JP2020067865A - Image processing apparatus, image processing system, and image processing program - Google Patents

Abstract

To provide an image processing apparatus capable of extracting an abnormal part from a medical image, image processing system, and an image processing program.SOLUTION: The image processing apparatus includes: an autoencoder configured to restore a medical image that is normal with respect to an input medical image and output the same; a dispersion estimation unit configured to output the dispersion of the anomalous image for the input medical image; image input acceptance means that accepts the input of the medical image of a person to be diagnosed; difference map generation means that calculates the difference between the input image and the restored image output by the autoencoder, and generates a difference map; standard deviation calculation means that calculates the standard deviation on the basis of the estimated dispersion output from the dispersion estimation unit; and diagnostic result image generation means configured to extract the difference between the input image and the restored image on the basis of the difference map and standard deviation and generate a diagnostic result image reduced with the differences other than the characteristic differences included in the input image.SELECTED DRAWING: Figure 8

Description

  The present invention relates to an image processing device, an image processing system, and an image processing program.

  The following biometric image processing devices are known. This biometric image processing apparatus accepts an input training image that is a biometric image that includes a target object, an output training image that is a biometric image that does not include the target object, and an input image that is a biometric image that includes the target object, Using the training image and the training image for output, learn a neural network that performs image processing to generate a biometric image that does not include the target from the biometric image that includes the target, and use the image processing of the neural network that is the learning result. An output image is generated by removing the object from the input image (for example, Patent Document 1).

JP, 2018-89301, A

  In a conventional biometric image processing apparatus, a neural network that performs image processing to generate a biometric image that does not include a target from a biometric image that includes a target is learned, and the image processing of the neural network that is the learning result is used to input the input image. The target image was removed from the output image. However, in the diagnosis using images, it is often required to find a site having a possibility of abnormality in the medical image. Therefore, there is a demand for a mechanism that can learn the image of a person without any abnormality and use the learning result to assist in identifying the abnormal portion when the image of the person with abnormality is input. However, in the conventional technology, no consideration has been given to that point.

An image processing apparatus according to the present invention is an auto encoder configured to learn an image pattern of a medical image having no abnormality and restore and output a medical image having no abnormality with respect to an input medical image. Input the medical image of the diagnosis target person, and the variance estimator configured to learn the variance of the medical image without abnormality and output the variance of the image without abnormality with respect to the input medical image An image input receiving unit that receives an input as an image, a difference between an input image input by the image input receiving unit and a restored image output by the auto encoder as a result of inputting the input image is calculated to generate a difference map. Difference map generating means, standard deviation calculating means for calculating the standard deviation based on the estimated variance output from the variance estimator as a result of inputting the input image, and the difference map Based on the difference map generated by the generating means and the standard deviation calculated by the standard deviation calculating means, the difference portion between the input image and the restored image is extracted, and the difference portion other than the characteristic difference portion included in the input image is extracted. And a diagnostic result image generating means for generating a diagnostic result image with reduced noise.
An image processing system according to the present invention includes an automatic encoder configured to learn an image pattern of a medical image having no abnormality and restore and output a medical image having no abnormality with respect to an input medical image. Input the medical image of the diagnosis target person, and the variance estimator configured to learn the variance of the medical image without abnormality and output the variance of the image without abnormality with respect to the input medical image An image input acceptance unit that accepts an input as an image, an input image input by the image input acceptance unit, and a difference between the restored image output by the auto encoder as a result of inputting the input image are calculated to generate a difference map. Difference map generating means, standard deviation calculating means for calculating standard deviation based on the estimated variance output from the variance estimator as a result of inputting the input image, and the difference Based on the standard deviation calculated by the standard deviation calculating unit and the difference map generated by the step-up generating unit, the difference portion between the input image and the restored image is extracted, and the difference portion other than the characteristic difference portion included in the input image is extracted. Diagnostic result image generating means for generating a diagnostic result image in which the difference is reduced.
An image processing program according to the present invention, an image input acceptance procedure for accepting an input of a medical image of a person to be diagnosed as an input image, and an image pattern of a medical image having no abnormality is learned, and an input medical image is input. Input the input image received in the image input procedure to the auto encoder configured to restore and output a medical image with no abnormality, and obtain the restored image by the auto encoder. The difference map generation procedure that calculates the difference between the input image received in the procedure and the restored image acquired in the restored image acquisition procedure and the difference map generation procedure to learn the variance of the medical image without abnormality and input Input the input image accepted in the image input procedure to the variance estimator configured to output the variance of the image that is normal to the medical image. Estimated variance acquisition procedure to obtain estimated variance by variance estimator, standard deviation calculation procedure to calculate standard deviation based on estimated variance acquired in estimated variance acquisition procedure, difference map generated in difference map generation procedure and standard deviation calculation Based on the standard deviation calculated in the procedure, the diagnostic result image that generates the diagnostic result image in which the differential part between the input image and the restored image is extracted and the differential part other than the characteristic differential part included in the input image is reduced A program for causing a computer to execute the generation procedure.

