JPS61179136A - Image processing of head part radiation image containing nostril - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to an image processing method for a radiographic image of the head including the paranasal sinuses.

(Prior Art) When diagnosing sinuses, radiographic images such as X-ray photographs are one of the means for providing the most important diagnostic information. In conventional otorhinolaryngological diagnosis, microscope X-rays using simple imaging methods such as the Caldwell method and the Waters method have been mainly used. This method is often unsatisfactory when used for the purpose of diagnosing otorhinolaryngological diseases, as the contrast in the paranasal sinuses is not good.

Conventionally, radiographic images such as X-ray images are widely used for medical purposes. Photography method is used.

In recent years, with the advancement of X-ray image diagnostic technology, CR
Methods for reproducing the image on T-screens and photosensitive materials began to be devised. As a result, more diagnostic information can be obtained from a single X-ray photograph, resulting in improved diagnostic performance and reduced X-ray exposure. It is also expected to improve the efficiency of storing and retrieving X-ray image information.

On the other hand, methods for obtaining radiographic images without using radiographic films made of silver salt photosensitive materials have been devised.

Such a method involves causing the radiation transmitted through the object to be absorbed by a certain type of phosphor, and then exciting this phosphor with, for example, light or thermal energy, so that the phosphor accumulates due to the absorption. There is a method in which radiation energy is emitted as fluorescence, and this fluorescence is detected and imaged. Specifically, for example, British patent 1,462,7
No. 69 and JP-A-51-29889 disclose a method of using a heat-stimulable phosphor as the phosphor. This method uses a radiation image conversion panel in which a heat-stimulable phosphor layer is formed on a support, and the radiation transmitted through the subject is absorbed by the heat-stimulable phosphor layer of this radiation image conversion panel. Radiation energy corresponding to the radiation transmittance of each part of the subject is accumulated to form a latent image, and then this thermally stimulable phosphor layer is heated to excite it, and the radiation energy accumulated in each part of the panel is It extracts the light as a light signal and obtains a radiographic image based on the intensity of this light.

Further, for example, US Pat. No. 3,859,527 and Japanese Patent Application Laid-Open No. 12144/1983 disclose a method of using a photoluminescent phosphor as the phosphor. This method uses a radiation image conversion panel in which a photostimulable phosphor layer is formed on a support. After forming a latent image as described above, this photoluminescent phosphor layer is photostimulated. By scanning with light,
The radiation energy accumulated in each part of the panel is extracted as a light signal to obtain a radiation image. This final image may be reproduced as a hard copy or
It may be played back on T.

Radiographic images of the head for the purpose of diagnosing sinuses can also be read from recording materials such as X-ray film and stimulable phosphors using the method described above, converted into electrical signals, and then reimaged. It is.

However, at present, there has been little study on an appropriate radiation image processing method for improving the contrast, sharpness, etc. of the reproduced image to obtain a head image that is easy to diagnose.

(Object of the Invention) An object of the present invention is to provide a radiographic image processing method for obtaining a head radiographic image that improves the image representation of the paranasal sinuses and provides more useful diagnostic information regarding otorhinolaryngological diseases. There is a particular thing.

(Structure of the Invention) The object of the present invention is to read radiographic image information by scanning a recording material in which radiographic images of the head including the paranasal sinuses are recorded, convert this into an electrical signal, and convert the radiographic image information from the electrical signal. When obtaining a reproduced image, the eye socket part in the head is
．． This is achieved by an image processing method of a head radiation image including the paranasal sinuses, which is characterized in that it is reproduced with an optical density in the range of 2 to 2.5.

When reconstructing images on a photosensitive material based on electrical signals obtained by converting radiation image information, K is a reference image for each part of the human body in order to finish all radiation images at an appropriate density that is easy to see. It is sufficient to designate a portion and perform gradation processing so that the image portion has a certain optical density on the reproduced image.

In this case, it is necessary that the image portion serving as the reference is located close to the image portion that is most important for diagnosis, and that the subject contrast in the vicinity of the reference image portion has relatively small individual differences among patients.

