendicular line YA is represented by YA = SIDtan ψ. As described above, the virtual rotation center axis VC is parallel to the rotation center axis C, and the rotation center axis C and the X-axis coordinate are the same, but at different positions for each view. That is, as shown in FIG. 12B, in B ′, the virtual rotation center axis VC is on the same point and coincides with the rotation center axis C, but in A ′ and C ′, the virtual rotation center axis VC. Is not on the same point as the rotation center axis C, but at a shifted position.
[0058]
Therefore, each time the back projection of each view is performed, reconstruction is performed by changing the parameter (Y (vr) in the above equation (1) according to YA = SID · tan ψ in accordance with the virtual rotation center axis VC each time. By performing the processing, the Feldkamp method can be applied as it is without changing the original image data (for example, without changing the surface of the original image data). As defined in the description of 1), the lowercase letter “v” represents a vector, and the same applies to the following.
[0059]
The change of the parameter Y (vr) will be specifically described. As shown in FIG. 12B, a point indicated by vr (referred to as r vector) at B ′ and a point indicated by vrc (referred to as rc vector) at C ′ are a three-dimensional lattice. Let it be the same point of group K. Considering the virtual rotation center axis VC as a reference, the relationship between the Y coordinates of the two points described above is expressed by the following equation (17). However, as indicated by C ′ in FIG. 12B, SOD is the distance from the X-ray tube R to the virtual rotation center axis VC, and rc_yIs the distance on the Y-axis from the virtual rotation center axis VC to the point indicated by vrc, and as indicated by B ′ in FIG._yIs the distance on the Y axis from the virtual rotation center axis VC to the point indicated by vr. The center point O of the specific tomographic plane of the subject M is on the virtual rotation center axis VC.
rc_y= R_y+ SOD ・ tan ψ (17)
[0060]
Then, the Y coordinate of the point at which the X-ray passing from the X-ray tube R through vrc is projected onto the detection plane of the flat panel X-ray detector D is defined as Yc (vr). Since the actual flat panel X-ray detector D is considered to be shifted by YA = SID · tan ψ, the coordinates on the detection plane of the actual flat panel X-ray detector D are expressed by the following equations: It is represented by (18).
Y (vr) = Yc (vr) −SID · tan ψ (18)
However, Yc (vr) = SID · rc_y/ SOD.
[0061]
In this way, when back projection of each view is performed, the parameter Y (vr) of the above-described equation (1) is changed based on the above-described equation (18) (coordinate correction is performed), and the re-projection for each pixel is performed. It is only necessary to perform the configuration process, and the Feldkamp method can be applied as it is without modifying the original image data (for example, without modifying the surface of the original image data).
[0062]
In this way, when the X-ray tube R and the flat panel X-ray detector D are moved in parallel straight lines, the X-ray tube R and the flat panel X-ray detector D are parallel straight lines across the subject M. Non-CT type tomography for performing image reconstruction by scanning and generating three-dimensional volume data of the region of interest of the subject M can be performed.
[0065]
The X-ray tube R and the flat panel X-ray detector D are scanned so as to move. For example, the X-ray tube R is fixed and the flat panel X-ray detector D and the subject M are moved. The X-ray tube R and the flat panel X-ray detector D can be scanned by moving the X-ray tube R and the subject M while the flat panel X-ray detector D is fixed. And the subject M may be scanned so as to move any two of them.
[0066]
(2) In each of the above-described embodiments, the flat panel X-ray detector D is employed as the surface detector. Various two-dimensional surface detectors such as an I tube and an imaging plate can also be employed.
[0067]
(3) The tomography apparatus of each of the embodiments described above can be used for medical purposes with the subject M as a human body or the like, and the subject M can be used for various electronic components such as a BGA (Ball Grid Array) substrate and a printed wiring board. For example, it can be used for nondestructive inspection.
[0068]
(4) In each of the above-described embodiments, the subject M is irradiated with the X-rays by the X-ray tube R. However, the present invention is not limited to the X-rays and has transparency to the subject M, for example, gamma rays, light, etc. Even when the electromagnetic wave is used, the same effect is obtained. Therefore, the tomography apparatus of the present invention is not limited to the X-ray tomography apparatus, but also to a tomography apparatus that performs tomography using electromagnetic waves that are transmissive to the subject M other than X-rays. Applicable.
[0069]
【The invention's effect】
  As apparent from the above description, the parallel linear scanning tomography apparatus according to claim 1 is a so-called non-CT type tomography apparatus in which the detector does not circulate around the subject. In addition, an image reconstruction method based on the Feldkamp method, which is a CT-type reconstruction method, can be applied, and a three-dimensional image of a region of interest of a subject can be obtained without performing tomography multiple times as in the past. A tomographic image can be generated quickly.
[0070]
  Even when the irradiation source and the surface detector are scanned in parallel straight lines, it is not necessary to change the plane of the projection data (original image data), and only by changing the parameters, this projection data ( The original image data) can be directly reconstructed.
[0071]
  According to the second aspect of the present invention, even when a flat panel detector or an image intensifier is used as the surface detector, a three-dimensional tomographic image of the region of interest of the subject can be quickly obtained. Can be generated.
[Brief description of the drawings]
FIG. 1 is a block diagram of an X-ray CT apparatus according to an embodiment of the present invention.
2A is a schematic side view showing one scanning mode of an X-ray tube and a flat panel X-ray detector in a non-CT type X-ray tomography apparatus, and FIG. 2B is a side view of FIG. It is a schematic perspective view.
FIG. 3 is a configuration diagram of a flat panel X-ray detector.
FIG. 4 is a perspective view showing a schematic configuration of a flat panel X-ray detector.
5A and 5B are cross-sectional views showing a layer structure of a flat panel X-ray detector.
FIG. 6 is a schematic diagram for explaining a Feldkamp algorithm according to an embodiment.
FIG. 7 is a schematic diagram for explaining cosine correction by the first weighting processing unit of the embodiment.
FIG. 8 is a schematic diagram for explaining a series of processes in the convolution unit of the embodiment.
FIGS. 9A and 9B are characteristic diagrams illustrating filter functions of the filtering unit according to the embodiment. FIGS.
FIG. 10 is a schematic diagram for explaining that the projection data after the convolution process is back-projected into a virtual three-dimensional lattice group.
FIG. 11 is a schematic diagram for explaining that the projection data after the convolution processing is back-projected into a virtual three-dimensional lattice group.
FIGS. 12A and 12B are style diagrams for explaining a method for applying a Feldkamp in the case of performing parallel linear scanning with a non-CT type X-ray tomography apparatus; FIGS.
[Explanation of symbols]
30 ... Drive unit
51. Image processing unit
54 ... Convolution processing section
56 ... Back projection processing unit
D ... Flat panel X-ray detector
M… Subject
R ... X-ray tube