  According to the present invention, an auto encoder configured to learn an image pattern of a medical image having no abnormality and restore and output a medical image having no abnormality with respect to an input medical image is used. , It is possible to extract the difference part between the input image and the restored image and generate the diagnosis result image in which the difference part other than the characteristic difference part included in the input image is reduced. It can assist in the identification of the part.

1 is a block diagram showing the configuration of an embodiment of an image processing device 100. FIG. It is the figure which showed the flow of the process by an auto encoder typically. It is the figure which showed typically the flow of a process by a dispersion estimator. It is the figure which showed typically the preparation method of a difference map. It is the figure which showed typically the method for reducing the difference part other than an abnormal part using a difference map and estimated standard deviation. It is the figure which summarized the flow of a training step and an estimation step into one, and was shown typically. It is a flowchart figure which shows the flow of a training step. It is a flowchart figure which shows the flow of an estimation step.

  FIG. 1 is a block diagram showing a configuration of an embodiment of an image processing apparatus 100 according to the present embodiment. The image processing apparatus 100 executes a process for extracting and visualizing an abnormal part from a medical image using a medical image captured in advance, for example, an MRI image, a CT image, and an X-ray image. As the image processing device 100, for example, a server device or a personal computer is used. FIG. 1 is a block diagram showing the configuration of an embodiment in which a personal computer is used as the image processing apparatus 100 in this embodiment.

  The image processing apparatus 100 includes an operation member 101, a control device 102, a storage medium 103, and a display device 104.

  The operation member 101 includes various devices operated by an operator of the image processing apparatus 100, such as a keyboard and a mouse.

  The control device 102 includes a CPU, a memory, and other peripheral circuits, and controls the entire image processing device 100. The memory forming the control device 102 is a volatile memory such as an SDRAM. This memory is used as a work memory for the CPU to expand the program when the program is executed and a buffer memory for temporarily recording data. For example, the data read via the connection interface 102 is temporarily recorded in the buffer memory.

  The storage medium 103 is a storage medium for recording various data stored in the image processing apparatus 100, data of a program to be executed by the control apparatus 102, and the like, and is, for example, an HDD (Hard Disk Drive) or SSD (Solid State). Drive) or the like is used. The program data recorded in the storage medium 103 is provided by being recorded in a recording medium such as a CD-ROM or a DVD-ROM or provided via a network, and stores the program data acquired by the operator. Installation on the medium 103 allows the controller 102 to execute the program. In the present embodiment, programs and various data used in the processing described below are recorded in the storage medium 103.

  The display device 104 is, for example, a liquid crystal monitor, and displays various display data output from the control device 102.

  In the present embodiment, in advance, a medical image having no abnormality, that is, a medical image obtained by capturing a specific region of a patient having no abnormality is captured, and the image pattern and pixel distribution of the medical image having no abnormality are dispersed. Training steps for learning are performed. A medical image in which there is no abnormality is an image that does not include a site recognized as medically abnormal in the image, for example, an image obtained by photographing a specific region of a person who does not have a medical abnormality, that is, a normal image. An image obtained by photographing a specific part of the patient is used. The training steps in the present embodiment will be described below.

  In the training step, as described above, the medical image having no abnormality is used as the input image. When the medical image is input, the control device 102 inputs the medical image into an auto encoder prepared in advance, and causes the auto encoder to learn an image pattern of medical image data in which no abnormality exists. Since the auto encoder can train the neural network so that the input and the output are the same, in the present embodiment, in the training step, the image pattern of the medical image having no abnormality in the auto encoder is learned. By doing so, when an image is input in the estimation step described later, it is possible to output an output image without any abnormality. It should be noted that since the auto encoder is a known technique, detailed description thereof will be omitted in the present embodiment, but here, for example, a deep auto encoder is assumed.

