Attenuation distribution image creation device, image processing device, radiation tomography system, attenuation distribution image creation method, and image processing method

Abstract

To provide a device capable of creating a more accurate attenuation distribution image from a tomographic image of an object of attenuation correction even if a medicine dosed to an analyte when the tomographic image of the object of attenuation correction is different from a medicine dosed to the analyte when a tomographic image for CNN learning is acquired.SOLUTION: An image processing device 20 comprises an image reconstitution part 21, an image conversion part 22, a CNN processing part 23, and an attenuation correction part 24. The image reconstitution part 22 creates a conversion image by changing pixel values of a partial area based upon tomographic images of respective slice positions of the analyte. The CNN processing part 23 inputs conversion images of the respective slice positions created by the image conversion part 22 to a CNN 31, and also inputs slice position information on the conversion images to an embedded layer 32. Output data from the embedded layer 32 are added to a feature map of the CNN 31. The CNN processing part 23 outputs an output image from the CNN 31 as an attenuation distribution image of the analyte.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to an apparatus and method for creating an attenuation distribution image, and an apparatus and method for performing attenuation correction of a tomographic image using the attenuation distribution image. [Background technology]

[0002] Radiation tomography apparatuses capable of acquiring tomographic images of a subject (living body) include PET (Positron Emission Tomography) apparatuses and SPECT (Single Photon Emission Computed Tomography) apparatuses.

[0003] The PET device is equipped with a detection section having a number of small radiation detectors arranged around the measurement space in which the subject is placed. The PET device detects photon pairs with energy of 511 keV generated by the annihilation of electrons and positrons in the subject into which a positron-emitting isotope (RI source) is introduced, using a coincidence method with the detection section, and collects the coincidence information. Then, based on the collected number of coincidence information, a tomographic image showing the spatial distribution of the occurrence frequency of photon pairs in the measurement space (i.e., the spatial distribution of the RI source) can be reconstructed. At this time, list data in which the coincidence information collected by the PET device is arranged in chronological order is divided into multiple frames in the order of collection, and image reconstruction processing is performed using data contained in each frame of the list data, thereby obtaining a dynamic PET image consisting of multiple tomographic images. This PET device plays an important role in the field of nuclear medicine, etc., and can be used to study, for example, biological functions and higher-level brain functions.

[0004] Photons generated within a subject are absorbed and attenuated when passing through the subject, and are also absorbed and attenuated when passing through the bed on which the subject is placed. Therefore, attenuation correction is required to obtain more accurate tomographic images. For example, a PET device is used to perform measurements (emission scan) on a subject administered with a drug containing an RI source to obtain a tomographic image (emission scan image), and measurements (transmission scan) are performed while rotating the RI source around the subject not administered with the drug to obtain a tomographic image (transmission scan image), and the attenuation correction of the emission scan image is performed using the transmission scan image.

[0005] For attenuation correction of emission scan images, in addition to using transmission scan images obtained by performing a transmission scan with a PET device as described above, X-ray CT scan images obtained by performing a scan with an X-ray CT device can also be used.

[0006] Hereinafter, the attenuation distribution image at energy 511 keV obtained from the transmission scan image and the X-ray CT scan image will be referred to as the attenuation distribution image. The attenuation distribution image shows the attenuation distribution of photons in the subject. Hereinafter, the emission scan image will be referred to as the tomographic image. The attenuation distribution image can be used to perform attenuation correction of the tomographic image.

[0007] To obtain more accurate tomographic images by performing attenuation correction, the PET device needs to have a transmission scan mechanism or an X-ray CT scan mechanism. This is one of the factors that hinder the simplification and cost reduction of PET devices. A technology intended to solve this problem is described in Non-Patent Document 1.

[0008] The technology described in Non-Patent Document 1 trains a convolutional neural network (CNN) using a large number of sets of learning tomographic images and attenuation distribution images as teacher images, creates an attenuation distribution image from a tomographic image to be attenuation corrected using this trained CNN, and performs attenuation correction of the tomographic image using this attenuation distribution image. This technology enables attenuation correction by using the trained CNN even if the PET device does not have a transmission scan mechanism or an X-ray CT scan mechanism. [Prior art documents] [Non-patent literature]

[0009] [Non-Patent Document 1] Hashimoto, et al. "Deeplearning-based attenuation correction for brain PET with various radiotracers." Annals of Nuclear Medicine, 2021. Summary of the Invention [Problem to be solved by the invention]

[0010] The present inventors have found that the technology described in Non-Patent Document 1 has the following problems. If the drug administered to the subject when the tomographic image to be attenuation corrected is acquired is the same as the drug administered to the subject when the tomographic image for CNN learning is acquired, the attenuation distribution image created from the tomographic image to be attenuation corrected is relatively accurate. However, if the drug administered to the subject when the tomographic image to be attenuation corrected is acquired is different from the drug administered to the subject when the tomographic image for CNN learning is acquired, the accuracy of the attenuation distribution image created from the tomographic image to be attenuation corrected is poor.

