JP4879472B2 - Cerebral blood flow quantitative analysis program, recording medium, and cerebral blood flow image data processing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a cerebral blood flow determination analysis program or the like, capable of performing brain image reconstitution without performing subtraction when using a split-dose method of twice or more number of times, avoiding dispersion in each operator in selection of a slice including a basal ganglia and in setting of ROI, preventing occurrence of dispersion in a determination image, and removing complexities for the operator. <P>SOLUTION: Split-dose of twice or more of a tracer labeled by RI is applied to a subject brain, and each piece of projection data acquired by performing SPECT measurement, after each split-dose, are inputted. Then, a reconstructed image after each split-dose is acquired, based on each inputted piece of projection data, and each standard brain image is acquired by performing anatomical normalization on the reconstituted image, after each split-dose. Each determination image can be acquired by performing prescribed determination to each acquired standard brain image. When the reconstituted image is acquired, IBL method or ABL method is used. Lassen correction is performed on a SPECT image, after anatomical normalization. <P>COPYRIGHT: (C)2006,JPO&amp;NCIPI

Description

  The present invention relates to a cerebral blood flow quantitative analysis program that performs quantitative analysis of cerebral blood flow by computer tomography measurement, and more particularly to cerebral blood flow by single photon emission computed tomography measurement using a tracer labeled with RI. The present invention relates to a cerebral blood flow quantitative analysis program for performing quantitative analysis.

  2. Description of the Related Art In recent years, an image diagnostic machine for capturing a state in a body as an image without imposing a heavy burden on a patient or the like and performing an accurate diagnosis has become indispensable in a medical field. Examples of such an image diagnostic machine include an X-ray tomography (CT), a magnetic resonance imaging (MRI), an ultrasonic diagnostic machine, and a radiological diagnostic machine. Image diagnosis using these image diagnosis machines is widely used as providing functional information such as early diagnosis of a disease of a patient, selection of a treatment method, prediction and determination of a treatment effect, and the like.

  In the clinical field of nuclear medicine, for example, as described in Non-Patent Document 1, a radioisotope (RI) is introduced into a patient and a single gamma ray emitted from the isotope is used. Photon emission computed tomography (SPECT) and positron emission tomography (PET) are used by using the respective apparatuses. In the medical image processing apparatus as described above, various computer programs are prepared so that various image processing such as image reconstruction and image analysis can be performed from the collected projection data.

  Tc (Technesium) -99mECD (ethyl cysteinate dimer, hereinafter abbreviated as “ECD”) is often used as a short half-life tracer which is a radiopharmaceutical for cerebral blood flow SPECT. . The feature of this tracer is that it has good stability after labeling in which some of the constituent elements of the compound are replaced with RI. Since the distribution in the brain is fixed within a few minutes after the tracer is taken into the brain tissue, it is a characteristic property to maintain the distribution for a long time. Utilizing this characteristic property, a split-dose method and a protocol for evaluating cerebral circulation reserve by acetazolamide (acetazolamide, ACZ) loading are used in clinical practice.

  As described in Non-Patent Document 1, the twice-divided administration method is a method in which a dose tracer that is usually used once is divided into two doses, and each dose is changed continuously. This is a method of taking a brain image of the above. FIG. 39 shows the protocol for two divided doses. In FIG. 39, the horizontal axis represents the elapsed time (minutes), and the time when the first ECD was intravenously administered is 0 minute. As shown in FIG. 39, collection of the first SPECT data (SPECT1, at rest) is started from time t1 (approximately 5 minutes). This is because the distribution of ECD in the brain is solidified after about 5 minutes. At time t1, Diamox (registered trademark) is intravenously injected for a load test. Diamox (registered trademark) is a carbonic acid dehydrogenase inhibitor acetazolamide, which has a selective and strong cerebral vasodilatory action. SPECT1 data is collected until time t2, and the second ECD is intravenously injected at time t2. The collection of the second SPECT data (SPECT2, when loaded) is started from time t3 when the distribution of the second ECD in the brain is solidified. SPECT2 data is collected until time t4. The effect of Diamox (registered trademark) hardly appears between time t1 and t2, but appears between time t2 and t3. As a result, the SPECT2 data is data in a state where the distribution of the ECD in the brain for the second time when the cerebral blood vessel is dilated is solidified. The collection time of SPECT1 data is a waiting time until the effect of Diamox (registered trademark) is manifested.

An image of the local cerebral blood flow distribution under the second condition (at the time of loading) can be obtained by subtraction (subtraction: SUB) of SPECT1 data from SPECT2 data. FIG. 40 shows the concept of the subtraction method. FIG. 40A shows SPECT1 data, and FIG. 40B shows SPECT2 data. As described above, SPECT detects γ-rays emitted from RI introduced into a patient, but 30 to 40% of the total collection count is scattered ray components, and most of these scattered rays are Compton scattered rays. . That is, the SPECT data includes a non-scattering component (primary component) and a scattering component caused by Compton photons. In FIG. 40A, reference numeral 104 denotes a primary component P _i ^[1st] of SPECT1 data, 102 denotes a scattering component S _i ^[1st] of SPECT data 1, and y _i ^[1st] obtained by combining both is SPECT1. It is data. In FIG. 40B, reference numeral 108 denotes a true primary component P _i ^{[2nd true]} of SPECT2 data, and 106 denotes a true scattering component S _i ^{[2nd true]} of SPECT2 data. True SPECT2 data. As shown in FIG. 40 (B), SPECT2 data y _i ^[2nd] is added with SPECT1 data y _i ^[1st], which is the residual radioactivity in the brain by the first ECD administration, as a baseline. Therefore, SPECT2 data y _i ^[2nd] is a combination of SPECT1 data y _i ^[1st] and true SPECT2 data. Therefore, in the conventional brain imaging in two divided doses method yielded obtain a reconstructed image of regional brain blood flow distribution by subtraction of SPECT1 data y _i ^[1st] from SPECT2 data y _{i ^[2nd]} . The reconstruction is to create a transverse image from the SPECT data described above using an FBP (filtered backprojection) method or an OSEM (ordered subsets-expectation maximization) method that is a successive approximation type image reconstruction method. That is, the SPECT1 data is reconstructed to create a transverse image (1st transverse image), and the data obtained by the above-described subtraction is reconstructed to create a transverse image (2 sub transverse image). These 1st transverse image and 2sub transverse image are called qualitative SPECT images.

  On the other hand, RIangography (blood vessel projection) was performed simultaneously with the first intravenous injection of ECD (time 0) (about 2 minutes), and the PA using RI accumulated in the aortic arch and RI accumulated in the brain was used. An average cerebral blood flow value in the normal hemisphere is calculated by a track plot method or a BUR (brain uptake ratio) method. Lassen correction using the average cerebral blood flow value is performed on the above-mentioned 1st cross-sectional image to convert it into a transverse image of a resting quantitative cerebral blood flow value. Since ECD uptake into brain tissue is not linearly related to cerebral blood flow, it is underestimated in high blood flow regions. Therefore, the high blood flow region is lifted and corrected by correcting the Lassen.

Next, the average cerebral blood flow value after loading is obtained from the basal ganglia SPECT count in the 2 sub transverse image, the basal ganglia SPECT count in the 1st transverse image, and the average cerebral blood flow value. FIG. 41 is a diagram for explaining selection of a slice (or frame) including a basal nucleus. In FIG. 41, slices 19 to 36 are shown. Of these, slices 28 to 30 indicated by symbol S are selected as slices including the base nucleus, and the slices are added. An ROI (region of interest) is set as a reference region of the normal hemisphere or the milder hemisphere of the slice-added image. 42A and 42B show images obtained by adding slices. In FIG. 42A, the symbol ROIa indicates the ROI set in the normal cerebral hemisphere or the milder cerebral hemisphere. When both cerebral hemispheres show a uniform distribution, ROIb is set so as to surround the entire cerebrum as shown in FIG. The average SPECT value in ROIa or ROIb set as described above is obtained. Using this average SPECT value and the average cerebral blood flow value after loading, as in the case of the 1st cross-sectional image, Lassen correction is performed on the 2 sub cross-sectional image to obtain a cross-sectional image of the quantified cerebral blood flow value on load Convert. The converted 1st transverse image and 2sub transverse image obtained as described above are referred to as a quantitative SPECT image. These quantitative images are displayed by desired display software, and a local cerebral blood flow analysis is performed.
Tsunehiko Nishimura, Clinical version of revised brain SPECT / PET-Examination of brain function, pp. 12-17, Medical View Inc.

The above-described two-part administration method has a problem that the signal / noise ratio (S / N ratio) of the load image is deteriorated by subtracting the SPECT1 data from the SPECT2 data. When the dose ratio is 1: 1, it is said that the dispersion of the second reconstructed image is deteriorated three times (that is, the standard deviation is three times worse than the root) due to subtraction. Theoretically, it is said that the first and second reconstructed images have the same S / N ratio when the dose ratio is 1: 2. To avoid such degradation of the image, it is necessary to increase the overall ECD dose, the increase in ECD dose cause excessive radiation exposure to the patient, because that would lead an increase in inspection costs It is not preferable.

  The selection of the slice including the basal ganglia and the setting of the ROI were performed by the operator, which was complicated and inevitable variation among operators. As a result, there is a problem that variation occurs in the quantitative image.

  Therefore, an object of the present invention is to solve the above-described problem, and to perform brain image reconstruction without performing subtraction when using a two-time (or more than two) divided dose method. It is to provide a cerebral blood flow quantitative analysis program and the like.

  The second object of the present invention is to avoid variations among operators in selecting slices including the basal ganglia and setting ROIs, so as not to cause variations in quantitative images, and to eliminate operator complexity. It is to provide a cerebral blood flow quantitative analysis program and the like.

The cerebral blood flow quantitative analysis program according to the present invention is a cerebral blood flow quantitative analysis program for performing quantitative analysis of cerebral blood flow by computer tomography measurement, wherein a predetermined drug is applied to the subject's brain n times (n is 2 or more), and a projection data input step for inputting each projection data obtained by performing the computed tomography measurement after each divided administration, based on each projection data input in the projection data input step, This is an image reconstruction step for obtaining a reconstructed image after divided administration , and using the data obtained by projecting the (n-1) th reconstructed image created by the successive approximation as a baseline in the successive approximation type image reconstruction method seek n-th reconstructed image by, which the baseline is updated each time, re after each divided doses obtained by the image reconstruction step An anatomical normalization step for obtaining each standard brain image by performing anatomical normalization on the adult image, and performing a predetermined quantification on each standard brain image obtained in the anatomical normalization step It is a cerebral blood flow quantitative analysis program for executing a quantification step for obtaining each quantitative image.

Here, in the cerebral blood flow quantitative analysis program of the present invention, the computed tomography measurement is a single photon emission computed tomography measurement or a positron emission tomography measurement, and the predetermined drug is labeled with a radioisotope (RI). The image reconstruction step includes: m is a total number of pixels on the reconstructed image, j (= 1 to m) is a pixel on the reconstructed image, and n is a radioisotope administered to the brain. The total number of pixels on the projection data onto which the photons or positrons emitted from the projection are projected or the total number of detectors that detect the photons or positrons, and i (= 1 to n) is the pixel on the projection data, where m and n are recorded in advance pixel number recording unit, and the probability that the photon or positron emitted from a pixel j in the reconstructed image the c _ij are detected on a pixel i on the projection data, wherein c _j is recorded in advance from j = 1 m and up i = 1 from the detection probability recording unit to n, the number L divided doses from h = 1 (L ≧ 2), the lambda _{j 0} ^[h] , The estimated RI density of the pixel j on the reconstructed image in the 0th successive approximation used for the hth divided administration, where λ _j ^{0 [h]} is estimated in the hth time from j = 1 to m in advance. Count of photons or positrons detected at one pixel i on the projection data after recording h _i of divided doses y _i ^[h] for h = 1 to L. Y _i ^[h] is recorded in the h-th measurement data recording section from i = 1 to n after detection, and S _i ^[h] is divided into the h-th division for h = 1 to L. and scatter correction term in 1 pixel i on the projection data after the administration, where S _{i ^[h]} is the i = 1 to n It is recorded in time scatter correction term recording unit, a reconstructed image of the divided administration of the first round, k-th pixel j in the reconstructed image in iterative estimation RI concentration lambda _{j k} ^[1] the formula A first reconstructed image generation step obtained by repeatedly calculating from k = 0 to a predetermined number of repetitions using (1);

Here, λ _j ^{0 [1]} is recorded in the first estimated RI concentration initial value recording unit, c _ij is recorded in the detection probability recording unit, and y _i ^[1] is recorded in the first measurement data recording unit. S _i ^[1] is recorded in the first scatter correction term recording unit, and the reconstructed image after the h-th divided administration from h = 2 to L is displayed for the k-th time. The h-th reconstructed image obtained by repeatedly calculating the estimated RI density λ _j ^{k [h]} of the pixel j on the reconstructed image in the successive approximation from k = 0 to a predetermined number of repetitions using Equation (2). Generating step,

Here, λ _j ^{0 [h]} is recorded in the h-th estimated RI concentration initial value recording unit, c _ij is recorded in the detection probability recording unit, and y _i ^[h] is recorded in the h-th measurement data recording unit. And S _i ^[h] may be recorded in the h-th scattering correction term recording unit.

  Here, in the cerebral blood flow quantitative analysis program of the present invention, the formula (3) is used instead of the formula (1).

Can be used.