In cranial radiology images for the purpose of diagnosing sinuses,
The most important image area for diagnosis is naturally the lateral nasal cavity area, especially the maxillary sinus, but the paranasal sinus area is inappropriate as the reference image area because radiation absorption varies greatly depending on the presence or absence of disease. Therefore, as a result of repeated studies on a large number of head radiation images, the present inventors found that it is most appropriate to designate the orbital part as the reference image part, and that the orbital part should be 1.2 It has been found that by finishing the optical density to an optical density in the range of ~2.5, a reproduced image that is easy to see and has an appropriate density can be obtained. In order to obtain an elephant in a reproduced image that is easier to see, the optical density of the eye socket part in the reproduced image must be set to 1.
It is preferable to finish in the range of 5 to 2.2.

According to the radiation image processing method of the present invention, reproduced images are expressed with easy-to-see density, and variations in density between individual reproduced images are eliminated. Therefore, for example, it becomes easier to compare head images of the same patient taken multiple times at intervals to evaluate the progress of the disease state and the effectiveness of treatment.

In addition, the degree of progression of sinusitis is usually diagnosed from the density of the sinus area in a head X-ray photograph, but according to the radiographic image processing method of the present invention, diagnosis is easier and more reliable. It gets expensive.

To express the gradation characteristics of the reproduced image, the electric signal level S obtained by converting the radiation image information is plotted on one axis of the orthogonal coordinates, and the optical density of the reproduced image is plotted on the other axis. A gradation conversion curve that represents the relationship in a graph is used. Since the axis of the electrical signal S is usually expressed on a scale corresponding to the logarithmic representation of the radiation exposure amount, the gradation conversion curve can be considered to correspond to the characteristic curve of X-ray film in the general screen/film method. .

By the way, in order to obtain a reproduced image that facilitates diagnosis of the sinuses, the gamma, that is, the gradient of the gradation conversion curve, in the sinus region should be in the range of 1.5 to 3.8, more preferably 1.8 to 3.8. It is effective to perform gradation processing to obtain a value in the range of 6.

It is natural that the slope of the tone conversion curve is positive or 0 in all electrical signal ranges in order to obtain a natural reproduced image, but there are various possible shapes of the curve. .

A preferred example of a gradation processing method for a head image including the paranasal sinuses is a gradation processing method that finishes the image so that the contrast is constant in all electrical signal ranges of the subject. This method is simple because it is only necessary to specify the optical density and gamma that correspond to the electrical signal level of the orbital region, and the sinus region can be finished to make it easier to see.

Another preferred example of a gradation processing method for a head image including the paranasal sinuses is to make the contrast of the paranasal sinuses higher than that of the teeth in the head image, and to make the contrast of the paranasal sinuses higher than that of the teeth in the head image. One example is a gradation processing method that makes the contrast higher than or equal to the contrast of the cervical soft tissue portion in the image. These gradation processing methods are extremely useful because they can increase the contrast in the sinus region and at the same time prevent "drops" from appearing in the lowest or highest density portions of the reproduced image.

According to the above-mentioned gradation processing method that increases the contrast of the sinus region, it becomes easier to diagnose the size and shape of benign tumors, malignant lies, cysts, mucosal thickening due to sinusitis, and the like. In particular, the above-mentioned gradation processing can be applied to the positions of bones and bone mass boundaries, the size of mucosal thickening, etc., which are difficult to diagnose from X-ray photographs taken using ordinary screen/film systems. This makes it easier to see and diagnose.

In addition to gradation processing, if appropriate spatial frequency emphasis is applied, a reproduced image that is easier to view can be obtained. Any conceivable method for spatial frequency emphasis may be used, but if the modulation transfer function of the image after spatial frequency emphasis is
1. It is preferable to take the maximum value in the range of 0 c/mm. Further, the maximum value of the modulation transfer function is preferably 5.0 parts or less of the value of the modulation transfer function at a spatial frequency of 0.01 c,/+tx or less. Note that the frequency enhancement may be performed before or after the gradation processing.

By performing the frequency enhancement, the edges of cysts and bone destruction areas due to cancer or severe sinusitis can be clearly seen, providing more useful diagnostic information.