[0011] The present invention has been made to solve the above problems, and aims to provide an apparatus and method that can create a more accurate attenuation distribution image from a tomographic image to be attenuation corrected, even if the drug administered to the subject when the tomographic image to be attenuation corrected was acquired is different from the drug administered to the subject when the tomographic image for CNN learning was acquired. [Means for solving the problem]

[0012] The attenuation distribution image generating apparatus of the present invention is an apparatus for generating an attenuation distribution image of a subject based on a tomographic image of the subject.

[0013] A first aspect of the attenuation distribution image creation device of the present invention includes: (1) an image conversion unit that generates a converted image in which pixel values ​​of a portion of a region are changed based on a tomographic image of each slice position of a subject; and (2) a CNN processing unit that inputs the converted image of each slice position generated by the image conversion unit to a convolutional neural network, inputs slice position information of the converted image to an embedding layer, adds output data from the embedding layer to a feature map of the convolutional neural network, and outputs an output image from the convolutional neural network as an attenuation distribution image of the subject.

[0014] A second embodiment of the attenuation distribution image creation device of the present invention, in addition to the first embodiment, further includes a CNN learning unit that uses a plurality of sets of learning tomographic images and attenuation distribution images as teacher images, inputs converted images of each slice position generated by the image conversion unit based on the learning tomographic images to a convolutional neural network, and when slice position information of the converted images is input to an embedding layer, evaluates an error between an output image from the convolutional neural network and the attenuation distribution image as the teacher image, and trains the convolutional neural network and the embedding layer based on the error evaluation result.

[0015] The image processing device of the present invention includes the above-mentioned attenuation distribution image creation device of the present invention that creates an attenuation distribution image of a subject based on a tomographic image of the subject, and an attenuation correction unit that performs attenuation correction of the tomographic image of the subject using the attenuation distribution image of the subject.

[0016] The radiation tomography system of the present invention comprises a radiation tomography device for acquiring a tomographic image of a subject, the above-mentioned attenuation distribution image creation device of the present invention for creating an attenuation distribution image of the subject based on the tomographic image of the subject, and an attenuation correction unit for performing attenuation correction of the tomographic image of the subject using the attenuation distribution image of the subject.

[0017] The attenuation distribution image creating method of the present invention is a method for creating an attenuation distribution image of a subject based on a tomographic image of the subject.

[0018] A first aspect of the attenuation distribution image creation method of the present invention includes: (1) an image conversion step of generating a converted image in which pixel values ​​of a portion of the image are changed based on a tomographic image of each slice position of the subject; and (2) a CNN processing step of inputting the converted image of each slice position generated in the image conversion step to a convolutional neural network, inputting slice position information of the converted image to an embedding layer, adding output data from the embedding layer to a feature map of the convolutional neural network, and outputting an output image from the convolutional neural network as an attenuation distribution image of the subject.

[0019] A second embodiment of the attenuation distribution image creating method of the present invention, in addition to the first embodiment, further includes a CNN learning step of using a plurality of sets of learning tomographic images and attenuation distribution images as teacher images, inputting converted images of each slice position generated in the image conversion step based on the learning tomographic images to a convolutional neural network, and inputting slice position information of the converted images to an embedding layer, evaluating an error between an output image from the convolutional neural network and the attenuation distribution image as the teacher image, and training the convolutional neural network and the embedding layer based on the error evaluation result.