  FIG. 2 is a diagram schematically showing a flow of processing by an auto encoder (Auto-encoder). When the input image Xh is input to the auto encoder, the encoder Enc converts the high-dimensional input image Xh into a low-dimensional latent representation z, and the decoder Dec converts the low-dimensional latent expression z into a high-dimensional reconstruction. Convert to image X'h.

  The control device 102 repeats the above conversion process while adjusting the weight, which is a parameter used in the calculation of the auto encoder, until the loss L = d (X′h, Xh) becomes equal to or less than a preset threshold value. The control device 102 sets the weight value set when the loss L becomes equal to or less than the preset threshold value as the weight of the auto encoder. As a result, the auto encoder can be made to learn to output an image in which there is no abnormality with respect to the input image.

  The latent expression z converted by the encoder Enc is input to the variance estimator DecV shown in FIG. FIG. 3 is a diagram schematically showing the flow of processing by the variance estimator DecV. The variance estimator DecV transforms the input latent expression z to obtain a restored image X′h, and estimates the variance Vh of the restored image X′h.

  The control device 102 adjusts the dispersion while adjusting the weight, which is a parameter used in the calculation of the dispersion estimator DecV, until the loss Lv = d (Vh, d (X′h, Xh)) becomes equal to or less than a preset threshold value. The estimation process of is repeated. The control device 102 sets the weight value set when the loss Lv becomes equal to or less than the preset threshold value as the weight of the dispersion estimator DecV. As a result, the variance estimator DecV can be trained to output an image in which the variance of an image in which there is no abnormality with respect to the input image is estimated. Note that, as shown in FIG. 3, if the variance of an image in which there is no abnormality can be estimated, the standard deviation can be easily calculated from the variance, and thus the estimated standard deviation Sh can be easily output.

  When the training steps described above are completed, it is possible to input the medical image of the patient to be diagnosed and execute the estimation step for outputting the image in which the abnormal part is extracted from the medical image. That is, since the training step described above needs to be completed in order to execute the estimation step, when the input image Xh is input to the auto encoder shown in FIG. 2 when the estimation step is executed. , Learning has been completed so as to output an image having no abnormality as the restored image X′h, and the variance estimator DecV shown in FIG. Learning is completed so that the variance of is output as the estimated variance Vh.

  For example, as shown in FIG. 4, when a brain image of a patient having an abnormal part 4a is input to the input image Xa as an input to the auto encoder, an image having no abnormal part is output as a restored image X'a. Therefore, if the difference between the input image Xa and the restored image X'a is calculated, the difference image Da in which the abnormal portion 4a is extracted can be generated. In the present embodiment, the difference image Da is called a difference map Da.

  The abnormal portion 4a can be estimated from the brain image of the patient by using the difference map Da, but there is a difference between the input image Xa and the restored image X'a other than the abnormal portion 4a. Therefore, in the difference map Da, the difference portion other than the abnormal portion 4a which is the characteristic difference portion included in the input image Xa is also extracted. For example, in the brain image data, the shape of the normal part differs depending on the person, and the difference in the shape appears as a difference on the difference map Da. Also in FIG. 4, differences appear on the difference map Da in addition to the abnormal part 4a. Therefore, in the estimation step according to the present embodiment, the estimated standard deviation Sa obtained based on the output of the variance estimator DecV is used to reduce the difference portion other than the characteristic difference portion included in the input image Xa. Process.

FIG. 5 is a diagram schematically showing a method for reducing the difference portion other than the abnormal portion 4a using the difference map Da and the estimated standard deviation Sa. As shown in FIG. 5, the control device 102 calculates the difference map Da and the estimated standard deviation Sa by using the calculation formula shown in the following formula (1) to obtain a difference portion other than the abnormal portion 4a appearing on the image. Can be reduced. In FIG. 5, a standard score map ZSa is obtained as a diagnostic result image showing the result.
ZSa = Da / Sa (1)

  The control device 102 outputs the standard score map ZSa obtained by the calculation to the display device 104 and displays it. Accordingly, the operator of the image processing apparatus 100 captures and inputs a medical image of a patient, for example, a brain image, into the image processing apparatus 100, and a difference portion from a brain image in a normal state without an abnormal portion is determined as an abnormal portion. Since the extracted standard score map can be confirmed on the display device 104, it is possible to diagnose whether the brain of the patient has an abnormality.

  FIG. 6 is a diagram schematically showing the flow of the training step and the estimation step described above in one. The processing flow shown in FIG. 6 is the same as the content described with reference to FIGS.