The stimulable phosphor used for forming a radiation image in the present invention is stimulated by optical, thermal, mechanical, chemical, or electrical stimulation (stimulable phosphor) after being irradiated with the first light or high-energy radiation. It is a phosphor that exhibits stimulated luminescence corresponding to the amount of initial light or high-energy radiation irradiated by excitation (excitation), but from a practical standpoint it preferably exhibits stimulated luminescence by stimulated excitation light of 500 μm or more. It is a phosphor.

As the photostimulable phosphor according to the present invention, for example, JP-A No. 48
A phosphor represented by Ba5O,:Ax (where A is at least one of Dlr, Tb, and Tm, and X is 0.001≦x<1 mol%) described in No.-80487. , MgSO described in JP-A-48-80488
4: Phosphor represented by Ax (where A is either Ho or Dy, and 0.0014X≦1 mol%), S rsO4: Ax described in JP-A-48-80489 (However, A is DY, Tb and T
At least one type of m, and x ha0.001≦x
(1 mol%), Na, So, described in JP-A-51-29889;
, Ca5O, and BaSO4, etc. with Mn, Dy, and T.
A phosphor containing a small amount of b (also 1),
-Bed, LIF, described in No. 30487,
Phosphors such as MgSO4 and CaF, JP-A-53-3
Li2B40y described in No. 9277: Cu, A
Li2O.(B202)x:Cu (where X is 2 (
x≦3), and Li20 ・(B201 ) x: cu,
Phosphors such as Ag (where X is 2 (x≦3), U.S. Patent 3
, 859,527: SrS:Ce, S
m, SrS: Eu, Sm.

La2O2S: Eu, Sm and (Zn, Cd)S
:Mn, X (where X is )-logen). In addition, ZnS:Cu, Pb phosphor, Ba0-xA described in JP-A-55-12142
Barium aluminate phosphor represented by l, O, :Eu (however, Q, 34
sjo, :A (However, MII is Mg, Ca, Sr, Zn,
Cd or Ba, A is Ce, Tb, Eu, Tm,
It is at least one of Pb, Tl, Bi, and Mn, and X satisfies 0.54X<2.5. ) Alkaline earth metal silicate-based phosphors represented by: Also, (Ba1-X-yMgxCay) FX
'e Eu2'' (However, X is a small amount of Br and C1 (
are one, and x, y and e are each O(x +
y≦0.6, x y % Q and 10-6≦e l,
Alkaline earth fluoride halide phosphor expressed by 5 x 10-'', JP-A-5
LnOX described in No. 5-12144: xA
(However, Ln is La, Y.

At least one of Gd and Lu, X is CI and/or Br, A is Ce and/or Tb, X is Ox (0
, 1. ), (Ba, -XMx)FX:yA described in JP-A-55-12145 (where MII is at least one of Mg, Ca, Sr, Zn and Cd) , X is one of CI, Br and l, A is Eu, Tb, Cs, Tm, Dy
, Pr, Ho, Nd+Yb and Er, X and y are 0≦X40.6 and O≦y≦0,2
represents the number that satisfies the condition. ), a phosphor represented by
BeFX described in JP-A No. 55-84389:
xCe, yA (where J is at least one of C1, Br and 1, and A is In.

is at least one of T1+Gct, Sm and Zr, and X and y are respectively O(x≦2XIO-' and Q
(y45X10-2), MriFX-XA described in JP-A-55-160078: yLn (where Mn is at least one of Mg, Ca, Sr, Zn and Cd) species, X is at least one of CI, Br and 1, A is BeO, MgO, CaO+SrO,
BaO, ZnO.

AI 10B , Y2O1+ La10311nto,
l S 10t lT'tol l Zr011・G
501.5n021 At least one of Nb, o, l Ta, 01 and The, Ln is Eu, Tb,
Co, Tm, Dy, Pr, Ho, Nb.

At least one of Yb, Er, Sm, and Gd, and each of X and y is 5 X to-'! x l-
This is a number that satisfies the conditions 0, 5 and 0<y≦0.2.

) Rare-earth element-activated two-li4 fluorono-ride phosphor, ZnS:ASCdS:A, (Zn,
Cd) S: A, SnS: A, X and CdS: A, X
(However, A is Cu + Ag + Au or Mu,
X is) Rogen. ) xM, (PO,)t→■, :yA Ms (POa )z・FA (where M and N are Mg and Ca+Sr, respectively) ,Ba,
At least one of Zn and Cd, X is F, CI,
Br.

and the lesser of 1 (both 1 food, A is Eu, Tb*Ce
.