  Here, in the cerebral blood flow quantitative analysis program according to the present invention, in the anatomical normalization step, with respect to the reconstructed images after the divided administration, the reconstructed images after the divided administration after the second time are displayed for the first time. A positioning step for performing correction to match the same position as the reconstructed image after divided administration, and a reconstructed image after the correction in the positioning step are deformed so as to match the standard brain using spatial deformation parameters Thus, a standardization step for obtaining each standard brain image and a smoothing step for applying a smoothing filter to the reconstructed image after the standardization step can be provided.

  Here, in the cerebral blood flow quantitative analysis program according to the present invention, the deformation parameter is obtained as total projection data that is a sum of projection data weighted with a predetermined weight, and a reconstructed image is obtained based on the total projection data. Thereafter, the reconstructed image is subjected to anatomical normalization to obtain a standard brain image, and can be obtained based on the processing.

  Here, in the cerebral blood flow quantification analysis program according to the present invention, the predetermined quantification in the quantification step uses an average cerebral blood flow value at rest of the subject input for each standard brain image. By performing Lassen correction, each cerebral blood flow can be obtained and each quantitative image can be obtained.

  Here, the cerebral blood flow quantitative analysis program according to the present invention may further include a display step of displaying each quantitative image obtained in the quantifying step.

  Here, in the cerebral blood flow quantitative analysis program according to the present invention, the increase amount and / or the increase rate of the local cerebral blood flow in a predetermined region are displayed in a predetermined display format based on each quantitative image obtained in the quantification step. An increase amount display step for displaying can be further provided.

  Here, in the cerebral blood flow quantitative analysis program of the present invention, the predetermined display format can display a local cerebral blood flow value for each divided administration in a graph.

  Here, in the cerebral blood flow quantitative analysis program according to the present invention, the predetermined display format includes the local cerebral blood flow value in the h-th (h is 2 or more) divided administration and the local brain in the divided administration before the h-th administration. The pixel-by-pixel difference from the blood flow value can be displayed in a distinguishable manner using a predetermined color.

  Here, in the cerebral blood flow quantitative analysis program according to the present invention, the predetermined display format is an average value of the local cerebral blood flow value in a predetermined region of interest in the h th (h is 2 or more) divided administration and the h th time. The difference from the average value of the local cerebral blood flow values in the predetermined region of interest in the earlier divided administration can be displayed using a predetermined color in an identifiable manner.

  Here, in the cerebral blood flow quantitative analysis program according to the present invention, prior to the image reconstruction step, projection of one slice constituting the projection data after the first divided administration input in the projection data input step A confirmation display step of applying a predetermined filter to the data and displaying the reconstructed image can be further provided.

  Here, in the cerebral blood flow quantitative analysis program according to the present invention, the average value of local cerebral blood flow values in a predetermined region of interest can be identified using a predetermined color based on each quantitative image obtained in the quantification step. Can be displayed.

The cerebral blood flow quantitative analysis program of the present invention is a cerebral blood flow quantitative analysis program for performing quantitative analysis of cerebral blood flow by computer tomography measurement, and a predetermined drug is administered to the brain of n subjects in the computer. A projection data input step for inputting projection data obtained by performing the computed tomography measurement after administration, and an image reconstruction step for obtaining a reconstructed image after administration based on the projection data input in the projection data input step In the successive approximation type image reconstruction method, the nth reconstruction image is obtained by using, as a baseline, data obtained by projecting the (n−1) th reconstruction image created by the successive approximation, and the baseline is obtained. is intended to update each time, target performs anatomical normalized to the reconstructed image after administration obtained by the image reconstruction step Anatomical normalization step for obtaining a brain image, cerebral blood flow quantification for executing a predetermined quantification step for obtaining a quantitative image by performing a predetermined quantification on the standard brain image obtained in the anatomical normalization step It is an analysis program.

  Here, in the cerebral blood flow quantitative analysis program according to the present invention, the anatomical normalization step is performed by deforming the reconstructed image after the administration so as to match the standard brain using a spatial deformation parameter. A standardization step for obtaining a standard brain image and a smoothing step for applying a smoothing filter to the reconstructed image after the standardization step can be provided.

  Here, in the cerebral blood flow quantitative analysis program according to the present invention, the deformation parameter is obtained as a standard brain image by obtaining a reconstructed image based on the projection data and then performing anatomical normalization on the reconstructed image. It is possible to perform the process of obtaining the above and obtain the process based on the process.

  Here, in the cerebral blood flow quantification analysis program of the present invention, the predetermined quantification in the quantification step is performed by using Lassen's average cerebral blood flow value of the subject at rest with respect to a standard brain image. By performing correction, a local cerebral blood flow can be obtained and a quantitative image can be obtained.

  Here, the cerebral blood flow quantitative analysis program according to the present invention may further include a display step for displaying the quantitative image obtained in the quantification step.

  Here, in the cerebral blood flow quantitative analysis program according to the present invention, prior to the image reconstruction step, a predetermined amount of projection data for one slice constituting the post-administration projection data input in the projection data input step is added. A confirmation display step of filtering and displaying the reconstructed image can be further provided.

  Here, in the cerebral blood flow quantitative analysis program of the present invention, based on the quantitative image obtained in the quantifying step, the average value of the local cerebral blood flow values in a predetermined region of interest can be identified using a predetermined color. Can be displayed.

  The recording medium of the present invention is a computer-readable recording medium on which any one or a plurality of cerebral blood flow quantitative analysis programs of the present invention are recorded.

The cerebral blood flow image data processing method according to the present invention is a cerebral blood flow image data processing method for processing cerebral blood flow image data by computer tomography measurement, wherein the computer applies a predetermined drug to the subject's brain n times. (n is 2 or more) comprising the steps of inputting the brain image data when it is divided doses, projection data input for inputting the projection data obtained said performing computed tomography measured after each divided doses And an image reconstruction step for obtaining a reconstructed image after each divided administration based on each projection data input in the projection data input step, which is created by successive approximation in the successive approximation type image reconstruction method. The nth reconstructed image is obtained by using the data obtained by projecting the (n-1) th reconstructed image as the baseline, and the baseline is updated every time. An anatomical normalization step for obtaining each standard brain image by performing anatomical normalization on the reconstructed image after each divided administration obtained in the image reconstruction step, and the anatomical normalization And a quantification step for obtaining each quantification image by performing a predetermined quantification on each standard brain image obtained in the quantification step.

Here, in the cerebral blood flow image data processing method of the present invention, the computed tomography measurement is a single photon emission computed tomography measurement or a positron emission tomography measurement, and the predetermined agent is labeled with a radioisotope (RI). The image reconstruction step includes: m is a total number of pixels on the reconstructed image, j (= 1 to m) is a pixel on the reconstructed image, and n is a radioisotope administered to the brain. The total number of pixels on the projection data onto which the photons or positrons emitted from the element (RI) are projected or the total number of detectors that detect the photons or positrons, and i (= 1 to n) as the pixels on the projection data. Here, m and n are recorded in the pixel number recording unit in advance, and c _ij is a probability that a photon or a positron emitted from the pixel j on the reconstructed image is detected at one pixel i on the projection data, This Here, c _ij is recorded in the detection probability recording unit from j = 1 to m and i = 1 to n in advance, and for h = 1 to the number L of divided doses (L is 2 or more), λ _j ^{0 [ h]} is the estimated RI density of the pixel j on the reconstructed image in the 0th successive approximation used for the h-th divided administration, where λ _j ^{0 [h]} is the first from j = 1 to m h times estimated RI concentration initial value recording unit, and for h = 1 to L, y _i ^[h] is a photon detected by one pixel i on the projection data after the h-th divided administration or The number of positrons is counted, where y _i ^[h] is recorded in the h-th measurement data recording section from i = 1 to n after detection, and for h = 1 to L, S _i ^[h] is the hth and scatter correction term in 1 pixel i on the projection data after divided doses times th, where S _{i ^[h]} from i = 1 Until being recorded to the h time scatter correction term recording unit, a reconstructed image of the divided administration of the first round, the estimated RI concentration lambda _{j k} pixel j in the reconstructed image in k-th iterative ^{[1 ]} and 1st reconstructed image generating step of obtaining by iteration until a predetermined number of iterations from k = 0 using equation (4),

Here, λ _j ^{0 [1]} is recorded in the first estimated RI concentration initial value recording unit, c _ij is recorded in the detection probability recording unit, and y _i ^[1] is recorded in the first measurement data recording unit. S _i ^[1] is recorded in the first scatter correction term recording unit, and the reconstructed image after the h-th divided administration from h = 2 to L is displayed for the k-th time. The h-th reconstructed image obtained by repeatedly calculating the estimated RI density λ _j ^{k [h]} of the pixel j on the reconstructed image in the successive approximation from k = 0 to a predetermined number of repetitions using Equation (5). Generating step,

Here, λ _j ^{0 [h]} is recorded in the h-th estimated RI concentration initial value recording unit, c _ij is recorded in the detection probability recording unit, and y _i ^[h] is recorded in the h-th measurement data recording unit. And S _i ^[h] may be recorded in the h-th scattering correction term recording unit.

Here, in the cerebral blood flow image data processing method of the present invention, instead of the formula (4), the formula (6)

Can be used.

Here, in the cerebral blood flow image data processing method according to the present invention, the anatomical normalization step uses a first reconstructed image after divided administration for the second and subsequent divided images. An alignment step for performing correction to match the same position as the reconstructed image after the divided administration for the second time, and a reconstructed image after the correction by the alignment step are deformed so as to match the standard brain using spatial deformation parameters Thus, a standardization step of obtaining each standard brain image and a smoothing step of applying a smoothing filter to the reconstructed image after the standardization step can be provided.

Here, in the cerebral blood flow image data processing method according to the present invention, the deformation parameter is obtained as total projection data that is a sum of projection data weighted with a predetermined weight, and a reconstructed image is obtained based on the total projection data. Thereafter, the reconstructed image is subjected to anatomical normalization to obtain a standard brain image, and can be obtained based on the processing.

Here, in the cerebral blood flow image data processing method of the present invention, the predetermined quantification in the quantification step is performed using the average cerebral blood flow value at rest of the subject input for each standard brain image. By performing Lassen's correction, each cerebral blood flow can be obtained and each quantitative image can be obtained.

Here, in the cerebral blood flow image data processing method of the present invention, prior to the image reconstruction step, one slice constituting the projection data after the first divided administration input in the projection data input step. A confirmation display step of applying a predetermined filter to the projection data and displaying the reconstructed image can be further provided.

Cerebral blood flow image data processing method of the present invention is a cerebral blood flow image data processing method for processing cerebral blood flow image data by a computer tomography measurements, given drug is administered to the n's subject's brain A projection data input step for inputting projection data obtained by performing the computed tomography measurement after that, and an image reconstruction step for obtaining a reconstructed image after administration based on the projection data input in the projection data input step In the successive approximation type image reconstruction method, the nth reconstruction image is obtained by using, as a baseline, data obtained by projecting the (n−1) th reconstruction image created by the successive approximation, and the baseline is obtained. each is intended to update, obtain the standard brain image by performing anatomical normalized to the reconstructed image after administration obtained by the image reconstruction step Anatomical normalization step, characterized in that a quantification step for obtaining the quantitative image by performing a predetermined quantification against a standard brain image obtained by the anatomical normalization step.

Here, in the cerebral blood flow image data processing method according to the present invention, the anatomical normalization step deforms the reconstructed image after the administration so as to match the standard brain using a spatial deformation parameter. Thus, a standardization step of obtaining a standard brain image and a smoothing step of applying a smoothing filter to the reconstructed image after the standardization step can be provided.

Here, in the cerebral blood flow image data processing method of the present invention, the deformation parameter is obtained by obtaining a reconstructed image based on the projection data, and then performing anatomical normalization on the reconstructed image to obtain a standard brain It is possible to perform processing for obtaining an image and obtain the image based on the processing.

Here, in the cerebral blood flow image data processing method according to the present invention, the predetermined quantification in the quantification step is performed by using the average cerebral blood flow value at rest of the subject input to the standard brain image. As a result of this correction, a local cerebral blood flow can be obtained and a quantitative image can be obtained.

  The cerebral blood flow quantitative analysis apparatus according to the present invention is a cerebral blood flow quantitative analysis apparatus that performs quantitative analysis of cerebral blood flow by computer tomography measurement, wherein a predetermined drug is applied to the subject's brain n times (n is 2 or more). Each divided dose based on the projection data input by the projection data input means and projection data input means for inputting the projection data obtained by performing the computed tomography measurement after each divided dose. Image reconstructing means for obtaining a reconstructed image later, and anatomical obtaining for each standard brain image by performing anatomical normalization on the reconstructed image after each divided administration obtained by the image reconstructing means The image processing apparatus includes a normalizing unit and a quantifying unit that performs predetermined quantification on each standard brain image obtained by the anatomical normalization unit to obtain each quantitative image.