Claims (2)

(A) an irradiation source for irradiating the subject with an electromagnetic wave having permeability to the subject in a divergent beam shape;
(B) a surface detector that is disposed opposite to the irradiation source across the subject and detects electromagnetic waves transmitted through the subject;
(C) The subject is imaged in a second direction opposite to the first direction in synchronism with linear movement of one of the irradiation source and the surface detector in the first direction. Scanning means for performing linear scanning for parallel linear movement across the power slice plane;
(D) a tomography apparatus including an image processing unit that performs image reconstruction for generating a three-dimensional tomographic image of a region of interest of a subject by calculation from projection data detected at each scanning position;
(E) The image processing unit performs a first weighting process for correcting the influence of the beam spread from the irradiation source, a convolution process, and a second for correcting the influence of the beam spread from the irradiation source. , The Feldkamp method consisting of weighting processing and back projection processing,
(F) In the back projection processing in the Feldkamp method, a tomographic axis orthogonal to the tomographic plane of the region of interest of the subject to be imaged and a straight line connecting the irradiation source and the center of the surface detector at each scanning position A tomographic apparatus characterized in that the projection data at each of the positions is corrected according to the angle formed by the projection, and then back-projected onto a predetermined grid point of the three-dimensional grid group.   The tomography apparatus according to claim 1, wherein the surface detector is a flat panel detector or an image intensifier.

Images