  FIG. 7 is a flowchart showing the flow of training steps in this embodiment. The processing illustrated in FIG. 7 is executed by the control device 102 as a program that is started when a medical image in a normal state with no abnormal parts, that is, a brain image of a healthy person is input.

  In step S10, the control device 102 inputs the input image Xh to the auto encoder, and the encoder Enc converts the high-dimensional input image Xh into a low-dimensional latent representation z. Then, it progresses to step S20.

  In step S20, the low-dimensional latent representation z is converted into the high-dimensional restored image X'h by the decoder Dec. Then, it progresses to step S30.

  In step S30, the control device 102 calculates the above-mentioned loss L = d (X'h, Xh). Then, it progresses to step S40.

  In step S40, the control device 102 inputs the latent expression z converted by the encoder Enc to the variance estimator DecV, and the variance estimator DecV calculates the variance of the restored image X′h obtained by transforming the latent expression z. The estimated estimated variance Vh is calculated. Then, it progresses to step S50.

  In step S50, the control device 102 calculates the above-mentioned loss Lv = d (Vh, d (X'h, Xh)). Then, it progresses to step S60.

  In step S60, the control device 102 determines whether the loss L and the loss Lv are equal to or less than a preset threshold value. If an affirmative decision is made in step S60, the weighting at that time is adopted and the processing ends. On the other hand, if a negative decision is made in step S60, the operation proceeds to step S70.

  In step S70, the control device 102 adjusts the weights of the auto encoder and the variance estimator as described above, and returns to step S10.

  FIG. 8 is a flowchart showing the flow of the estimation step in this embodiment. The processing illustrated in FIG. 8 is executed by the control device 102 as a program that is activated when a medical image of a diagnosis target patient is input.

  In step S110, the control device 102 inputs the input image Xa into an auto encoder, and the encoder Enc converts the high-dimensional input image Xa into a low-dimensional latent representation z. Then, it progresses to step S120.

  In step S120, the low-dimensional latent representation z is converted into the high-dimensional restored image X'a by the decoder Dec. Then, it progresses to step S130.

  In step S130, the control device 102 takes the difference between the input image Xa and the restored image X′a to generate the difference map Da, as described above. Then, it progresses to step S140.

  In step S140, the control device 102 inputs the latent expression z converted by the encoder Enc to the variance estimator DecV, and the variance estimator DecV estimates the estimated variance of the restored image X′a obtained by transforming the latent expression z. Calculate Va. Then, it progresses to step S150.

  In step S150, the control device 102 calculates the estimated standard deviation Sa based on the calculated estimated variance Va. Then, it progresses to step S160.

  In step S160, the control device 102 calculates the standard score map ZSa by the formula (1). Then, it progresses to step S170.

  In step S170, the control device 102 outputs and displays the standard score map ZSa obtained by the calculation on the display device 104. Then, the process ends.

According to the embodiment described above, the following operational effects can be obtained.
(1) The control device 102 receives an input of a medical image of a person to be diagnosed as an input image, learns an image pattern of the medical image having no abnormality, and outputs the medical image having no abnormality to the input medical image. Input an input image to an auto encoder configured to restore and output an image, obtain a restored image by the auto encoder, calculate a difference between the input image and the restored image, and generate a difference map. did. The control device 102 inputs the input image to a variance estimator configured to train the variance of the medical image having no abnormality and output the variance of the image having no abnormality with respect to the input medical image. , The estimated variance by the variance estimator is obtained, the standard deviation is calculated based on the obtained estimated variance, and the difference part between the input image and the restored image is extracted based on the difference map and the standard deviation and included in the input image. The diagnostic result image in which the difference portion other than the characteristic difference portion is reduced is generated. This can assist a doctor or the like in identifying an abnormal site not found in a normal patient from the input medical image and making a diagnosis. Further, although it is possible to extract the abnormal portion with the difference map obtained by calculating the difference between the input image and the restored image, it is possible to generate the diagnosis result image in which the standard deviation is also added by the above-mentioned formula (1). Therefore, it is possible to reduce the influence of the normal site that appears on the image and that is different for each patient, and to generate the diagnostic result image in which the site that is likely to be the abnormal site is clarified.

(2) The control device 102 displays the generated diagnostic result image on the display device 104. As a result, an operator such as a doctor can confirm the diagnosis result image on the display device 104.