ここで、λ_ｊ ^０[1]は前記第１回推定ＲＩ濃度初期値記録部に記録され、ｃ_ｉｊは前記検出確率記録部に記録され、ｙ_ｉ ^[1]は前記第１回測定データ記録部に記録されており、Ｓ_ｉ ^[１]は前記第１回散乱補正項記録部に記録されており、ｈ＝２からＬまで、第ｈ回目の分割投与後の再構成画像を、ｋ回目の逐次近似における再構成画像上の画素ｊの推定ＲＩ濃度λ_ｊ ^ｋ[h]を式（２）を用いてｋ＝０から所定の繰返し回数まで繰返し計算することにより求める第ｈ回再構成画像生成手段とを備え、

ここで、λ_ｊ ^０[h]は前記第ｈ回推定ＲＩ濃度初期値記録部に記録され、ｃ_ｉｊは前記検出確率記録部に記録され、ｙ_ｉ ^[h]は前記第ｈ回測定データ記録部に記録され、Ｓ_ｉ ^[ｈ]は前記第ｈ回散乱補正項記録部に記録されているものとすることができる。 Here, in the cerebral blood flow quantitative analysis apparatus according to the present invention, the computed tomography measurement is a single photon emission computed tomography measurement or a positron emission tomography measurement, and the predetermined agent is labeled with a radioisotope (RI). The image reconstructing means includes m as a total number of pixels on the reconstructed image, j (= 1 to m) as pixels on the reconstructed image, and n as a radioisotope administered to the brain. The total number of pixels on the projection data onto which the photons or positrons emitted from the projection are projected or the total number of detectors that detect the photons or positrons, and i (= 1 to n) is the pixel on the projection data, where m And n are previously recorded in the pixel number recording unit, and c _ij is a probability that a photon or a positron emitted from the pixel j on the reconstructed image is detected by one pixel i on the projection data, where c _ij In advance = 1 from to and i = 1 to n m is recorded in the detection probability recording unit, the number L divided doses from h = 1 (L ≧ 2), the lambda _{j 0} ^[h], th the h times The estimated RI concentration of pixel j on the reconstructed image in the 0th successive approximation used for divided administration of λ _j ^{0 [h]} is the h-th estimated RI concentration initial value from j = 1 to m in advance For y = 1 to L, y _i ^[h] is the count of photons or positrons detected at one pixel i on the projection data after the h-th divided administration, Y _i ^[h] is recorded in the h-th measurement data recording section from i = 1 to n after detection, and S _i ^[h] is projected after the h-th divided administration for h = 1 to L and scatter correction term in 1 pixel i on the data, where S _{i ^[h]} is the h time scatter auxiliary i = 1 to n Is recorded in section recording unit, a reconstructed image of the divided administration of the first round, k-th pixel j in the reconstructed image in iterative estimation RI concentration lambda _{j k} ^[1] Equation (1) A first reconstructed image generation means that is obtained by repeatedly calculating from k = 0 to a predetermined number of repetitions using

Here, λ _j ^{0 [1]} is recorded in the first estimated RI concentration initial value recording unit, c _ij is recorded in the detection probability recording unit, and y _i ^[1] is recorded in the first measurement data recording unit. S _i ^[1] is recorded in the first scatter correction term recording unit, and the reconstructed image after the h-th divided administration from h = 2 to L is displayed for the k-th time. The h-th reconstructed image obtained by repeatedly calculating the estimated RI density λ _j ^{k [h]} of the pixel j on the reconstructed image in the successive approximation from k = 0 to a predetermined number of repetitions using Equation (2). Generating means,

Here, λ _j ^{0 [h]} is recorded in the h-th estimated RI concentration initial value recording unit, c _ij is recorded in the detection probability recording unit, and y _i ^[h] is recorded in the h-th measurement data recording unit. And S _i ^[h] may be recorded in the h-th scattering correction term recording unit.

を用いることができる。 Here, in the cerebral blood flow quantitative analysis apparatus according to the present invention, the formula (3) is used instead of the formula (1).

Can be used.

  Here, in the cerebral blood flow quantitative analysis apparatus according to any one of the aspects of the invention, the anatomical normalization unit may calculate a reconstructed image after divided administration for the second and subsequent times with respect to the reconstructed image after each divided administration. An alignment unit that performs correction to match the same position as the reconstructed image after the first divided administration, and a reconstructed image that has been corrected by the alignment unit so as to match the standard brain using a spatial deformation parameter It is possible to provide a standardization unit that obtains each standard brain image by transforming to a smoothing unit that applies a smoothing filter to the reconstructed image after the standardization unit.

  Here, in the cerebral blood flow quantitative analysis apparatus according to the present invention, the deformation parameter is obtained as total projection data that is a sum of projection data weighted with a predetermined weight, and a reconstructed image is obtained based on the total projection data. Thereafter, the reconstructed image is subjected to anatomical normalization to obtain a standard brain image, and can be obtained based on the processing.

  Here, in any one of the cerebral blood flow quantification analysis apparatuses according to the present invention, the predetermined quantification in the quantification means uses an average cerebral blood flow value at rest of the subject input for each standard brain image. By performing Lassen correction, each cerebral blood flow can be obtained and each quantitative image can be obtained.

  Here, the cerebral blood flow quantitative analysis apparatus according to any one of the present invention may further include display means for displaying each quantitative image obtained by the quantifying means.

  Here, in the cerebral blood flow quantification analysis apparatus according to any one of the present inventions, an increase amount and / or an increase rate of the local cerebral blood flow in a predetermined region is determined based on each quantitative image obtained by the quantification means. An increase amount display means for displaying in a display format can be further provided.

  Here, in the cerebral blood flow quantitative analysis apparatus according to the present invention, the predetermined display format can display a local cerebral blood flow value for each divided administration in a graph.

  Here, in the cerebral blood flow quantitative analysis apparatus according to the present invention, the predetermined display format is the local cerebral blood flow value in h-th (h is 2 or more) divided administration and the local brain in divided administration before the h-th administration. The pixel-by-pixel difference from the blood flow value can be displayed in a distinguishable manner using a predetermined color.

  Here, in the cerebral blood flow quantitative analysis apparatus according to the present invention, the predetermined display format is the average value of the local cerebral blood flow value in the predetermined region of interest in the h th (h is 2 or more) divided administration and the h th time. The difference from the average value of the local cerebral blood flow values in the predetermined region of interest in the earlier divided administration can be displayed using a predetermined color in an identifiable manner.

  Here, in the cerebral blood flow quantitative analysis apparatus according to any one of the present inventions, one slice constituting the projection data after the first divided administration input by the projection data input means prior to the image reconstruction means It is possible to further include confirmation display means for applying a predetermined filter to the projection data for the minute and displaying the reconstructed image.

  Here, in the cerebral blood flow quantitative analysis device according to any one of the present invention, based on each quantitative image obtained by the quantifying means, an average value of local cerebral blood flow values in a predetermined region of interest is used in a predetermined color. Can be displayed in an identifiable manner.

  A second cerebral blood flow quantitative analysis apparatus according to the present invention is a cerebral blood flow quantitative analysis apparatus that performs quantitative analysis of cerebral blood flow by computer tomography measurement, wherein a predetermined drug is administered to the brain of a subject, and after administration Projection data input means for inputting projection data obtained by performing the computed tomography measurement, image reconstruction means for obtaining a reconstructed image after administration based on the projection data input by the projection data input means, An anatomical normalization means for obtaining a standard brain image by performing anatomical normalization on the reconstructed image after administration obtained by the image reconstruction means, and obtained by the anatomical normalization means Quantifying means for obtaining a quantitative image by performing predetermined quantification on the standard brain image is provided.

  Here, in the second cerebral blood flow quantitative analysis apparatus according to the present invention, the anatomical normalization means deforms the reconstructed image after the administration so as to match the standard brain using a spatial deformation parameter. By doing so, it is possible to include standardization means for obtaining a standard brain image, and smoothing means for applying a smoothing filter to the reconstructed image after the standardization means.

  Here, in the second cerebral blood flow quantitative analysis apparatus according to the present invention, the deformation parameter is obtained by performing anatomical normalization on the reconstructed image after obtaining a reconstructed image based on the projection data. A process for obtaining a standard brain image can be performed and obtained based on the process.

  Here, in the second cerebral blood flow quantification analysis apparatus according to any one of the present inventions, the predetermined quantification in the quantification means is performed by comparing the standard cerebral image with the average cerebral blood flow at rest of the subject input. By performing Lassen correction using the value, a local cerebral blood flow can be obtained and a quantitative image can be obtained.

  Here, the second cerebral blood flow quantitative analysis apparatus according to any one of the present invention may further comprise display means for displaying a quantitative image obtained by the quantifying means.

  Here, in the second cerebral blood flow quantification analysis apparatus according to any one of the present inventions, prior to the image reconstruction unit, one slice constituting the projection data after administration input by the projection data input unit A confirmation display means for applying a predetermined filter to the projection data and displaying the reconstructed image can be further provided.

  Here, in the second cerebral blood flow quantification analysis apparatus according to any one of the present inventions, an average value of local cerebral blood flow values in a predetermined region of interest is determined in a predetermined color based on the quantitative image obtained by the quantification means. Can be displayed in an identifiable manner.

  According to the cerebral blood flow quantitative analysis program or the like of the present invention, a predetermined drug, for example, a drug containing a tracer labeled with RI is administered to the subject's brain n times (n is 2 or more), and a computer is used after each divided administration. Each projection data obtained by performing tomographic measurement is input. As the computed tomography measurement, SPECT measurement or PET measurement is suitable, but MRI measurement or CT measurement may be used. Next, based on each input projection data (for example, SPECT data), a reconstructed image after each divided administration is obtained, and an anatomical normalization is performed on the reconstructed image after each divided administration. A standard brain image is obtained. Each quantitative image can be obtained by performing predetermined quantification on each standard brain image obtained. When obtaining a reconstructed image, by using the IBL method or ABL method described later, it is possible to perform brain image reconstruction without performing subtraction in the case of using two (or two or more) divided administration methods. it can. In the cerebral blood flow quantitative analysis program of the present invention, Lassen correction is performed on an image such as SPECT after anatomical normalization by using the inputted mean cerebral blood flow value (mCBF) at rest of the subject. Thus, each cerebral blood flow is obtained to obtain each quantitative image. For this reason, the basal ganglia count at the time of loading and the basal ganglia count at rest can be automatically set to the value of the slice image selected as Range 31-40. In the cerebral blood flow quantitative analysis program of the present invention, when setting an ROI in the normal cerebral hemisphere or the milder cerebral hemisphere of the slice-added image, the ROI template can be used. For this reason, there is an effect that the operator can be prevented from being varied in the setting of the ROI, the quantitative image is not varied, and the operator's complexity can be eliminated.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In the computer tomography measurement in the present invention, a predetermined drug, for example, a drug containing a tracer labeled with RI is administered to the subject's brain n times (n is 2 or more), and each projection data is measured after each divided administration. Is. For example, although a SPECT measurement or a PET measurement is suitable, an MRI measurement or a CT measurement may be used. Below, it demonstrates using SPECT measurement as an example.

  FIG. 1 is a flowchart showing a flow of a cerebral blood flow quantitative analysis program or the like in Embodiment 1 of the present invention. First, an outline of the flow of the cerebral blood flow quantitative analysis program and the like will be described, and then each step will be described in detail. As shown in FIG. 1, a tracer labeled with RI is divided into two or more doses to the subject's brain, and each projection data (SPECT data) obtained by performing SPECT measurement after each divided dose is input (projection). Data input step, step S10). Next, a reconstructed image after each divided administration is obtained based on the projection data input in the projection data input step (step S10) (image reconstruction step, step S20). Anatomical normalization is performed on the reconstructed image after each divided administration obtained in the image reconstruction step (step S20) to obtain each standard brain image (anatomical normalization step, step S40). Each standard brain image obtained in the anatomical normalization step (step S50) is subjected to predetermined quantification to obtain each quantitative image (quantification step, step S50).

  In the above description, the cerebral blood flow quantitative analysis program according to the first embodiment of the present invention is described in units of steps. However, since the cerebral blood flow quantitative analysis program is computer software, it goes without saying that each step is a function of the program. Accordingly, the projection data input step, the image reconstruction step, the anatomical normalization step, the quantification step, the display step (described later), etc. are respectively the projection data input function (means), the image reconstruction function (means), and the anatomy. Of course, it may be rephrased as a normal normalization function (means), a quantification function (means), a display function (means), or the like.

  FIG. 2 is a diagram for explaining the projection data input step (step S10) in the first embodiment of the present invention. In the projection data input step (step S10), the double divided administration method (see FIG. 38) described in the background art can be used. Further, as shown in FIG. 2, the dose usually used once is used. These three tracers can be divided into three doses, and three consecutive brain images can be taken under different conditions (three-fold divided dose method). The three-fold divided administration method shown in FIG. 2 is the same as that shown in FIG. As shown in FIG. 2, collection of SPECT2 data (loading time 1) is performed until time t4, and the third ECD is intravenously injected at time t4. The collection of the third SPECT data (SPECT3, when loaded) is started from time t5 when the distribution of ECD in the brain is solidified. SPECT3 data is collected until time t6. Although the three-fold divided administration method has been described with reference to FIG. 2, it is possible to use four or more divided administration methods as necessary in the projection data input step (step S10).

  Next, the image reconstruction step (Step S20) will be described. The cerebral blood flow quantitative analysis program of the present invention is characterized by using an algorithm in which the effect of the baseline by ECD administration up to the previous time (for example, once) is previously incorporated in the successive approximation image reconstruction method (OSEM method). is there. First, the successive approximation type image reconstruction method will be described.

  As a method of estimating the unknown RI distribution in the body as the most likely case using statistical methods from the collected projection data, Maximum Likelihood-Expectation Maximization (ML-EM) There is a construction method. This ML-EM method is a method combining a maximum likelihood estimation method and an expected value maximization method. The ML-EM method is a method of repeatedly correcting the RI distribution in the body using a successive approximation method so that the projection data calculated from the estimated RI distribution in the body is as close as possible to the actually collected projection data.