[0020] The image processing method of the present invention comprises the above-mentioned attenuation distribution image creation method of the present invention, which creates an attenuation distribution image of the subject based on a tomographic image of the subject, and an attenuation correction step of performing attenuation correction of the tomographic image of the subject using the attenuation distribution image of the subject. Effect of the Invention

[0021] According to the present invention, even if the drug administered to the subject when the tomographic image to be attenuation corrected was acquired is different from the drug administered to the subject when the tomographic image for CNN learning was acquired, a more accurate attenuation distribution image can be created from the tomographic image to be attenuation corrected. [Brief description of the drawings]

[0022] [Figure 1] FIG. 1 is a diagram showing the configuration of a radiation tomography system 1. As shown in FIG. [Diagram 2] FIG. 2 is a diagram showing the configuration of the attenuation distribution image creating device. [Diagram 3] FIG. 3 is a diagram illustrating the configuration of the CNN processing unit 23. [Figure 4] FIG. 4 shows images of a subject administered [18F]FDG. [Diagram 5] FIG. 5 shows images of subjects administered with [11C]Raclopride. [Figure 6] FIG. 6 shows images of subjects administered [11C]PIB. [Figure 7] FIG. 7 shows images of subjects administered with [11C]PBB3. [Figure 8] FIG. 8 shows images of a subject administered with [11C]Methionine. [Figure 9] FIG. 9 shows images of subjects administered [11C]DPA713. [Figure 10] FIG. 10 is a graph showing the PSNR of the attenuation distribution image of all slices obtained based on the tomographic image to be attenuation corrected for each drug. [Figure 11]FIG. 11 is a graph showing the PSNR of the attenuation distribution image of slice 5 obtained based on the tomographic image to be attenuation corrected for each drug. [Figure 12] FIG. 12 is a graph showing the PSNR of the attenuation distribution image of the slice 20 obtained based on the tomographic image to be attenuation corrected for each drug. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0023] Hereinafter, the embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the description of the drawings, the same elements are given the same reference numerals, and duplicated descriptions are omitted. The present invention is not limited to these examples, but is indicated by the claims, and is intended to include all modifications within the meaning and scope equivalent to the claims.

[0024] 1 is a diagram showing the configuration of a radiation tomography system 1. The radiation tomography system 1 includes a radiation tomography device 10 and an image processing device 20. The image processing device 20 includes an image reconstruction unit 21, an image conversion unit 22, a CNN processing unit 23, and an attenuation correction unit 24. Among these, the image conversion unit 22 and the CNN processing unit 23 constitute an attenuation distribution image creation device that creates an attenuation distribution image based on a tomographic image to be attenuation corrected.

[0025] The radiation tomography apparatus 10 is an apparatus that performs an emission scan to collect list data (emission scan data) for reconstructing a tomographic image (emission scan image) of a subject. Examples of the radiation tomography apparatus 10 include a PET apparatus and a SPECT apparatus. In the following, the radiation tomography apparatus 10 will be described as a PET apparatus.

[0026] The radiation tomography apparatus 10 includes a detection unit having a large number of small radiation detectors arranged around a measurement space in which a subject placed on a bed is placed. The radiation tomography apparatus 10 detects photon pairs with energy of 511 keV generated by annihilation of electrons and positrons in a subject into which a positron-emitting isotope (RI source) has been introduced, by a coincidence method using the detection unit, and accumulates this coincidence information. The radiation tomography apparatus 10 then outputs the accumulated coincidence information arranged in chronological order to the image processing device 20 as list data (emission scan data).

[0027] The list data includes identification information and detection time information of a pair of radiation detectors that have counted the photon pairs coincidentally. The list data may further include energy information of the photons detected by each radiation detector, and detection time difference information of the pair of radiation detectors.

[0028] The image processing device 20 is a device that reconstructs a tomographic image of the subject after attenuation correction based on the emission scan data of the subject output from the radiation tomography device 10. The image processing device 20 reconstructs an emission scan image (a tomographic image to be attenuation corrected) by an image reconstruction unit 21 based on the emission scan data, creates an attenuation distribution image by an attenuation distribution image creation device (image conversion unit 22, CNN processing unit 23) based on this tomographic image to be attenuation corrected, and performs attenuation correction of the tomographic image by an attenuation correction unit 24 based on this attenuation distribution image.

[0029] A computer having a CPU, a GPU, a RAM, a ROM, a hard disk drive, etc. is used as the image processing device 20. The image processing device 20 also includes an input unit (e.g., a keyboard and a mouse) that accepts input from an operator, and a display unit (e.g., a liquid crystal display) that displays images, etc.