(3) The automatic encoder repeats the process while changing the value of the parameter used in the calculation until the difference between the input medical image and the output image reaches a predetermined threshold value, thereby obtaining a medical image having no abnormality. I tried to learn the image pattern. As a result, the learning accuracy of the auto encoder can be improved, and a medical image in a normal state can be output as a restored image even when various images are input.

-Modification-
The image processing system according to the above-described embodiment can be modified as follows.
(1) In the above-described embodiment, the example in which the image processing apparatus 100 is a personal computer and the control apparatus 102 executes the above-described processing has been described. However, the terminal operated by the operator is allowed to connect to the image processing apparatus 100 via a communication line such as the Internet, and the above-described input images Xh and Xa are input by being transmitted to the image processing apparatus 100 from the operation terminal. May be done. Further, in that case, the standard score map ZSa displaying the abnormal part on the image of the diagnosis target person may be transmitted to the operation terminal and displayed on the operation terminal. Thus, by connecting the operation terminal and the image processing apparatus 100 via the communication line, a client-server type or cloud type image processing system can be constructed, while the image processing apparatus 100 is used alone. It can also be used as a stand-alone device.

(2) In the above-described embodiment, the processing for the training step is described as an example in which the control device 102 executes the image processing device 100. However, the processing for the training step may be executed by another device, and the data obtained by the training step may be recorded in the storage medium 103. In this case, the processing for the training step in the image processing apparatus 100 becomes unnecessary.

  The present invention is not limited to the configurations of the above-described embodiments as long as the characteristic functions of the present invention are not impaired. Further, a configuration in which the above-described embodiment and a plurality of modified examples are combined may be adopted.

100 image processing device 101 operation member 102 control device 103 storage medium 104 display device

An image processing apparatus according to the present invention is an auto encoder configured to learn an image pattern of a medical image having no abnormality and restore and output a medical image having no abnormality with respect to an input medical image. Input the medical image of the diagnosis target person, and the variance estimator configured to learn the variance of the medical image without abnormality and output the variance of the image without abnormality with respect to the input medical image An image input receiving unit that receives an input as an image, a difference between an input image input by the image input receiving unit and a restored image output by the auto encoder as a result of inputting the input image is calculated to generate a difference map. Difference map generating means, standard deviation calculating means for calculating the standard deviation based on the estimated variance output from the variance estimator as a result of inputting the input image, and the difference map Based on the difference map generated by the generation means and the standard deviation calculated by the standard deviation calculation means, the difference portion between the input image and the restored image is extracted, and the difference portion other than the characteristic difference portion included in the input image is extracted. And a diagnostic result image generating means for generating a diagnostic result image with reduced noise.
An image processing system according to the present invention includes an automatic encoder configured to learn an image pattern of a medical image having no abnormality and restore and output a medical image having no abnormality with respect to an input medical image. Input the medical image of the diagnosis target person, and the variance estimator configured to learn the variance of the medical image without abnormality and output the variance of the image without abnormality with respect to the input medical image An image input receiving unit that receives an input as an image, a difference between an input image input by the image input receiving unit and a restored image output by the auto encoder as a result of inputting the input image is calculated to generate a difference map. Difference map generating means, standard deviation calculating means for calculating standard deviation based on the estimated variance output from the variance estimator as a result of inputting the input image, and the difference Based on the standard deviation calculated by the standard deviation calculating unit and the difference map generated by the step-up generating unit, the difference portion between the input image and the restored image is extracted, and the difference portion other than the characteristic difference portion included in the input image is extracted. Diagnostic result image generating means for generating a diagnostic result image in which the difference is reduced.
An image processing program according to the present invention, an image input acceptance procedure for accepting an input of a medical image of a person to be diagnosed as an input image, and an image pattern of a medical image having no abnormality is learned, and an input medical image is input. Input the input image that was accepted in the image input acceptance procedure to the auto encoder configured to restore and output a medical image without abnormality, and obtain the restored image by the auto encoder. A difference map generation procedure for generating a difference map by calculating the difference between the input image received in the input reception procedure and the restored image acquired in the restored image acquisition procedure, and learning the variance of the medical image having no abnormality, input image the input of which is accepted by the image input accepting procedure configured dispersed estimator to output the variance of abnormal there is no image on the input medical image , The estimated variance acquisition procedure for obtaining the estimated variance by the variance estimator, the standard deviation calculation procedure for calculating the standard deviation based on the estimated variance acquired by the estimated variance acquisition procedure, and the difference generated by the difference map generation procedure. Based on the map and the standard deviation calculated by the standard deviation calculation procedure, the difference part between the input image and the restored image is extracted, and the diagnostic result image is obtained by reducing the difference part other than the characteristic difference part included in the input image. It is a program for causing a computer to execute the diagnostic result image generation procedure to be generated.