  FIG. 3 is a diagram for explaining the application of the maximum likelihood estimation method to image reconstruction. In FIG. 3, reference numeral 10 denotes a tomographic image of the brain image, 3 denotes one pixel having an RI density of λ (MBq) in the matrix of the tomographic image 10, and 30 denotes a detector which indicates one pixel on the projection data 20. c indicates a detection probability that a photon is detected at the pixel 30, and x indicates a count of photons from the pixel 3 in one pixel 30 on the projection data 20 (or one detector 30 assuming that the pixel is a detector). . When photons emitted from RI (pixel 3) having an RI concentration of λ (MBq) are measured n times by the detector 30 (pixel 30), the maximum likelihood estimated value of the count x is assumed to be xi for each measurement. Is given by Equation 7.

  The maximum likelihood estimated value of the RI concentration λ is given by Equation 8 because λ × c = x.

Next, a case where RI is measured with an n-direction detector will be described. FIG. 4 is a diagram for explaining the application of the maximum likelihood estimation method to image reconstruction when RI is measured by an n-direction detector. In FIG. 4, the parts denoted by the same reference numerals as those in FIG. In FIG. 4, reference numeral 5 indicates one pixel j (j = 1 to the total number m of pixels in the matrix of the tomographic image 10) having an RI density of λ _j (MBq) in the matrix of the tomographic image 10. Reference numerals 31 to 33 denote detectors each indicating one pixel on the projection data 20 or the like. Although there are n detectors, only three are shown for convenience of drawing. c _ij (i = 1 to the number n of detectors or the number n of each pixel on the projection data, j = 1 to m) is the photon emitted from the pixel j (5) is the pixel i (or the detector i). X _ij indicates the number of photons emitted from pixel j (5) and detected by pixel i (or detector i). Under the above conditions, the maximum likelihood estimated value of the density λ _j of the pixel j (5) is given by Equation 9.

Actually, since RI exists also around the pixel j (5), the photon number x _ij cannot be directly measured. FIG. 5 is a diagram for explaining the application of the maximum likelihood estimation method to image reconstruction when the RI around the pixel j (5) is considered. In FIG. 5, the same reference numerals as those in FIG. In FIG. 5, y _i indicates the total sum (projection count) of the measurement values from each pixel j (5) and the like in the detector i. It is necessary to estimate the number of photons x _ij from the sum y _i . Here, when the maximum posterior probability estimation (MAP estimation) is used, the MAP estimation value of the number of photons x _ij is given by Equation 10.

That is, when the RI density λ _j of the pixel j (5), which is an unknown number to be obtained by the maximum likelihood estimation method, is estimated from the detection probability c _ij and the photon number x _ij , the photon number x _ij cannot be directly measured. It is obtained from the estimated MAP value of x _ij . Therefore, it is possible by substituting MAP estimate of photon number _{x ij} of formula 10 to the number of photons _{x ij} of the right side of Equation 9 is obtained Equation 11.

Equation 11 cannot be solved for λ _j because there is also an RI concentration λ _{j on the} right side. Therefore, a sequential formula is created by the EM method. When λ _j ^k is an estimated value in the k-th successive approximation, Expression 12 can be obtained from Expression 11.

FIG. 6 shows a calculation example in which the RI concentration λ _j is estimated using Equation 12. As shown in FIG. 6, first, the number n of projection directions is set to 2, the matrix size of the tomographic image is set to 3 × 3 (m = 3 × 3 = 9), and each detection probability c _ij is set to 1 only in the projection direction. It is assumed that the direction is 0 (step S60). That is, it is assumed that measurement is not performed from the pixel j (5) other than the projection direction. That is, only c ₁₁ , c ₁₄ , and c ₁₇ are 1. As shown in step S62 of FIG. 6, when the true value is 1 to 9 from m = 1 to 9, the projection count number y _i in each projection direction is the sum of the values of the pixels j in the projection direction. Become. For example, in the direction of m = 1, 4 and 7, the value of the pixel j is 1, 4, and 7, respectively, so that the projection count number y _i = 1 + 4 + 7 = 12. The same applies to the other projection count numbers y _i , and thus the description thereof is omitted. As shown in step S62, the estimated value becomes λ ₁ to λ ₉ from m = 1 to 9. Then, all the lambda _j ⁰ of k = 0 assuming 1 to estimate calculated by the equation 12 (step S64). Details are as shown in step S66. For example, for λ ₁ ¹ , λ ₁ ⁰ = 1.0, and for the sum from c ₁₁ λ ₁ ⁰ to c ₁₉ λ ₉ ⁰ , only c ₁₁ , c ₁₄ , and c ₁₇ are 1 and the others are 0. Therefore, it becomes 3.0. Since y ₁ c ₁₁ = 12 × 1, y ₂ c ₂₁ = 6 × 1, and c ₁₁ + c ₂₁ = 2, λ ₁ ¹ = 3.0.

As shown in step S68 as the first estimated value, from m = 1 to 9, λ ₁ ¹ = 3.0, λ ₂ ¹ = 3.5, λ ₃ ¹ = 4.0, λ ₄ ¹ = 4 .5, λ ₅ ¹ = 5.0, λ ₆ ¹ = 5.5, λ ₇ ¹ = 6.0, λ ₈ ¹ = 6.5, and λ ₉ ¹ = 7.0. Next, λ _j ² is similarly obtained from Equation 12 using λ _j ¹ thus obtained (step S70). As shown in step S72 as the second estimated value, from m = 1 to 9, λ ₁ ² = 2.2, λ ₂ ² = 2.8, λ ₃ ² = 3.3, λ ₄ ² = 4 .3, λ ₅ ² = 5.0, λ ₆ ² = 5.8, λ ₇ ² = 6.4, λ ₈ ² = 7.3, and λ ₉ ² = 8.1. The above calculation is repeated as many times as desired (step S74). The end of repetition may be determined according to a desired convergence condition. As a result, a true value (or approximate value) can be obtained (step S76).

The image reconstruction step (step S20) in the cerebral blood flow quantitative analysis program of the present invention uses the above-described successive approximation image reconstruction method. 7 and 8 show a series of flowcharts of the image reconstruction step in the cerebral blood flow quantitative analysis program of the present invention. In this flowchart, the number h of divided doses is set to 2 for convenience of explanation. FIG. 7 is a flow chart (first reconstructed image generation step) for generating a reconstructed image after the first divided dose (h = 1) (first reconstructed image generating step), and FIG. 8 shows the processing from step S26 in FIG. That is, this is a flowchart (second reconstructed image generation step) of generating a reconstructed image after the second (h = 2) divided administration. As shown in FIG. 7, first, the total number m of pixels on the reconstructed image and the total number n of pixels on the projection data are read from the pixel number recording unit 84 (described later, see FIG. 37) (step S21). Pixel j on the reconstructed image is set to 1 (step S22), and the number of repetitions k described above is set to 0 (step S23). The estimated RI density λ _j ^{k [1]} = of the pixel j on the reconstructed image is read from the first (h = 1) estimated RI density initial value recording unit 86-1 (described later, see FIG. 37) (step S24). . For pixels i = 1 to n on the projection data, the scatter correction term S _i ^[1] is appropriately read out from the first scatter correction term recording unit 88-1 (described later, see FIG. 37), and the detection probability c _ij is determined as the detection probability. The photon count y _i ^[1] detected by the pixel i on the projection data is appropriately read out from the recording unit 85 (described later, see FIG. 37), and the first (h = 1) -th measurement data recording unit 87-1 ( It reads suitably from below-mentioned (refer FIG. 37), and calculates Formula 12 (step S25). However, considering the scattering correction term S _i ^[1] in Equation 12 and considering the number of divided doses h = 1, Equation 12 is as shown in Equation 1. In Formula 1, [1] means h = 1.

It is determined whether or not the number of repetitions has been repeated a predetermined number of times (step S26). If the number of repetitions has not been repeated, the number of repetitions k is incremented by 1, and λ _j ^{k [1]} obtained by Equation 1 is used as in step S25. Equation 1 is calculated for each new k. If it is determined in step S26 that the number of repetitions has been completed, the generation of the reconstructed image after the first (h = 1) divided administration is completed, and the process then proceeds to A.

As shown in FIG. 8, first, the total number m of pixels on the reconstructed image and the total number n of pixels on the projection data are read from the pixel number recording unit 84 (step S30). Since step S30 is similar to step S21, it may be omitted. Pixel j on the reconstructed image is set to 1 (step S31), and the number of repetitions k described above is set to 0 (step S32). The estimated RI density λ _j ^{k [2]} = of the pixel j on the reconstructed image is read out from the second (h = 2) estimated RI density initial value recording unit 86-2 (step S33). For pixels i = 1 to n on the projection data, the detection probability c _ij is appropriately read from the detection probability recording unit 85, and the scatter correction term S _i ^[2] is appropriately read from the second scatter correction term recording unit 88-2. The count number y _i ^[2] of photons detected at the pixel i on the projection data is appropriately read out from the second (h = 2) -th measurement data recording unit 87-2, and the following equation 2 is calculated (step S34). ).

  The characteristic of the image reconstruction step in the cerebral blood flow quantitative analysis program of the present invention uses data obtained by projecting the 1st reconstructed image created by the (k + 1) th successive approximation as the baseline radioactivity as shown in Equation 2. In the point. For this reason, the baseline radioactivity is updated every time. This method is called an IBL (iterative baseline estimation) method.

Next, it is determined whether or not the process has been repeated a predetermined number of times (step S35). If it has not been repeated, the number of repetitions k is incremented by 1, and the step is performed using λ _j ^{k [h]} obtained by Equation 2. Similarly to S34, Equation 2 is calculated for a new k. If it is determined in step S35 that the number of repetitions has been completed, the generation of the reconstructed image after the second (h = 2) divided administration ends.

  In the above IBL method, the following formula 3 can be used instead of the formula 1.

Hereinafter, Formula 3 will be described. The component of SPECT2 can be expressed as y _i ^[2] = P _i ^[1] + P _i ^{[2 true]} + S _i ^[2] . If the primary component (P _i ^[1] ) of residual radioactivity by administration of 1st contained in y _i ^[2] can be extracted and used in the 1st reconstruction, the S / N of the 1st reconstructed image It is thought that it contributes to improvement. Considering the data (y _i ^[1] + y _i ^[2] ) obtained by adding the 1st and 2nd projection data, this is 2P _i ^[1] + P _i ^{[2 true]} + S _i ^[1] + S _i ^{[2 ] ]} become. That is, when viewed from the 1st reconstruction, the net primary component (P _i ^{[2 true]} ) by the 2nd administration can be regarded as the baseline. Therefore, in this case, it can be incorporated into the OSEM reconstruction in the same manner. That is, the 1st reconstruction is obtained as shown in Equation 3 above. The terms λ _j ^{k + 1 [2]} and λ _j ^{k + 1 [1]} in the denominators of Equation 3 and Equation 2 can be obtained alternately from the k + 1th iteration of Equation 3 and the k + 1 iteration of Equation 2, respectively. Therefore, a reconstructed image can be obtained by substituting them for each repetition. This method is called an ABL (alternate baseline estimation) method.

Evaluation of convergence.
In order to evaluate the convergence of each reconstruction method, a cold region (5 cm in diameter) and a hot region (5 cm in diameter) having a low radioactivity concentration in a cylindrical phantom (18 cm in diameter) in which background (BG) radioactivity is present. ) Was assumed. The radioactivity concentration ratio of each region was assumed to be BG: Hot: Cold = 1: 4: 0.25, and the matrix size of the phantom was assumed to be 128 × 128 (pixel size: 3.125 mm). In order to focus only on the problem of convergence, it was assumed that there was no γ-ray absorption, scattering, and statistical noise, and 120-direction projection data was created by numerical integration. The dose ratio between 1st and 2nd was assumed to be 0.667: 1.333 in the subtraction (SUB) method described in the background art and 1: 1 in the other methods. Taking a circular ROI (diameter: 4.375 cm) in the background region, the Hot region, and the Cold region on the 1st and 2nd images obtained by each reconstruction method, the contrast recovery coefficient calculated by the following Equation 13 ( The change in contrast recovery coefficient (CRC) was calculated as a function of the number of iterations.

Here, BG represents the background region, S represents either the Hot region or the Cold region, (Contrast) _meas is the contrast calculated from the ROI, and (Contrast) _true is calculated from the assumed radioactivity concentration. Contrast. When it converges to the true value, it approaches CRC = 1.0. The number of repetitions was 1 to 150 times, and subset =.

Effect of optimal dose ratio and statistical noise.
Next, in each reconstruction method, the statistical noise was evaluated by Monte Carlo simulation when the dose ratio (2nd / 1st) was changed. The simulation method and parameter conditions are 8 types of dose ratios of 0.5, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, and the total dose (ie, The total count of 2nd projection data) was always constant (5.40 Mcounts / slice). As the numerical phantom, a cylindrical phantom (18 cmφ) was used, and a uniform absorber (μ = 0.15 cm ⁻¹ ) was assumed. In reconstruction, the number of iterations was set to 4 and subset = 20. Since the quantitative property of the SPECT image is impaired by the absorption (attenuation) and scattering of γ rays, it is necessary to correct the absorption (attenuation) and scattering to obtain a quantitative SPECT image. In Example 1, the absorption correction was performed by the Chang method using the assumed μ map, and the scattering component was estimated by using the Triple Energy Window (TEW) method. The% COV (= standard deviation / ROI value × 100%) within the ROI (5 cmφ) set at the center of the cylindrical phantom was calculated, and the optimal dose ratio at which the% COV was the same between 1st and 2nd was determined. For comparison, the SUB method and the MBL method were also simulated. The MBL method uses the following equation 14 instead of equation 2 for the 2nd reconstruction in the IBL method.