[0030] The image reconstruction unit 21 reconstructs a tomographic image based on the emission scan data acquired by the radiation tomography apparatus 10. As a technique for reconstructing a tomographic image based on the emission scan data in the image reconstruction unit 21, the maximum likelihood expectation maximization (ML-EM) method and an improved iterative image reconstruction technique using a block iteration method are known. In addition, as the iterative image reconstruction technique using the block iteration method, the ordered subset ML-EM (OSEM) method, the row-action maximum likelihood algorithm (RAMLA) method, and the dynamic RAMLA (DRAMA) method are known.

[0031] The image conversion unit 22 generates a converted image in which pixel values ​​of a partial region are changed based on a tomographic image at each slice position of the subject. The converted image is a converted image in which pixel values ​​of a partial region are changed based on a tomographic image to be attenuation corrected, and preferably, a converted image in which pixel values ​​of a partial region of a subject region in a tomographic image to be attenuation corrected are changed. The pixel value change of the converted image for the tomographic image can be in any manner. The partial region in which pixel values ​​are changed may be of any shape, any size, and any number. The change of pixel values ​​in the partial region may be, for example, a nonlinear conversion of pixel values, an exchange of pixel values ​​between a plurality of pixels, a replacement with a fixed pixel value, a replacement with a random pixel value, or the like.

[0032] The CNN processing unit 23 inputs the converted image of each slice position generated by the image conversion unit 22 and slice position information of the converted image. The CNN processing unit 23 inputs the converted image of each slice position to a trained convolutional neural network (CNN) 31, and inputs the slice position information of the converted image to an embedding layer 32. Output data from the embedding layer 32 is added to the feature map of the CNN 31. Then, the CNN processing unit 23 outputs the output image from the CNN 31 as an attenuation distribution image of the subject.

[0033] The attenuation correction unit 24 inputs the emission scan data and the attenuation distribution image created by the CNN processing unit 23. Then, the attenuation correction unit 24 performs image reconstruction with attenuation correction based on the emission scan data and the attenuation distribution image, and creates a tomographic image after the correction.

[0034] The attenuation distribution image creating method of this embodiment is a method for creating an attenuation distribution image of a subject based on a tomographic image of the subject. The image processing method of this embodiment includes the attenuation distribution image creating method of this embodiment, and is a method for performing attenuation correction of a tomographic image using the created attenuation distribution image.

[0035] In the image processing method of this embodiment, an emission scan image (tomographic image to be subjected to attenuation correction) is reconstructed based on the emission scan data in an image reconstruction step, an attenuation distribution image is created based on the tomographic image to be subjected to attenuation correction in an attenuation distribution image creation step (image conversion step, CNN processing step), and attenuation correction of the tomographic image is performed based on the attenuation distribution image in an attenuation correction step.

[0036] In the image reconstruction step, the image reconstruction unit 21 reconstructs a tomographic image based on the emission scan data acquired by the radiation tomography apparatus 10. In the image conversion step, the image conversion unit 22 generates a converted image in which pixel values ​​of a partial region are changed based on the tomographic image at each slice position of the subject.

[0037] In the CNN processing step, the CNN processing unit 23 inputs the converted image of each slice position to the trained CNN 31, and inputs slice position information of the converted image to the embedding layer 32. The output data from the embedding layer 32 is added to the feature map of the CNN 31. Then, in the CNN processing step, the output image from the CNN 31 is output as an attenuation distribution image of the subject.

[0038] In the attenuation correction step, the attenuation correction unit 24 performs image reconstruction with attenuation correction based on the mission scan data and the attenuation distribution image, and creates a tomographic image after the correction.

[0039] FIG. 2 is a diagram showing the configuration of an attenuation distribution image creation device. The attenuation distribution image creation device includes a CNN learning unit 25 in addition to an image conversion unit 22 and a CNN processing unit 23. This diagram shows the configuration of a system for learning the CNN 31 and the embedding layer 32 used in the CNN processing unit 23. When learning the CNN 31 and the embedding layer 32, first, a plurality of sets of learning tomographic images and attenuation distribution images as teacher images are prepared. The learning tomographic images prepared here may be tomographic images of a subject to which any drug has been administered. In addition, the attenuation distribution images prepared here as teacher images may be transmission scan images obtained by performing a transmission scan with a PET device, or may be X-ray CT scan images obtained by performing a scan with an X-ray CT device.