Claims (9)

An auto encoder configured to learn an image pattern of a medical image having no abnormality and restore and output a medical image having no abnormality with respect to an input medical image,
A variance estimator configured to train the variance of the medical image having no abnormality and output the variance of the image having no abnormality with respect to the input medical image,
Image input receiving means for receiving input as a medical image of the person to be diagnosed as an input image,
A difference map generating unit that calculates a difference between the input image input by the image input receiving unit and a restored image output by the auto encoder as a result of inputting the input image;
Standard deviation calculating means for calculating a standard deviation based on the estimated variance output from the variance estimator as a result of inputting the input image;
Based on the difference map generated by the difference map generating means and the standard deviation calculated by the standard deviation calculating means, a difference portion between the input image and the restored image is extracted and included in the input image. An image processing apparatus, comprising: a diagnostic result image generating unit that generates a diagnostic result image in which a difference portion other than a characteristic difference portion is reduced. The image processing apparatus according to claim 1,
The image processing apparatus further comprising image display means for displaying the diagnostic result image generated by the diagnostic result image generating means on a display device. The image processing device according to claim 1,
In the auto encoder, by repeating the process while changing the value of the parameter used in the calculation until the difference between the input medical image and the output image reaches a predetermined threshold value, an image pattern of a medical image having no abnormality An image processing apparatus characterized by learning. An auto encoder configured to learn an image pattern of a medical image having no abnormality and restore and output a medical image having no abnormality with respect to an input medical image,
A variance estimator configured to train the variance of the medical image having no abnormality and output the variance of the image having no abnormality with respect to the input medical image,
Image input receiving means for receiving input as a medical image of the person to be diagnosed as an input image,
A difference map generating unit that calculates a difference between the input image input by the image input receiving unit and a restored image output by the auto encoder as a result of inputting the input image;
Standard deviation calculating means for calculating a standard deviation based on the estimated variance output from the variance estimator as a result of inputting the input image;
Based on the difference map generated by the difference map generating means and the standard deviation calculated by the standard deviation calculating means, a difference portion between the input image and the restored image is extracted and included in the input image. An image processing system, comprising: a diagnostic result image generating unit that generates a diagnostic result image in which a difference portion other than a characteristic difference portion is reduced. The image processing system according to claim 4,
The image processing system further comprising image display means for displaying the diagnostic result image generated by the diagnostic result image generating means on a display device. The image processing system according to claim 4 or 5,
In the auto encoder, by repeating the process while changing the value of the parameter used in the calculation until the difference between the input medical image and the output image reaches a predetermined threshold value, an image pattern of a medical image having no abnormality An image processing system characterized by learning. An image input acceptance procedure for accepting input of a medical image of a person to be diagnosed as an input image,
Accept the input in the image input procedure to the auto encoder configured to learn the image pattern of the medical image without abnormality and restore and output the medical image without abnormality with respect to the input medical image And a restored image acquisition procedure for inputting the input image to obtain a restored image by the auto encoder,
A difference map generation procedure for calculating a difference between the input image received in the image input procedure and the restored image acquired in the restored image acquisition procedure to generate a difference map;
The input accepting the input in the image input procedure to a variance estimator configured to learn the variance of the medical image having no abnormality and output the variance of the image having no abnormality with respect to the input medical image. An estimated variance acquisition procedure for obtaining an estimated variance by the variance estimator by inputting an image,
A standard deviation calculation procedure for calculating a standard deviation based on the estimated variance acquired in the estimated variance acquisition procedure,
Based on the difference map generated in the difference map generation procedure and the standard deviation calculated in the standard deviation calculation procedure, a difference portion between the input image and the restored image is extracted, and a characteristic included in the input image An image processing program for causing a computer to execute a diagnostic result image generating procedure for generating a diagnostic result image in which a different portion other than the different portion is reduced. The image processing program according to claim 7,
An image processing program, further comprising an image display procedure for displaying the diagnostic result image generated in the diagnostic result image generating procedure on a display device. The image processing program according to claim 7,
In the auto encoder, by repeating the process while changing the value of the parameter used in the calculation until the difference between the input medical image and the output image reaches a predetermined threshold value, an image pattern of a medical image having no abnormality An image processing program characterized by learning.

Images