Evaluation of image quality and quantitativeness of reconstructed images.
A Monte Carlo simulation was similarly performed using a 3D digital brain phantom by Zubal et al. As the blood flow pattern, 1st was a resting brain blood flow pattern, 2nd was a pattern in which blood flow increase was observed only in the right hemisphere due to ACZ load, and this was combined to evaluate the image quality and quantitativeness of each reconstruction method. Data was created as follows. First, projection data (y _i ^[1] ) was created by Monte Carlo simulation assuming a gray matter: white matter radioactivity concentration ratio of resting brain blood flow pattern of 4: 1. Next, the ACZ load blood flow pattern is 1.4 times the radioactivity concentration of the gray matter in the right hemisphere of the resting blood flow pattern, and similarly, projection data (y _i ^{[2 true]} ) is created by simulation. did. The amount of baseline radioactivity component (y _i ^{[1 base]} ) contained in the 2nd projection data is essentially the same as the 1st projection data (y _i ^[1] ), but to make statistical noise independent. The simulation was performed using another random number. The 2nd projection data is created by adding these two projection data (y _i ^[2] = (y _i ^{[1 base]} + y _i ^{[2 true]} ). The dose ratio is 0.667 by the SUB method. : 1.333, 1: 1 in other methods, and the total count of 2nd projection data was constant (1.95Mcounts) to keep the dose constant in all methods, the preprocessing filter was a Butterworth filter (Fc = 0.26cycles / pixel, n = 8) Absorption correction was performed using the μ map created from a digital brain phantom and the Chang method, and ROI analysis was performed using a three-dimensional stereotaxic ROI template (3DSRT). The quantitative values in each brain segment were compared.

result.
Convergence characteristics and contrast.
FIGS. 9 and 10 are graphs showing convergence characteristics of the reconstruction methods in the Cold region and Hot region of the 1st image, respectively. FIGS. 11 and 12 are reconstructions in the Cold region and Hot region of the 2nd image, respectively. The graph which showed the convergence characteristic of the method is shown. As shown in FIG. 9, in the Cold region of the 1st image, the characteristics were almost the same in all the methods, and the CRC became a constant value when the number of repetitions was about 50-60. On the other hand, as shown in FIG. 10, the ABL method showed slightly different characteristics in the Hot region, but all methods reached the true value at about 50 repetitions. When evaluated with the 2nd image, as shown in FIG. 11, in the Cold region, the MBL method was slower in convergence than the other three methods, and converged at about 90 iterations. The other three methods have almost the same characteristics, and converged in almost the same manner as the 1st image. As shown in FIG. 12, the Hot region showed different characteristics in each region. The IBL method smoothly asymptotically converged to a true value according to repetition. On the other hand, in the ABL method, a large overshoot was made at a high number of repetitions, and then approached the true value. In the SUB method and the MBL method, a slightly overshooting tendency was observed. The IBL method reached the true value with the same number of repetitions (about 30 times) as the SUB method. The MBL method was the slowest convergence characteristic. Overall, the Cold region converged slower than the Hot region in any method.

Evaluation of optimal dose ratio and statistical noise.
FIGS. 13 to 16 are graphs showing changes in% COV with respect to the dose ratio (2nd / 1st) in the MBL method, SUB method, IBL method and ABL method, respectively. In FIGS. 13 to 16, the horizontal axis represents the injection dose ratio, and the vertical axis represents% COV. As shown in FIGS. 13 and 14, in the MBL method and the SUB method, when the dose ratio is 1.0 and 2.0, the 1st and 2nd% COVs are almost the same, and about 53% and about 73 respectively. %Met. On the other hand, as shown in FIG. 15 and FIG. 13, the IBL method had a slightly smaller 2nd% COV than the MBL method, and as a result, the optimum dose ratio shifted to a slightly smaller value of about 0.85. The% COV of the IBL method at this time was about 50%, which was slightly better than the MBL method. As shown in FIG. 16, the ABL method has a 1%% COV that is much better than that of the 2nd, and the dose ratio at which the% COV is the same for the 1st and 2nd was not found within the scope of this simulation. As shown in FIGS. 15 and 16, the 2nd reconstruction formulas of the IBL method and the ABL method are exactly the same, but the ABL method has better% COV under any conditions.

Evaluation of image quality and quantitative values.
FIG. 17 shows 1st (REST, rest), 2nd (ACZ, load) images of the brain digital phantom obtained by the above reconstruction methods. For comparison, a true blood flow increase image (True ACZ) obtained by directly reconstructing y _i ^{[2nd true]} is also shown. As shown in FIG. 17, from the viewpoint of image quality, the ABL method was the best and the SUB method was the worst for the 1st image. Since the MBL method and the IBL method are exactly the same calculation formula, the result images are the same. On the other hand, in the 2nd image, the ABL method is still the best, and the IBL method and the MBL method are almost the same. The SUB method was the worst. These agreed with the results from the pool phantom (FIG. 14). In the MBL method, a slight decrease in contrast is observed in the central low count region as compared with the IBL method. In the ABL method, the 1st image quality was slightly better than the 2nd image quality. As described above, in the case of using twice (or two or more) divided administration methods, brain image reconstruction can be performed without performing subtraction.

  Next, the anatomical normalization step (step S40) in the cerebral blood flow quantitative analysis program of the present invention will be described. FIG. 18 is a flowchart showing details of the anatomical normalization step (step S40) in the cerebral blood flow quantitative analysis program of the present invention. As shown in FIG. 18, for the reconstructed images after each divided administration, the reconstructed images after the second divided administration (SPECT2 and later) are the same as the reconstructed images after the first divided administration (SPECT1). (Alignment step, step S42) Next, the reconstructed image after the correction in the alignment step is deformed so as to match the standard brain using the deformation parameters of the space. A brain image is obtained (standardization step, step S44), where the deformation parameter is the same parameter as the deformation parameter in SPM99, SPM2, etc. However, in the first embodiment, each of the projection data is given a predetermined weight. After obtaining total projection data that is the sum of the projection data and obtaining a reconstructed image based on this total projection data, A standard brain image is obtained by performing anatomical normalization on the component image, and a deformation parameter is obtained based on this processing.Examples of predetermined weighting are 1 for SPECT1 data and 1 for SPECT2 data. In this case, the total projection data is SPECT1 + SPECT2 data × (1/2) + SPECT3 data × (1/3). The deformation parameters calculated as described above are used as deformation parameters for anatomical normalization of the first, second, and third reconstructed images, and then the reconstructed image after the standardization step (step S44). Is applied with a smoothing filter (smoothing step, step S46).

  The above-described anatomical normalization step (step S40) can be executed by using, for example, free software SPM (statistical parametric mapping) 99 or SPM2. Alternatively, 3D-SSP (three-dimensional stereotactic surface projections) may be used.

Next, the quantification step (step S50) in the cerebral blood flow quantitative analysis program of the present invention will be described. SPECT images include qualitative images and quantitative images. In the qualitative image, the pixel value indicates the count value that is actually counted. Therefore, if there are many drugs or the like, the pixel value increases accordingly. In addition, the longer the time the target is imaged, the larger the pixel value accordingly. Therefore, the pixel value itself has little meaning. On the other hand, in the quantitative image, the pixel value indicates a value obtained by converting the actually counted count value into a desired unit, for example, a flow rate (ml / 100 g / min). Therefore, the pixel value has a meaning as a flow rate, for example. The quantification (absolute ^{assay), 123 I-IMP (N} -isopropyl-p- [123 I] iodoamphetamine. Hereinafter abbreviated as "IMP".) And ECD microsphere method used as a tracer, the IMP as a tracer There are an autoradiography method used, a patlak plot method using an ECD developed from the microsphere method as a tracer, and a BUR using the ECD as a tracer. In the quantification of the first embodiment, the case where the patlak plot method is used will be described, but it is needless to say that other methods described above may be used.

  FIG. 19 shows the timing for performing the patlak plot method. 19 is an enlarged view of a part of FIG. 2 (from time 0 to t4), so the description of the same parts as in FIG. 2 is omitted. As shown in FIG. 19, when the ECD is intravenously injected at time 0, the ECD accumulates in the aortic arch of the heart and the brain along the blood flow. In the patlak plot method, ECD accumulated in the aortic arch and the brain is imaged by SPECT. Although it takes about 2 minutes in terms of time, it ends before the collection of SPECT1 (at rest) data is started. ROI is set for the aortic arch and the brain (the cerebral hemisphere on the healthy side).

  FIGS. 20A and 20B show time-activity curves (TAC) of ECD accumulated in the aortic arch and brain. 20 (A) and (B), the horizontal axis is time (t), the vertical axis is the amount of radioactivity, the TAC of the aortic arch (Aorta) is A (t), and the TAC of the brain (Brain) is This is indicated by B (t). Since ECD flows from the aortic arch to the brain, the peak position Pa of A (t) and the peak position Pb of B (t) are deviated as shown in FIG. Therefore, as shown in FIG. 20B, B (t) is moved in parallel to align these peak positions Pa and Pb.

  Next, a patlak plot as shown in FIG. 20C is obtained from the TAC shown in FIG. Specifically, the horizontal axis represents the following equation:

And the vertical axis is plotted as shown in FIG. 20C with B (t) / A (t). As a result, the obtained straight line is expressed by Equation 15 below.

  Here, the slope Ku is a velocity inflow constant (parameter corresponding to cerebral blood flow), and the intercept Vn is an initial distribution volume (parameter corresponding to cerebral blood volume). Next, BPI (Brain perfusion index) is calculated by the following equation 16 using the slope Ku.

  Subsequently, the cerebral hemisphere average CBF is calculated from a conversion formula between BPI (ECD-BPI) and average cerebral blood flow (mCBF) obtained in advance. Examples of the conversion formula include formulas 17 to 19.

¹³³ Xe-mCBF = 2.59 × ECD-BPI + 19.8 (17)
¹³³ Xe-mCBF = 3.34 × ECD-BPI + 19.7 (18)
¹²³ IMP-mCBF = 3.69 × ECD-BPI + 4.61 (19)

  Equation 17 is the EP method (early picture method), Equation 18 is the SP method (sequential picture method), Equation 19 is the continuous artery voting method, and any conversion equation may be used as long as it is constant for each facility. In the quantification step (step S50) in the cerebral blood flow quantification analysis program of the present invention, the input average cerebral blood flow value (mCBF) of the subject at rest is used for each standard brain image as the predetermined quantification. Each cerebral blood flow is obtained by correcting Lassen mentioned in the background art to obtain each quantitative image. Specifically, the Lassen correction is performed using the following Lassen correction formula 20.

  In the case of the 1st cross-sectional image (at rest), Fr is the above-mentioned mCBF (average cerebral blood flow value at rest), and C ′ is set by the value of the 1st cross-sectional image of region i / 3DSRT ROI template (described later) It is the average value of the values of the 1st cross-sectional image of the ROI part. On the other hand, in the case of a 2sub cross-sectional image (1 when loaded), Fr is the above-described mCBF × basal ganglia count during loading / basal ganglia count during rest. Here, the loaded basal ganglia count is the value of the slice image (2nd) automatically selected as Range 31-40, and the resting basal ganglia count is the slice image (Range 31-40 automatically selected). 1st). C ′ is the average value of the 2sub cross-sectional image value of the region i / the 2sub value of the ROI portion set by the 3DSRT ROI template.

  FIG. 21 shows the above-described ROI template. In the cerebral blood flow quantitative analysis program of the present invention, a standard ROI as shown in FIG. 21 is prepared. When setting an ROI for the normal cerebral hemisphere or the milder cerebral hemisphere of the slice-added image, this ROI template can be used. For this reason, it is possible to avoid variation among operators in setting the ROI, to prevent variation in the quantitative image, and to eliminate the complexity of the operator.

  As described above, in the cerebral blood flow quantitative analysis program of the present invention, Lassen correction is performed on the SPECT image after anatomical standardization using the average cerebral blood flow value (mCBF) at the time of input of the subject. Each quantitative image is obtained by determining the local cerebral blood flow. This analysis method is examined as follows, and as a result, as described above, the loading basal ganglia count and resting basal ganglia count are automatically set to the values of the slice image selected as Range 31-40. I was able to.

Equipment used and collection conditions.
The gamma camera used for imaging was a two-detector-facing gamma camera GCA7200A / PI (manufactured by Toshiba) equipped with a low energy general-purpose collimator (LEGP), and the image processing apparatus used was GMS5500PI. Dynamic data was collected at 128 × 128 matrix, 1 second per frame. The SPECT was 64 × 64 matrix, the pixel size was 4.3 mm, and the collection time was 2 minutes 30 seconds / 1 rotation with continuous repeated 4 rotation collection (10 minutes). Scattered ray correction is not performed. Image reconstruction was performed by a filter back projection method using a ramp filter, Butterworth (order 8, cutoff frequency 0.50 cycles / cm) was used as a preprocessing filter, and Chang (attenuation coefficient 0.09 cm ⁻¹ ) was used as attenuation correction. Using the input function obtained from the dynamic data, the average cerebral blood flow value on the normal side was calculated by the BUR method.

Target case.
Basal ganglia slice setting: A 65-year-old woman diagnosed with left cerebral infarction visually observed a left-right difference and rested on all 39 slices calculated with a mean cerebral blood flow value of 45 ml / 100 g / min on the right side by the BUR method The cerebral blood flow SPECT image (FIG. 22) was used. A comparison between a conventional manual quantitative analysis (conventional method) and a quantitative analysis method (3DSRT method) corrected by Lassen after normalization was performed using data of 11 clinical cases.