[0040] A converted image of each slice position generated by the image conversion unit 22 based on the learning tomographic image is input to the CNN 31, and slice position information of the converted image is input to the embedding layer 32. The CNN learning unit 25 evaluates an error (e.g., L1 norm, L2 norm, negative log-likelihood in Poisson distribution, etc.) between the output image from the CNN 31 and the attenuation distribution image as a teacher image, and corrects the parameters of the CNN 31 and the embedding layer 32 based on the error evaluation result. By performing such a series of processes using multiple sets of learning tomographic images and attenuation distribution images as teacher images, the CNN 31 and the embedding layer 32 are trained (CNN learning step).

[0041] 3 is a diagram illustrating the configuration of the CNN processing unit 23. The CNN 31 has a U-net structure including an encoder and a decoder.

[0042] The encoder unit of the CNN 31 has a convolution layer, a normalization layer, a nonlinear activation function, and downsampling. The encoder unit inputs the converted image output from the image conversion unit 22, and updates the feature map using the convolution layer, normalization layer, and nonlinear activation function. After performing processing using the convolution layer, normalization layer, and nonlinear activation function several times, the encoder unit reduces the spatial size of the feature map by downsampling and increases the number of channels of the feature map. For example, the spatial size of the feature map is initially 128×128, but is reduced to 64×64, 32×32, 16×16, and so on.

[0043] The decoder unit of the CNN31 is provided after the encoder unit. The decoder unit has a convolution layer, a normalization layer, a nonlinear activation function, and upsampling. The decoder unit expands the spatial size of the feature map and reduces the number of channels of the feature map by upsampling. For example, the spatial size of the feature map is expanded to 16x16, 32x32, 64x64, 128x128, and so on. After performing the upsampling process, the decoder unit updates the feature map by convolution, a normalization layer, and a nonlinear activation function.

[0044] The embedding layer 32 inputs slice position information of the converted image input to the CNN 31, converts the slice position information into a data string, and outputs the data string. The output data from the embedding layer 32 is added to the feature map of the CNN 31. In this figure, the embedding layer 32 outputs a data string [0.2, 0.5, ..., 0.3, ...], of which the first data, 0.2, is added to the feature map of the first channel after the first downsampling of the encoder unit of the CNN 31, and the second data, 0.5, is added to the feature map of the second channel after the first downsampling of the encoder unit of the CNN 31. The output data from the embedding layer 32 is added to the feature map of each channel after downsampling or upsampling of the CNN 31. The output data from the embedding layer 32 may be added to the feature map of each channel after all downsampling and all upsampling of the CNN 31.

[0045] Next, an example will be described. In this example, 188 sets of uncorrected tomographic images and attenuation distribution images of the subject's head were obtained using a head-dedicated PET device. 18 [F]FDG, 11 C]Raclopride, 11 C]PIB, 11 C]PBB3, 11 C]Methionine and [ 11 Either [C]DPA713 or [ 18 The tomographic images and attenuation distribution images without attenuation correction obtained for the subject administered [F]FDG were used for learning. The attenuation distribution images created by CNN in this embodiment and the comparative example were compared using the technology described in Non-Patent Document 1 as a comparative example.

[0046] 4 to 9 are diagrams showing images obtained in the examples and comparative examples. 18 Figure 5 shows images of a subject administered [F]FDG. 11 C]Raclopride was administered to the subject. 11FIG. 7 shows images of subjects administered [C]PIB. 11 FIG. 8 shows images of subjects administered [C]PBB3. 11 C]Methionine was administered to the subject. 11 C] Images of subjects administered DPA713 are shown.

[0047] In each of Figs. 4 to 9, the top row shows a tomographic image without attenuation correction, the second row shows an attenuation distribution image (correct image), the third row shows an attenuation distribution image created by CNN in the comparative example, and the bottom row shows an attenuation distribution image created by CNN in the embodiment. In each of Figs. 4 to 9, the left column shows images of slice 5, and the right column shows images of slice 20. The larger the slice number, the higher the position of the subject's head (the closer to the top of the head). In a head-only PET device, an emission scan and a transmission scan are performed using a jig that positions the subject's head, so even if the subject is different, the same slice number indicates approximately the same position of the head.

[0048] 10 to 12 are graphs showing the PSNR of attenuation distribution images obtained based on tomographic images to be attenuation corrected for each drug in the examples and comparative examples. Fig. 10 shows the PSNR of the attenuation distribution images of all slices, Fig. 11 shows the PSNR of the attenuation distribution image of slice 5, and Fig. 12 shows the PSNR of the attenuation distribution image of slice 20. PSNR (Peak Signal to Noise Ratio) represents image quality in decibels (dB), and a higher value means better image quality.