Tm, Dy+Pr, Ha, Nd, Yb, Er, Sb, T
I, Mn, and Sn (each represents one kind).

Further, X and y are numbers that satisfy the condition 0x≦6.0≦y≦1. ), and nReX1 ・mAX, '2 xEunReX3
IImA) 5': xEu, ysm (in the formula,
Re is the least of La, Gd, Y, Lu (all 1 type,
A represents an alkaline earth metal, and at least one of Ba, Sr, and Ca, and X and X' represent at least one of F, CI, and Br. Moreover, X and y are 1 xlo-
'<x(3XIO"", t xtO-'< y <
i Xl0-' is a number that satisfies the condition, and n7m satisfies the condition that txto-'<n7m<7×1O-1. ) and the phosphor represented by M'X-aMIIX2'・bM"4' : cA(
However, M! is at least one alkali metal selected from Li, Na, Rh, and Cs;
is at least one divalent metal selected from Be, Mg, Ca, Sr, Ba, Zn, Cd, Cu and Ni. Mll [kt S e .

Y, LatCs, Pr, Nd, Pm, Sm, Eu, Gd
+Tb+Dy+HO+Er, Tm, Yb, Lu, A
At least one trivalent metal selected from I, Ga, and In.

x, x' and
u, Tb.

Co, Ttn, Dy, Pr, He, Nd1Yb, Er,
It is at least one metal selected from Gd, Lu18m, Y, T1#Na, Ag, Cu, and Mg. Also a
is O≦a (a numerical value in the range of 0, 5, b is 0-4b
(The numerical value is in the range of 0.5, and C is the numerical value in the range of O<C60.2).

However, the stimulable phosphor according to the present invention is not limited to the above-mentioned phosphors, but any phosphor that exhibits stimulated luminescence when irradiated with radiation and then irradiated with stimulable excitation light. It may be the body.

Hereinafter, the present invention will be described in detail based on the drawings.

FIG. 1 schematically shows a radiation image conversion system using a photostimulable phosphor, which is one embodiment of the present invention.

When the radiation emitted from the radiation generating device 11 and transmitted through the subject 12 is absorbed by the stimulable phosphor material 13, the stimulable phosphor material 13 accumulates radiation energy and forms a latent image. By scanning this stimulable phosphor material with the stimulable excitation light from the stimulable excitation light source 14, the accumulated radiation energy is detected as fluorescence by the photoelectric converter 15 and converted into an electrical signal. After passing through an amplifier, the electric signal is A/D converted and sent to the arithmetic unit 1 in the form of a digital signal.
6, is subjected to the image processing of the present invention, and is reproduced and displayed by the radiation image reproduction device 17.

FIG. 2 schematically shows an X-ray photography system using an X-ray film on which X-ray image information is recorded, which is another embodiment of the present invention.

In the X-ray image reading device 21, the original X-ray photograph n
is scanned by light from a reading light source n.

The light transmitted through the original X-ray photograph is converted into an electrical signal by a photoelectric converter. As an example of this X-ray image reading device 21, it has an aperture and a photodetector on the outside of a transparent drum with a reading light source placed inside, and an original photograph is placed closely on the outer periphery of the drum, and the drum is rotated at the same time. An example is a device that is moved axially and directs the light that has passed through the original photograph through an aperture and onto a photoelectric converter.

The electric signal obtained by converting the original photographic density passes through an amplifier, is A/D converted, is stored as a digital signal in the arithmetic unit δ, and is subjected to the image processing of the present invention.

The digital signal may be temporarily recorded on a recording material such as a magnetic tape or an optical disk, and then read out to a calculation device to perform calculations.

FIG. 3 shows a schematic diagram of a skull image obtained by the Caldwell method. The maxillary sinus in Figure 3 is also important for the diagnosis of the paranasal sinuses. In the image processing method of the present invention, it is preferable to select the image area inside the eye socket 31 in FIG. 3 as the image area serving as the density reference, or more specifically, select, for example, point A in FIG. 3. FIG. 4 shows a preferred example of a gradation processing method for a head image including the paranasal sinuses, in which gamma is constant in the electrical signal range from the minimum value Sm1n to the maximum value Smax of the electrical signal level of the subject part. This is a gradation conversion curve representing such a gradation processing method.