Methods and items to consider.
Qualitative SPECT images were anatomically normalized with 3DSRT using SPM99 (FIG. 23). The basal ganglia slice was selected for 59 SPECT images after anatomical normalization (although there are 68 images in actuality but 59 images with 3DSRT displayed). As the selection of the basal ganglia slice, the upper limit is changed to 11, 16, 21, 26, 31 slices so as to exclude the most parietal portion and the cerebellum which are generally omitted, and the lower limit is 20, 21, 31, 36, 40, We examined 19 combinations of 47 combinations. As the reference ROI setting on the normal side after anatomical normalization, the ROI of 3DSRT was used as it was, and the average count included in the selected slice was obtained. Since 3DSRT is displayed as an average count, the average count was calculated again after correction using each pixel number. In order to compare the conventional method and the 3DSRT method for data of 11 clinical cases, 3DSRT analysis was performed on each quantitative SPECT image, and local cerebral blood flow values in 12 regions on both sides were calculated. Both measured values were analyzed by Bland-Altman (the vertical axis represents the difference between the conventional method and 3DSRT method, and the horizontal axis represents the average value).

result.
3DSRT was performed on the qualitative SPECT image, and the difference in selection of basal ganglia slices was compared with the anatomically normalized qualitative SPECT image (FIG. 23). As a result, as shown in FIG. 24, the whole blood flow showed a high value when a slice at the top of the head was included, and a difference in local quantitative value due to a difference in selection was confirmed. In FIG. 24, the horizontal axis indicates the selection of the basal ganglia slice (upper limit slice-lower limit slice), and the vertical axis indicates the average count. As shown in FIG. 24, the comparison was performed in 24 regions from Anterior (R) to Cerebellum (L). As shown in FIG. 24, the cerebral blood flow value obtained by the conventional method was the closest to the cerebral blood flow value obtained by the 3DSRT method, as shown in FIG. Approximate to 3DSRT values created with 19 to 22 slices (4.3 mm x 4 = 17.2 mm), which were considered to be optimal for visual analysis of basal ganglia slices by SPECT before anatomical normalization, and also visually 31-40 slices (2 mm × 10 = 20 mm) after anatomical normalization close to the selection of the basal ganglia were defined as the basal ganglia and compared in 11 clinical cases. As a result of Bland-Altman analysis of 24 regions of both the conventional method and the 3DSRT method, 9 cases out of 11 cases were included in the average value and 2SD. The remaining two cases (Case 10.11) were slightly different.

Discussion.
At present, the poor reproducibility of the cerebral blood flow analysis results is pointed out, and it is considered that this is caused by the choice of the basal ganglia slice and the normal ROI setting by the operator. However, in the 3DSRT method described above, the RODS setting of 3DSRT itself is not set to a part that lowers the average count such as brain quality, so there is a difference depending on the selection, but the difference is not so large. The criteria for selection of basal ganglia slices need to be examined again, but by performing anatomical normalization, it is possible to always select the same slice regardless of the case or the operator, and the ROI on the normal side Since a 3DSRT ROI value was used for setting, the created quantitative SPECT image was reproducible. Of the 11 clinical cases, two cases that differed from the conventional method may be approximated to the 3DSRT method by performing normal ROI setting and slice selection again using the conventional method.

  As described above, the method of performing anatomical normalization on the qualitative SPECT image and performing Lassen correction on the image after the anatomical normalization is performed by selecting the basal ganglia slice and ROI setting of the operator. This method is excluded, and is considered to be a method with very high reproducibility in local cerebral blood flow quantification using 3DSRT. It is also consistent with local cerebral blood flow created by conventional methods, and is a new quantitative analysis method that improves reproducibility without changing the current clinical evaluation.

  The cerebral blood flow quantitative analysis program of the present invention may further include a display step for displaying each quantitative image obtained in the quantification step (step S50). FIG. 25 shows an example of the display result by the display function in the cerebral blood flow quantitative analysis program of the present invention. In FIG. 25, reference numeral 40 denotes a file selection field for selecting projection data, file 00REST is selected as REST (resting time), and file 00ACZ_7_5 is selected as ACZ1 (loading time 1). The file 00ACZ_20 is selected as ACZ2 (load 2). FIG. 25 shows the case of using the three-part administration method, but when using the two-part administration method, only REST (at rest) and ACZ1 (at the time of loading) should be selected. Reference numeral 42 denotes a Butterworth filter parameter column in the reconstruction. The dose, the collection time correction coefficient (Weight), the cut-off (Cutoff) which is the coefficient of the Butterworth filter, and the order are described above. Can be specified for each file. Reference numeral 44 denotes an acquisition parameter column, where 360 CCW is selected as the rotation, 64 × 64 is selected as the size of the matrix, and 180 degrees is selected as the start angle. A state in which 4.30 (mm) is selected as the pixel size and 60 is selected as the steps (Steps) is shown. Reference numeral 46 denotes an attenuation correction (Attenuation) column, where “Chang” is selected, 0.09 (fixed) is selected as the coefficient (Atten.), And 20% is selected as the threshold (Theshold). Reference numeral 48 denotes a Lassen correction field. Lassen correction is selected (On), the left side (left) is selected as the normal side (Normal) of the cerebral hemisphere, and 40 is selected as the mean blood flow value (mCBF) at rest. This shows a state where Range 23-32 is selected as the basal ganglia slice. When Lassen correction is selected (On), a quantitative image is created, and when Lassen correction is not selected (Off), a quantitative image is not created and a qualitative image is created. Reference numeral 50 is a SPECT data display field (Projection) in a state where reconstruction is not performed, and SPECT data captured at an arbitrary time point can be displayed by moving the right or lower bar (Frame). Reference numeral 52 denotes a patient information column, which can display or input the hospital name, patient name, patient ID (identifier), and patient gender. Reference numeral 54 denotes a scatter correction column, which indicates a state where the TEW method is not used (Off) for estimating the scatter component. Reference numeral 56 denotes an OSEM parameter column, which is 40 times as the number of iterations, 1 as the number of subsets, and 7 as an interval between rest (REST) and load 1 (ACZ1). .5 (minutes), 20 (minutes) is selected as the interval (Interval) between loaded time 1 (ACZ1) and loaded time 2 (ACZ2). Reference numeral 58 denotes a Go button, which is clicked to start processing of the cerebral blood flow quantitative analysis program of the present invention.

  FIG. 26 shows a display example in a state where analysis is started by clicking the Go button 58. A part of the various parameters selected in FIG. 25 is displayed in the left window of FIG. 26, and a projection data input step (step S10) and an image reconstruction step (step S20) proceed in the right window. The message that is in is displayed.

  FIG. 27 shows a display example in a state where the anatomical normalization step (step S40) by SPM99 or SPM2 is being executed. As shown in FIG. 27, when the image reconstruction step (step S20) is completed, SPM99 or SPM2 automatically rises and anatomical normalization is executed.

  The cerebral blood flow quantification analysis program of the present invention can further include a display step for displaying each quantitative image obtained in the quantification step (step S50). FIG. 28 shows a display example of display steps when Lassen correction (On) is selected in the Lassen correction field 48, that is, when creation of a quantitative image is selected. FIG. 29 shows a display example of the display step when Lassen correction (Off) is selected in the Lassen correction field 48, that is, when creation of a qualitative image is selected. FIG. 30 shows an enlarged example of the qualitative image at rest in FIG.

  In the cerebral blood flow quantitative analysis program of the present invention, the increase amount and / or the increase rate of the local cerebral blood flow in a predetermined region are displayed in a predetermined display format based on each quantitative image obtained in the quantification step (step S50). An increase amount display step for displaying can be further provided. As a predetermined display format, the local cerebral blood flow value for each divided administration can be displayed in a graph. FIG. 31 shows a method of selecting data at rest, load 1 and load 2 when performing the graphing. Specifically, as shown in FIG. 31, copy the data of load 1 from the 70th line of the right file 3DSRT_ntCBFtcACZ1.csv and paste it to the 4th line of the left quantitative graph.xls. Copy the resting data from the 70th line of 3DSRT_ntCBFtcREST.csv and paste it into the 5th line of the left quantitative graph.xls. Paste to the 6th line of the left quantitative graph.xls. Thereafter, by clicking the arrow 60, the increase amount and / or the increase rate of the local cerebral blood flow in the predetermined region can be displayed in a graph format.

  FIG. 32 shows a graph of 12 regions displayed in the increase amount display step. FIGS. 33 (A) to 33 (L) are enlarged examples of the 12-region graph shown in FIG. 32. At rest (REST), load 1 after 7.5 minutes, and load after 20 minutes. For time 2, the amount of increase is displayed in a graph. FIG. 33A shows an example of an increase amount display at the corpus callosum edge displayed in the increase amount display step, and FIG. 33B shows an increase amount display in front of the center displayed in the increase amount display step. FIG. 33 (C) shows an example of an increase amount display at the center displayed at the increase amount display step, and FIG. 33 (D) shows an increase display at the top displayed at the increase amount display step. FIG. 33 (E) shows an example of the increase amount display at the angular rotation displayed in the increase amount display step, and FIG. 33 (F) shows the increase amount in the temporal region displayed in the increase amount display step. An example of display is shown, FIG. 33 (G) shows an example of an increase amount display in the cerebrum after being displayed in the increase amount display step, and FIG. 33 (H) is an area around the corpus callosum displayed in the increase amount display step. An example of an increase amount display in FIG. An example of an increase amount display in the lens nucleus displayed in the addition amount display step is shown, and FIG. 33 (J) shows an example of an increase amount display in the thalamus displayed in the increase amount display step, and FIG. Shows an example of an increase amount display in the hippocampus displayed in the increase amount display step, and FIG. 33 (L) shows an example of an increase amount display in the cerebellum hemisphere displayed in the increase amount display step.

  As the above-mentioned predetermined display format, a difference for each pixel between the local cerebral blood flow value in the h-th (h is 2 or more) divided administration and the local cerebral blood flow value in the divided administration before the h-th is given. It can also be displayed in an identifiable manner using color. 34A shows a quantitative image at rest, FIG. 34B shows a quantitative image at load, and FIG. 34C shows the local brain at rest from the local cerebral blood flow value at load. The image of the increase amount at the time of subtracting a blood flow value for every pixel is shown. As the predetermined color described above, as shown in the scale on the right side of FIG. 34C, pink, red, orange, green,. . . It can be blue.

  The cerebral blood flow quantitative analysis program of the present invention is a slice constituting the projection data after the first divided administration input in the projection data input step (step S10) prior to the image reconstruction step (step S20). A confirmation display step of displaying a reconstructed image by applying a predetermined filter to the projection data of the minute can be further provided. FIG. 35 shows another example of the display result by the display function in the cerebral blood flow quantitative analysis program of the present invention. 35, the same reference numerals as those in FIG. 25 indicate the same elements, and the description thereof is omitted. In FIG. 35, reference numeral 57 denotes a reconstructed image displayed in the confirmation display step. As shown in FIG. 35, when the Ref button 59 is clicked, the projection data can be filtered and displayed in the preview field 57. A smoothing filter is suitable as the filter.

  As the above-mentioned predetermined display format, an average value of local cerebral blood flow values in a predetermined region of interest in h-th (h is 2 or more) divided administration and the predetermined ROI (interest in divided administration prior to the h-th administration). It is also possible to display the difference from the average value of the local cerebral blood flow value in the region using a predetermined color so as to be identifiable. The ROI template described above can be used for setting the ROI. 36A shows a quantitative image at rest, FIG. 36B shows a quantitative image at the time of loading, and FIG. 36C shows the above from the average value of local cerebral blood flow values in the ROI at the time of loading. The increase amount at the time of subtracting the average value of the local cerebral blood flow value in the same ROI at the time of rest is shown. FIG. 36D shows the rate of increase from the average value of local cerebral blood flow values in the ROI at rest to the average value of local cerebral blood flow values in the same ROI at the time of loading. As the predetermined color described above, as shown in the scales on the right side of FIGS. 36 (C) and 36 (D), pink, red, orange, green, . . . It can be blue.

  The cerebral blood flow quantitative analysis program of the present invention uses a predetermined color for the average value of local cerebral blood flow values in a predetermined ROI (region of interest) based on each quantitative image obtained in the quantification step (step S50). Can be displayed in an identifiable manner. The ROI template described above can be used as the ROI. FIG. 37 shows an average ROI count image when the ROI template is used. As shown in FIG. 37, the average value of local cerebral blood flow values in units of ROI can be displayed. As the predetermined color, as shown in the scale on the right side of FIG. 37, pink, red, orange, green,. . . It can be.

  As described above, according to Example 1 of the present invention, the RI-labeled tracer is divided into two or more doses to the subject's brain, and each projection data (SPECT measurement obtained after each divided dose) ( SPECT data). Next, based on each input projection data, a reconstructed image after each divided administration is obtained, and each reconstructed image after each divided administration is subjected to anatomical normalization to obtain each standard brain image . Each quantitative image can be obtained by performing predetermined quantification on each standard brain image obtained. When the reconstructed image is obtained, the brain image can be reconstructed without performing subtraction by using the IBL method or the ABL method and using the divided administration method twice (or two times or more).