[0049] As shown in Figs. 4 to 12, when the tomographic images to be attenuation corrected were acquired, the drug administered to the subject was 18 In the case of [F]FDG, accurate attenuation distribution images could be created even in the comparative example. 11 C]DPA713 is [ 18 F]FDG, the drug administered to the subject when the tomographic image to be attenuation corrected was obtained was [ 11In both the case of [C]DPA713 and the comparative example, accurate attenuation distribution images could be created. 18 When different from [F]FDG, in this embodiment, a more accurate attenuation distribution image could be created from the tomographic image subject to attenuation correction, as compared with the comparative example.

[0050] As described above, in this embodiment, even if a transmission scan mechanism or an X-ray CT scan mechanism is not provided, an attenuation distribution image of a subject can be created based on a tomographic image of the subject that is to be attenuation corrected, and an attenuation-corrected tomographic image can be created, so that the device can be simplified and the cost can be reduced. Also, in this embodiment, a more accurate attenuation distribution image can be created, and by using this, attenuation correction of a tomographic image can be performed with higher accuracy. [Explanation of symbols]

[0051] 1...Radiation tomography system, 10...Radiation tomography device, 20...Image processing device, 21...Image reconstruction unit, 22...Image conversion unit, 23...CNN processing unit, 24...Attenuation correction unit, 25...CNN learning unit, 31...Convolutional neural network (CNN), 32...Embedding layer.

Claims

1. 1. An apparatus for creating an attenuation distribution image of a subject based on a tomographic image of the subject, comprising: an image conversion unit that generates a converted image by changing pixel values ​​of a partial region based on the tomographic image of each slice position of the subject; a CNN processing unit that inputs the converted image of each slice position generated by the image conversion unit to a convolutional neural network, inputs slice position information of the converted image to an embedding layer, adds output data from the embedding layer to a feature map of the convolutional neural network, and outputs an output image from the convolutional neural network as an attenuation distribution image of the subject; An attenuation distribution image creating device comprising:

2. The method further includes a CNN learning unit that uses a plurality of sets of learning tomographic images and attenuation distribution images as teacher images, inputs converted images of each slice position generated by the image conversion unit based on the learning tomographic images to the convolutional neural network, and when slice position information of the converted images is input to the embedding layer, evaluates an error between an output image from the convolutional neural network and an attenuation distribution image as a teacher image, and trains the convolutional neural network and the embedding layer based on the error evaluation result. The attenuation distribution image generating device according to claim 1 .

3. an attenuation distribution image creating device according to claim 1, which creates an attenuation distribution image of a subject based on a tomographic image of the subject; an attenuation correction unit that performs attenuation correction of the tomographic image of the subject using the attenuation distribution image of the subject; An image processing device comprising:

4. a radiation tomography apparatus for acquiring a tomographic image of a subject; an attenuation distribution image creating device according to claim 1 , which creates an attenuation distribution image of the subject based on the tomographic image of the subject; an attenuation correction unit that performs attenuation correction of the tomographic image of the subject using the attenuation distribution image of the subject; A radiation tomography system comprising:

5. 1. A method for creating an attenuation distribution image of a subject based on a tomographic image of the subject, comprising: an image conversion step of generating a converted image by changing pixel values ​​of a partial region based on the tomographic images of each slice position of the subject; a CNN processing step of inputting the converted image of each slice position generated in the image conversion step into a convolutional neural network, inputting slice position information of the converted image into an embedding layer, adding output data from the embedding layer to a feature map of the convolutional neural network, and outputting an output image from the convolutional neural network as an attenuation distribution image of the subject; The attenuation distribution image creating method includes the steps of:

6. The method further includes a CNN learning step of using a plurality of sets of learning tomographic images and attenuation distribution images as teacher images, inputting converted images of each slice position generated in the image conversion step based on the learning tomographic images to the convolutional neural network, and inputting slice position information of the converted images to the embedding layer, evaluating an error between an output image from the convolutional neural network and an attenuation distribution image as a teacher image, and learning the convolutional neural network and the embedding layer based on the error evaluation result. The method for producing an attenuation distribution image according to claim 5.

7. The attenuation distribution image creating method according to claim 5, further comprising creating an attenuation distribution image of the subject based on a tomographic image of the subject; an attenuation correction step of performing attenuation correction of the tomographic image of the subject using the attenuation distribution image of the subject; An image processing method comprising:

Images