Here, the electrical signal level at the orbital region is represented by Ss, and the optical density corresponding to S is represented by Do. The De is 1.2 to 2.
5, more preferably set to 1.5 to 2.2.

FIG. 5 shows another preferred example of the gradation processing method for a head image including the paranasal sinuses. In FIG. 5(1), the gamma of the paranasal sinuses is higher than that of the teeth.

Further, FIG. 5(2) shows a gradation processing method in which the gamma of the paranasal sinus region is made higher than the gamma of the tooth region and the cervical soft tissue region. In the method of FIG. 5(1), the electric signal S is intermediate between the electric signal of the tooth portion and the electric signal of the sinus portion, and in the method of FIG. 5(2), the electric signal S and the electric signal of the sinus portion are The curve may be created so that the gradient of the gradation conversion curve changes at the electrical signal S2 between the electrical signal of It is natural that the key conversion curve be a smooth curve as shown by the solid line in FIG.

The electric signal level of the orbital part necessary for creating a gradation conversion curve is determined by using a radiation image read from a recording material or an image obtained by thinning out pixels from the radiation image to reduce the number of pixels.
This can be determined by displaying it on the RT and specifying the position of the eye socket using a pointing device such as a trackball or joystick. The electrical signal ranges of the paranasal sinuses, teeth, and neck soft tissue regions can be similarly determined using a pointing device or by using a histogram of electrical signal levels. The histogram of a head image including the paranasal sinuses does not necessarily show the same shape due to the type of radiation image recording material, the imaging method, and individual differences between patients; FIG. 6 shows the outline of the histogram of the head image taken using the method. Here again, the horizontal axis is expressed on a scale corresponding to the logarithmic representation of the radiation exposure amount. A peak 61 that appears on the side where the electrical signal level is low corresponds to the tooth portion, and a peak 64 that appears on the side where the signal is high in the subject area corresponds to the soft tissue of the neck. The sinuses including the maxillary sinus, the orbits, etc. exist in the complex mountain-shaped signal region between the two. Therefore, it is sufficient to specify the electrical signal levels of Sl and S as shown in FIG. 6, respectively. Alternatively, find the minimum value Sm1n and maximum value Smax of the electrical signal level in the subject area, set the electrical signal level higher than the minimum value Sm1n by 5 to 35% of the difference between the two to 81, and
A similar effect can be obtained even if the electric signal level that is higher than the minimum value Sm1n by 60 to 90% of the difference between m1n and Smax is defined as S.

(Example) Next, the present invention will be specifically explained with reference to Examples.

Example I Four patients with A: sinusitis, B: benign tumor, C malignant tumor, and Dll (all diseased areas are in the maxillary sinus) were subjected to cranial X-rays using a conventional screen/film system. I took a photo. The shooting conditions were as follows.

Head X-ray image information was read from the X-ray photograph obtained as described above and the original photograph 1-1 by the method of the present invention, and the following image processing of the present invention was applied to each.

Processing 1) Gradation processing was performed using a gradation conversion curve as shown in FIG. Here, as Se, the electrical signal level for point A in FIG. 3 was taken, and De=1.9. Gamma in Se was set to 2.0.

Processing 2) Gradation processing was performed using a gradation conversion curve as shown in FIG. 5(1). Here, Se and Da were set in the same manner as in Process 1. Sl was determined by a method using a histogram of electrical signal levels, and gamma in the electrical signal region between Sl and Smax was set to 3.0.

Processing 3) Gradation processing was performed using a gradation conversion curve as shown in FIG. 5(2)K. Here, Se and De were set in the same manner as in the case of [Processing]. SI and S2 are determined by a method using a histogram of electrical signal levels, and S and 82
Gamma and ma in the electrical signal region between the two were set to 3.4.

Processing 4) For the head image subjected to processing], the modulation transfer function becomes maximum when the spatial frequency is 0.6c/tomo,
(Spatial) frequency emphasis was performed such that the maximum value was 2.0.