  In the cerebral blood flow quantitative analysis program of the present invention, local correction is performed by performing Lassen correction on the SPECT image after anatomical normalization using the inputted mean cerebral blood flow value (mCBF) of the subject at rest. Each quantitative image is obtained by determining cerebral blood flow. For this reason, the basal ganglia count at the time of loading and the basal ganglia count at the time of rest could be automatically set to the value of the slice image selected as Range 31-40. In the cerebral blood flow quantitative analysis program of the present invention, when setting an ROI in the normal cerebral hemisphere or the milder cerebral hemisphere of the slice-added image, the ROI template can be used. For this reason, it is possible to avoid variation among operators in setting the ROI, to prevent variation in the quantitative image, and to eliminate the complexity of the operator.

  The display step (function or means) described above can be executed by using an image viewer such as MRICRO or OSIRIS, which is free software. However, there is a possibility that a predetermined ROI cannot be set.

  The cerebral blood flow quantitative analysis program of the present invention can be easily applied to projection data (for example, resting data) when divided administration is not performed. In this case, the cerebral blood flow quantitative analysis program according to the present invention includes a projection data input step of inputting a projection data obtained by performing computer tomography measurement after administration of a predetermined drug to the subject's brain, Based on the projection data input in the projection data input step, an image reconstruction step for obtaining a reconstructed image after administration, and an anatomical normalization for the reconstructed image after administration obtained in the image reconstruction step An anatomical normalization step for obtaining a standard brain image and performing a predetermined quantification step on the standard brain image obtained in the anatomical normalization step to obtain a quantitative image. it can. For example, instead of the projection data (SPECT1, SPECT2, and SPECT3) for three times for one subject, the resting data for three people in an examination that does not perform divided administration can be processed once.

  FIG. 38 is a block diagram showing an internal circuit 70 of a computer that executes the computer program of the present invention. As shown in FIG. 38, the CPU 71, ROM 72, RAM 73, image control unit 76, controller 77, input control unit 79, and external interface (Interface: I / F) unit 81 are connected to a bus 82. 38, the above-described computer program of the present invention is recorded on a recording medium (including a removable recording medium) such as a ROM 72, a disk 78a, or a CD-ROM 78n. This computer program is loaded into the RAM 73 from the ROM 72 via the bus 82 or from a recording medium such as the disk 78a or CD-ROM 78n via the controller 77 via the bus 82. The image control unit 76 sends brain image data to the VRAM 75, and the display unit 74 is a display device such as a display that displays a brain image based on the brain image data sent from the VRAM 75. The VRAM 75 is an image memory having a capacity corresponding to the data capacity of one screen of the display unit 74. The input operation unit 80 is an input device such as a mouse or a numeric keypad for inputting to a computer. The input control unit 79 is connected to the input operation unit 80 and performs input control. The external I / F unit 81 has an interface function when connecting to an external communication network (not shown) such as the Internet. The disk 78a includes a pixel number recording unit 84, a detection probability recording unit 85, an h-th estimated RI concentration initial value recording unit 86-h, an h-th measurement data recording unit 87-h, and an h-th scattering correction term recording unit. 88-h etc. can be recorded.

  As described above, the CPU 71 executes the computer program of the present invention to achieve the object of the present invention. The computer program can be supplied to the computer CPU 71 in the form of a recording medium such as a CD-ROM 78n as described above, and a recording medium such as a CD-ROM 78n recording the computer program also constitutes the present invention. It will be. As the recording medium on which the computer program is recorded, for example, a memory card, a memory stick, a DVD, an optical disk, an FD, or the like can be used in addition to the recording medium described above.

  As an application example of the present invention, the present invention can be applied to detection of diffuse cerebral blood flow reduction, follow-up observation (before and after treatment), objective comparison between different patients, comparison between rest and load, and the like.

It is a flowchart which shows flows, such as a cerebral blood flow quantitative analysis program in Example 1 of invention. It is a figure for demonstrating the projection data input step (step S10) in Example 1 of this invention. It is a figure explaining application to the image reconstruction of the maximum likelihood estimation method. It is a figure explaining application to the image reconstruction of the maximum likelihood estimation method in case RI is measured with the detector of n direction. It is a figure explaining application to the image reconstruction of the maximum likelihood estimation method in the case of considering the RI around the pixel j (5). FIG. 10 is a diagram illustrating a calculation example for estimating an RI concentration λ _j using Equation 12. It is a flowchart which shows a series of flows of the image reconstruction step in the cerebral blood flow quantitative analysis program of this invention. It is a flowchart which shows a series of flows of the image reconstruction step in the cerebral blood flow quantitative analysis program of this invention. It is the graph which showed the convergence characteristic of each reconstruction method in the Cold area | region of a 1st image. It is the graph which showed the convergence characteristic of each reconstruction method in the Hot area | region of 1st image. It is the graph which showed the convergence characteristic of each reconstruction method in the Cold area | region of a 2nd image. It is the graph which showed the convergence characteristic of each reconstruction method in the Hot area | region of a 2nd image. It is a graph which shows the change of% COV with respect to dosage ratio (2nd / 1st) in MBL method. It is a graph which shows the change of% COV with respect to dosage ratio (2nd / 1st) in the SUB method. It is a graph which shows the change of% COV with respect to dosage ratio (2nd / 1st) in IBL method. It is a graph which shows the change of% COV with respect to dosage ratio (2nd / 1st) in IBL method. It is a figure which shows the 1st (REST. At rest), 2nd (ACZ. At load) image of the brain digital phantom obtained by each above-mentioned reconstruction method. It is a flowchart which shows the detail of the anatomical normalization step (step S40) in the cerebral blood flow quantitative analysis program of this invention. It is a figure which shows the timing which performs a patlak plot method. It is a figure which shows the time activity curve TAC of ECD integrated | stacked on the aortic arch part and the brain. It is a figure which shows TAC at the time of combining peak position Pa and Pb. It is a figure which shows a patlak plot. It is a figure which shows a ROI template. It is a rest cerebral blood flow SPECT image figure of all 39 slices calculated with the average cerebral blood flow value of 45 ml / 100g / min on the right side by the BUR method. It is a figure which shows the SPECT data after anatomical normalization. It is the figure which compared the difference in selection of a basal ganglia slice. It is a figure which shows the example of the display result by the display function in the cerebral blood flow quantitative analysis program of this invention. It is a figure which shows the example of a display of the state by which Go button 58 was clicked and the analysis was started. It is a figure which shows the example of a display of the state in which the anatomical normalization step (step S40) by PM99 or SPM2 is performed. It is a figure which shows the example of a display of a display step when Lassen correction | amendment (On) is selected in the Lassen correction column 48, ie, when creation of a quantitative image is selected. It is a figure which shows the example of a display step when Lassen correction (Off) is selected in the Lassen correction column 48, that is, when creation of a qualitative image is selected. It is a figure which shows the enlarged example of the qualitative image at the time of rest in FIG. It is a figure which shows the selection method of each data at the time of rest, the time of load 1, and the time of load 2 at the time of performing the said graph. It is a figure which shows the graph of 12 area | regions displayed by the increase amount display step. It is a figure which shows the example of the increase amount display in the corpus callosum edge displayed by the increase amount display step. It is a figure which shows the example of the increase amount display in front of the center displayed by the increase amount display step. It is a figure which shows the example of the increase amount display in the center displayed by the increase amount display step. It is a figure which shows the example of the increase amount display in the top of the head displayed by the increase amount display step. It is a figure which shows the example of the increase amount display in the corner rotation displayed by the increase amount display step. It is a figure which shows the example of the increase amount display in the temporal region displayed by the increase amount display step. It is a figure which shows the example of the increase amount display in the postcerebrum displayed by the increase amount display step. It is a figure which shows the example of the increase amount display around the corpus callosum displayed at the display step of increase amount etc. It is a figure which shows the example of the increase amount display in the lens nucleus displayed by the increase amount display step. It is a figure which shows the example of the increase amount display in the thalamus displayed by the increase amount display step. It is a figure which shows the example of the increase amount display in the hippocampus displayed by the increase amount display step. It is a figure which shows the example of the increase amount display in the cerebellar hemisphere displayed by the increase amount display step. It is a figure which shows the fixed_quantity | quantitative_assay image at the time of rest. It is a figure which shows the fixed quantity image at the time of load. It is a figure which shows the image of the increase amount at the time of subtracting the local cerebral blood flow value in the said resting for every pixel from the local cerebral blood flow value in the said load. It is a figure which shows the other example of the display result by the display function in the cerebral blood flow quantitative analysis program of invention. It is a figure which shows the fixed_quantity | quantitative_assay image at the time of rest. It is a figure which shows the fixed quantity image at the time of load. It is a figure which shows the increase amount at the time of subtracting the average value of the local cerebral blood flow value in the same ROI at the time of the said rest from the average value of the local cerebral blood flow value in the ROI at the time of the said load. It is a figure which shows the increase rate from the average value of the local cerebral blood flow value in the ROI at the time of the said rest to the average value of the local cerebral blood flow value in the same ROI at the time of the said load. It is a figure which shows the average ROI count image in the case of using a ROI template. It is a block diagram which shows the internal circuit 210 of the computer which performs the computer program of this invention. It is a figure which shows the protocol of a 2 times divided dose method. It is a figure which shows the concept of a subtraction method. It is a figure for demonstrating selection of the slice containing a basal nucleus. It is a figure which shows the image by which the slice addition was carried out. It is a figure which shows the image by which the slice addition was carried out.

Explanation of symbols

  3, 5 pixels, 10 tomographic images, 20 projection data, 30, 31, 32, 33 detector, 40 pixel number recording section, 42 detection probability recording section, 40 file selection field, 42 filter parameter field, 44 acquisition parameter field, 46 attenuation correction column, 48 Lassen correction column, 50 SPECT data display column, 52 patient information column, 54 scatter correction column, 56 OSEM parameter column, 57 preview column, 58 Go button, 59 Ref button, 60 arrow, 70 internal circuit 71 CPU, 72 ROM, 73 RAM, 75 VRAM, 76 Image control unit, 78a disk, 78n CD-ROM, 77 controller, 80 input operation unit, 79 input control unit, 81 External I / F unit, 82 bus, 84 Pixel number recording section, 5 Detection probability recording unit, 86-h h-th estimated RI concentration initial value recording unit, 87-h h-th measurement data recording unit, 88-h h-th scattering correction term recording unit, 102, 106 scattering component, 104 , 108 Primary component.

Claims (32)

A cerebral blood flow quantitative analysis program for quantitative analysis of cerebral blood flow by computer tomography measurement,
A projection data input step in which a predetermined drug is divided and administered to the brain of a subject n times (n is 2 or more), and each projection data obtained by performing the computed tomography measurement after each divided administration is input;
An image reconstruction step for obtaining a reconstructed image after each divided administration based on each projection data input in the projection data input step, wherein n− created by successive approximation in the successive approximation type image reconstruction method The data obtained by projecting the first reconstructed image is used as a baseline to obtain the nth reconstructed image, and the baseline is updated each time.
An anatomical normalization step for obtaining each standard brain image by performing anatomical normalization on the reconstructed image after each divided administration obtained in the image reconstruction step,
A cerebral blood flow quantification analysis program for executing a quantification step in which each quantification image is obtained by performing predetermined quantification on each standard brain image obtained in the anatomical normalization step. In the cerebral blood flow quantitative analysis program according to claim 1,
The computed tomography measurement is a single photon emission computed tomography measurement or a positron emission tomography measurement, and the predetermined agent comprises a tracer labeled with a radioisotope (RI);
The image reconstruction step includes:
Projection in which m is the total number of pixels on the reconstructed image, j (= 1 to m) is a pixel on the reconstructed image, and n is a photon or positron emitted from a radioisotope administered to the brain The total number of pixels on the data or the total number of detectors that detect the photons or positrons, and i (= 1 to n) are the pixels on the projection data, where m and n are recorded in the pixel number recording unit in advance. And
and the probability of photon or positron emission the c _ij from pixel j in the reconstructed image is detected on a pixel i on the projection data, wherein c _ij is pre j = 1 from to m and i = 1 from the n Is recorded in the detection probability recording part,
From h = 1, for the number L of divided doses (L is 2 or more), λ _j ^{0 [h]} is used to estimate pixel j on the reconstructed image in the 0th successive approximation used for the hth divided dose. RI concentration, where λ _j ^{0 [h]} is previously recorded in the h-th estimated RI concentration initial value recording unit from j = 1 to m,
For h = 1 to L, y _i ^[h] is the count of photons or positrons detected at one pixel i on the projection data after the h-th divided dose, where y _i ^[h] is detected It is recorded in the h-th measurement data recording part from i = 1 to n later,
from h = 1 for L, S i ^_[h] was a scatter correction term in 1 pixel i on the projection data after divided doses of the h-th, where S _{i ^[h]} is the h i = 1 to n It is recorded in the time scatter correction term recording part,
The reconstructed image after the first divided administration is obtained by calculating the estimated RI density λ _j ^{k [1]} of the pixel j on the reconstructed image in the k-th successive approximation from k = 0 to a predetermined value using equation (1). A first reconstructed image generation step obtained by repeatedly calculating up to the number of repetitions;

Here, λ _j ^{0 [1]} is recorded in the first estimated RI concentration initial value recording unit, c _ij is recorded in the detection probability recording unit, and y _i ^[1] is recorded in the first measurement data recording unit. S _i ^[1] is recorded in the first scattering correction term recording unit,
From h = 2 to L, the reconstructed image after the h-th divided administration is used, and the estimated RI density λ _j ^{k [h]} of the pixel j on the reconstructed image in the k-th successive approximation is expressed by Equation (2). A h-th reconstructed image generation step obtained by repeatedly calculating from k = 0 to a predetermined number of repetitions,

Here, λ _j ^{0 [h]} is recorded in the h-th estimated RI concentration initial value recording unit, c _ij is recorded in the detection probability recording unit, and y _i ^[h] is recorded in the h-th measurement data recording unit. The cerebral blood flow quantitative analysis program is recorded in the recording section, and S _i ^[h] is recorded in the h-th scattering correction term recording section. The cerebral blood flow quantitative analysis program according to claim 2, wherein formula (3) is used instead of formula (1).

Quantitative analysis program for cerebral blood flow characterized by using The cerebral blood flow quantitative analysis program according to any one of claims 1 to 3, wherein the anatomical normalization step includes:
For the reconstructed image after each divided administration, an alignment step for correcting the reconstructed image after the second divided administration to the same position as the reconstructed image after the first divided administration;
A standardization step of obtaining each standard brain image by transforming the reconstructed image after the correction in the alignment step so as to match the standard brain using a spatial deformation parameter;
A cerebral blood flow quantitative analysis program comprising: a smoothing step of applying a smoothing filter to the reconstructed image after the standardization step. In the cerebral blood flow quantitative analysis program according to claim 4,
The deformation parameter is obtained as total projection data which is a sum of projection data weighted with a predetermined weight, and after obtaining a reconstructed image based on the total projection data, anatomical normalization is performed on the reconstructed image. A cerebral blood flow quantitative analysis program obtained by performing a process of obtaining a standard brain image by performing the process.   The cerebral blood flow quantitative analysis program according to any one of claims 1 to 5, wherein the predetermined quantification in the quantification step is performed by inputting the average cerebral blood flow at rest of the subject with respect to each standard brain image. A cerebral blood flow quantitative analysis program characterized by obtaining each quantitative image by obtaining local cerebral blood flow by correcting Lassen using values.   The cerebral blood flow quantitative analysis program according to any one of claims 1 to 6, further comprising a display step for displaying each quantitative image obtained in the quantification step. .   The cerebral blood flow quantitative analysis program according to any one of claims 1 to 7, wherein an increase amount and / or an increase rate of a local cerebral blood flow in a predetermined region is calculated based on each quantitative image obtained in the quantification step. A program for quantitative analysis of cerebral blood flow, further comprising a step of displaying an increase amount or the like to be displayed in a predetermined display format.   The cerebral blood flow quantification analysis program according to claim 8, wherein the predetermined display format displays a local cerebral blood flow value for each divided administration as a graph.   9. The cerebral blood flow quantitative analysis program according to claim 8, wherein the predetermined display format includes a local cerebral blood flow value in h-th (h is 2 or more) divided administration and local cerebral blood in divided administration before the h-th administration. A program for quantitative analysis of cerebral blood flow, characterized in that a difference from a flow value for each pixel is displayed using a predetermined color.   9. The cerebral blood flow quantitative analysis program according to claim 8, wherein the predetermined display format is an average value of local cerebral blood flow values in a predetermined region of interest in h-th (h is 2 or more) divided administration and h-th. A cerebral blood flow quantitative analysis program, characterized in that a difference with an average value of local cerebral blood flow values in the predetermined region of interest in the previous divided administration is displayed using a predetermined color so as to be identifiable.   The cerebral blood flow quantitative analysis program according to any one of claims 1 to 11, wherein one sheet constituting projection data after the first divided administration input in the projection data input step prior to the image reconstruction step A cerebral blood flow quantitative analysis program further comprising a confirmation display step of displaying a reconstructed image by applying a predetermined filter to the projection data of the slices.   The cerebral blood flow quantitative analysis program according to any one of claims 1 to 12, wherein an average value of local cerebral blood flow values in a predetermined region of interest is determined in a predetermined color based on each quantitative image obtained in the quantifying step. A program for quantitative analysis of cerebral blood flow, characterized in that it is displayed in an identifiable manner using a computer. A cerebral blood flow quantitative analysis program for quantitative analysis of cerebral blood flow by computer tomography measurement,
A projection data input step in which a predetermined drug is administered to the brains of n subjects, and projection data obtained by performing the computed tomography measurement after the administration is input;
An image reconstruction step for obtaining a reconstructed image after administration based on the projection data input in the projection data input step, wherein the n−1th person created by successive approximation in the successive approximation type image reconstruction method The nth reconstructed image is obtained by using the data obtained by projecting the reconstructed image as a baseline, and the baseline is updated each time.
An anatomical normalization step of obtaining a standard brain image by performing anatomical normalization on the reconstructed image after administration obtained in the image reconstruction step,
A cerebral blood flow quantification analysis program for executing a quantification step of obtaining a quantification image by performing predetermined quantification on the standard brain image obtained in the anatomical normalization step. The cerebral blood flow quantitative analysis program according to claim 14, wherein the anatomical normalization step includes:
About the reconstructed image after administration, a standardization step of obtaining a standard brain image by deforming the reconstructed image to match the standard brain using a spatial deformation parameter;
A cerebral blood flow quantitative analysis program comprising: a smoothing step of applying a smoothing filter to the reconstructed image after the standardization step. In the cerebral blood flow quantitative analysis program according to claim 15,
The deformation parameter is obtained based on the processing after obtaining a reconstructed image based on the projection data and performing anatomical normalization on the reconstructed image to obtain a standard brain image. A cerebral blood flow quantitative analysis program characterized by being.   The cerebral blood flow quantitative analysis program according to any one of claims 14 to 16, wherein the predetermined quantification in the quantification step is performed by inputting a mean cerebral blood flow value at rest of the subject input to a standard brain image. A program for quantitative analysis of cerebral blood flow, which obtains quantitative images by obtaining local cerebral blood flow by correcting Lassen using.   The cerebral blood flow quantitative analysis program according to any one of claims 14 to 17, further comprising a display step for displaying the quantitative image obtained in the quantification step.   The cerebral blood flow quantitative analysis program according to any one of claims 14 to 18, prior to the image reconstruction step, for one slice constituting the post-administration projection data input in the projection data input step. A cerebral blood flow quantitative analysis program, further comprising a confirmation display step for displaying a reconstructed image by applying a predetermined filter to projection data.   The cerebral blood flow quantification analysis program according to any one of claims 14 to 19, wherein an average value of local cerebral blood flow values in a predetermined region of interest is expressed in a predetermined color based on the quantitative image obtained in the quantification step. A program for quantitative analysis of cerebral blood flow, characterized in that it is displayed so that it can be identified.   A computer-readable recording medium in which the cerebral blood flow quantitative analysis program according to any one of claims 1 to 20 is recorded. A cerebral blood flow image data processing method for processing cerebral blood flow image data by computer tomography measurement, the computer comprising:
Given drug is n times to the subject's brain (n is 2 or more) comprising the steps of inputting the brain image data when it is divided doses, the obtained by performing a computed tomography measured after each divided doses A projection data input step for inputting each projection data;
An image reconstruction step for obtaining a reconstructed image after each divided administration based on each projection data input in the projection data input step, wherein n− created by successive approximation in the successive approximation type image reconstruction method The data obtained by projecting the first reconstructed image is used as a baseline to obtain the nth reconstructed image, and the baseline is updated each time.
An anatomical normalization step of obtaining each standard brain image by performing anatomical normalization on the reconstructed image after each divided administration obtained in the image reconstruction step;
A cerebral blood flow image data processing method comprising: a quantification step of performing predetermined quantification on each standard brain image obtained in the anatomical normalization step to obtain each quantification image. The cerebral blood flow image data processing method according to claim 22,
The computed tomography measurement is a single photon emission computed tomography measurement or a positron emission tomography measurement, and the predetermined agent comprises a tracer labeled with a radioisotope (RI);
The image reconstruction step includes:
m is the total number of pixels on the reconstructed image, j (= 1 to m) is a pixel on the reconstructed image, and n is a photon or positron emitted from a radioisotope (RI) administered to the brain The total number of pixels on the projection data or the total number of detectors that detect the photons or positrons, i (= 1 to n) are pixels on the projection data, where m and n are stored in the pixel number recording unit in advance. Recorded,
and the probability of photon or positron emission the c _ij from pixel j in the reconstructed image is detected on a pixel i on the projection data, wherein c _ij is pre j = 1 from to m and i = 1 from the n Is recorded in the detection probability recording part,
From h = 1, for the number L of divided doses (L is 2 or more), λ _j ^{0 [h]} is used to estimate pixel j on the reconstructed image in the 0th successive approximation used for the hth divided dose. RI concentration, where λ _j ^{0 [h]} is previously recorded in the h-th estimated RI concentration initial value recording unit from j = 1 to m,
For h = 1 to L, y _i ^[h] is the count of photons or positrons detected at one pixel i on the projection data after the h-th divided dose, where y _i ^[h] is detected It is recorded in the h-th measurement data recording part from i = 1 to n later,
from h = 1 for L, S i ^_[h] was a scatter correction term in 1 pixel i on the projection data after divided doses of the h-th, where S _{i ^[h]} is the h i = 1 to n It is recorded in the time scatter correction term recording part,
The reconstructed image after the first divided administration is obtained by calculating the estimated RI density λ _j ^{k [1]} of the pixel j on the reconstructed image in the k-th successive approximation from k = 0 to a predetermined value using equation (4). A first reconstructed image generation step obtained by repeatedly calculating up to the number of repetitions;

Here, λ _j ^{0 [1]} is recorded in the first estimated RI concentration initial value recording unit, c _ij is recorded in the detection probability recording unit, and y _i ^[1] is recorded in the first measurement data recording unit. S _i ^[1] is recorded in the first scattering correction term recording unit,
From h = 2 to L, the reconstructed image after the h-th divided administration is used, and the estimated RI density λ _j ^{k [h]} of the pixel j on the reconstructed image in the k-th successive approximation is expressed by Equation (5). A h-th reconstructed image generation step obtained by repeatedly calculating from k = 0 to a predetermined number of repetitions,

Here, λ _j ^{0 [h]} is recorded in the h-th estimated RI concentration initial value recording unit, c _ij is recorded in the detection probability recording unit, and y _i ^[h] is recorded in the h-th measurement data recording unit. The cerebral blood flow image data processing method is characterized in that S _i ^[h] is recorded in the h th scatter correction term recording unit. The cerebral blood flow image data processing method according to claim 23, wherein the formula (6) is substituted for the formula (4).

A method for processing cerebral blood flow image data, comprising: The cerebral blood flow image data processing method according to any one of claims 22 to 24, wherein the anatomical normalization step includes:
For the reconstructed image after each divided administration, an alignment step for correcting the reconstructed image after the second divided administration to the same position as the reconstructed image after the first divided administration;
A standardization step of obtaining each standard brain image by transforming the reconstructed image after the correction in the alignment step so as to match the standard brain using a spatial deformation parameter;
A cerebral blood flow image data processing method comprising: a smoothing step of applying a smoothing filter to the reconstructed image after the standardization step. The cerebral blood flow image data processing method according to claim 25,
The deformation parameter is obtained as total projection data which is a sum of projection data weighted with a predetermined weight, and after obtaining a reconstructed image based on the total projection data, anatomical normalization is performed on the reconstructed image. A method for processing cerebral blood flow image data obtained by performing a process of obtaining a standard brain image by performing the above-described process. 27. The cerebral blood flow image data processing method according to any one of claims 22 to 26, wherein the predetermined quantification in the quantification step is performed by inputting, for each standard brain image, an average cerebral blood pressure of the subject at rest. A method for processing cerebral blood flow image data, wherein a local cerebral blood flow is obtained by performing Lassen correction using a flow value to obtain each quantitative image . The cerebral blood flow image data processing method according to any one of claims 22 to 27 , wherein the projection data after the first divided administration input in the projection data input step is configured prior to the image reconstruction step. A method for processing cerebral blood flow image data, further comprising a confirmation display step for displaying a reconstructed image by applying a predetermined filter to projection data for one slice . A cerebral blood flow image data processing method for processing cerebral blood flow image data by computer tomography measurement,
A projection data input step of inputting projection data obtained by performing the computed tomography measurement after a predetermined drug is administered to the brain of n subjects;
An image reconstruction step for obtaining a reconstructed image after administration based on the projection data input in the projection data input step, wherein the n−1th person created by successive approximation in the successive approximation type image reconstruction method The nth reconstructed image is obtained by using the data obtained by projecting the reconstructed image as a baseline, and the baseline is updated each time.
An anatomical normalization step of obtaining a standard brain image by performing anatomical normalization on the reconstructed image after administration obtained in the image reconstruction step;
A cerebral blood flow image data processing method comprising: a quantification step of performing predetermined quantification on the standard brain image obtained in the anatomical normalization step to obtain each quantification image. The cerebral blood flow image data processing method according to claim 29 , wherein the anatomical normalization step comprises:
About the reconstructed image after administration, a standardization step of obtaining a standard brain image by deforming the reconstructed image to match the standard brain using a spatial deformation parameter;
A cerebral blood flow image data processing method comprising: a smoothing step of applying a smoothing filter to the reconstructed image after the standardization step. The cerebral blood flow image data processing method according to claim 30 ,
The deformation parameter is obtained based on the processing after obtaining a reconstructed image based on the projection data and performing anatomical normalization on the reconstructed image to obtain a standard brain image. A method for processing cerebral blood flow image data. 32. The cerebral blood flow image data processing method according to any one of claims 29 to 31 , wherein the predetermined quantification in the quantification step is performed by comparing the standard cerebral image with the average cerebral blood flow at rest of the subject input. A method for processing cerebral blood flow image data, wherein a quantitative image is obtained by obtaining local cerebral blood flow by performing Lassen correction using values .

Images