Reproducing the image subjected to the image processing on an X-ray film,
A reproduced image was obtained.

When an otorhinolaryngologist was asked to evaluate these reproduced images, it was found that all the reproduced images obtained by the image processing method of the present invention were easy to compare with the original photographs and were evaluated as having higher diagnostic ability.

In particular, treatments 2 and 3 were highly effective for patients A and B. The reasons for this are: A: It has become easier to determine the size and shape of the mucosal thickening area.

B: It is easy to see the boundary between the maehara and the cheekbone.

etc. are listed.

Furthermore, for patients C and D, treatment 4 was highly effective. The reasons for this are: C: The degree of virility of bone destruction has become easier to diagnose.

D: The margin of the cyst became clear.

etc. are listed.

Example 2 For the four patients who were the subjects in Example 1, their head images were recorded on a stimulable phosphor material by the method of the present invention, and the images were read and subjected to the same procedure as in Example 1 for each patient. Four types of image processing similar to those described above were performed. Note that the radiation irradiation conditions at the time of recording radiation image information are the same as in the first embodiment.

Reproducing the image subjected to the image processing on an X-ray film,
A reproduced image was obtained.

We asked an otorhinolaryngologist to evaluate these reproduced images. The X-ray photograph used as the original photograph in Example 1 was selected for comparison. As a result, all images reproduced by the image processing method of the present invention were easier to see than X-ray photographs taken using a conventional screen/film system, and were evaluated as having higher diagnostic ability.

Evaluations for individual patients were almost the same as in Example 1.

[Brief explanation of drawings]

Figure 1 is a schematic diagram of a radiographic image conversion system that is one embodiment of the present invention, Figure 2 is a schematic diagram of a radiographic system that is another embodiment of the present invention, and Figure 3 is a skull using the Caldwell method. 4 is a graph showing an example of preferable gradation processing for a head image including sinuses, and FIG. 5 is a graph showing another example of preferable gradation processing for a head image including sinuses. , FIG. 6 is a histogram of a head image taken using a stimulable phosphor using the Caldwell method. 13... Stimulable phosphor material, 15... Photoelectric conversion device, 16 and 5... Arithmetic device, 17 and addition... X-ray image reproducing device, 31... Eye socket, 32, E and Synonyms...Sinuses, Applicant Roku Konishi Photo Industry Co., Ltd. t/: Hg'r13 j;;1::°-;1=
:, ji1:;! '! r72: Renkoma 173; 1 exhaustion) Take 5 + Masu 14: Zen Gaku Senkitan 15 - Difference Sphing Hinoki 7t: Lumbar wheel device / 7: 轡A 畔KYOME 1 Fig. 2 X rope shell nuyo Figure 3: 4 Geta ＼ name, Pe..., Figure 5: Denki Baisaki, Figure 6: Otsul-Yasuke

Claims (5)

[Claims] (1) By scanning a recording material that records radiographic images of the head including the paranasal sinuses, the radiographic image information is read, converted into an electrical signal, and a reproduced image is obtained from the electrical signal. 1. An image processing method for a head radiation image including sinuses, characterized in that an orbital region is reproduced with an optical density in a range of 1.2 to 2.5. (2) In the reproduced image, the gamma of the sinus area is 1
．． The image processing method of a head radiation image including the paranasal sinuses according to claim 1, characterized in that it is in the range of 5 to 3.8. (3) An image of a head radiation image including the paranasal sinuses according to claim 1 or 2, wherein the reproduced image has constant contrast over the entire electrical signal range of the subject part. Processing method. (4) In the reproduced image, the contrast of the paranasal sinus region is higher than the contrast of the tooth region in the head, and the contrast of the paranasal sinus region is higher than the contrast of the cervical soft tissue region in the head. Claim 1 or 2 is characterized in that:
An image processing method for a head radiation image including the paranasal sinuses as described in Section 1. (5) The electrical signal is characterized by applying spatial frequency emphasis such that the value of the spatial frequency when the modulation transfer function takes the maximum value is in the range of 0.1 to 1.0 c/mm. An image processing method for a head radiation image including the paranasal sinuses according to any one of claims 1 to 4.