WO2021184576A1

WO2021184576A1 - Medical image generation method and apparatus, electronic device and medium

Info

Publication number: WO2021184576A1
Application number: PCT/CN2020/098947
Authority: WO
Inventors: 王俊; 张丞; 高鹏; 谢国彤
Original assignee: 平安科技（深圳）有限公司
Priority date: 2020-03-17
Filing date: 2020-06-29
Publication date: 2021-09-23
Also published as: CN111476760B; CN111476760A

Abstract

Disclosed is a medical image generation method, relating to artificial intelligence and wise information technology of med. Said method comprises: acquiring an original medical image and pre-processing same to obtain a standard medical image; according to the standard medical image, generating a first sample image set by using a medical image generation model; and performing valid information quantity calculation on the first sample image set, and selecting first sample images corresponding to K valid information quantities, so as to obtain a valid image set, inputting, for determination, the valid image set and the standard medical image into an image determination model having a relationship of mutual constraint with the image generation model, so as to obtain a determination result, adjusting parameters of the image determination model according to the determination result, so as to obtain a standard image generation model, and inputting the original medical image into the standard image generation model to generate a final medical image. The problem of medical image model training consuming a lot of human resources can be solved.

Description

Medical image generation method, device, electronic equipment and medium

This application claims the priority of a Chinese patent application filed with the Chinese Patent Office on March 17, 2020, the application number is CN 202010186679.3, and the invention title is "Medical Image Generation Method, Apparatus, Electronic Equipment, and Medium". The entire content of the application is approved The reference is incorporated in this application.

Technical field

This application relates to the field of artificial intelligence technology, and in particular to a method, device, electronic device, and computer-readable storage medium for generating medical images.

Background technique

With the development of science and technology, the level of medical care is getting higher and higher. In the field of medical and health, medical images are of great significance for studying disease conditions, predicting diseases and developing medical technology. However, the number of medical images containing enough medical information is small and the demand is great. Therefore, how to train an efficient and accurate medical image generation model to generate more medical images is becoming more and more important.

The inventor realizes that the current generation of medical images needs to rely on medical experts with a high level of professional knowledge to manually screen a large number of samples, and then input them into a pre-built model for training, resulting in a lot of waste of human resources.

Summary of the invention

This application provides a method, device, electronic device, and computer-readable storage medium for generating medical images.

A method for generating medical images provided by this application includes:

Acquiring an original medical image, performing conversion processing on the original medical image, to obtain an initial medical image;

Performing cell enhancement processing on the initial medical image to obtain a standard medical image;

Acquiring distribution data of each pixel information in the standard medical image, and generating a plurality of first sample images similar to the standard medical image by using an image generation model according to the distribution data to obtain a first sample image set;

The effective information amount calculation is performed on the first sample image set to obtain the effective information amount of each first sample image in the first sample image set. Select K first sample images corresponding to the effective amount of information from the sample image set to obtain an effective image set;

The effective image set and the standard medical image are input into an image discrimination model that has a mutual constraint relationship with the image generation model for image discrimination, and the discrimination result is obtained, and the parameters of the image discrimination model are adjusted according to the discrimination result , Until the discrimination result meets the preset requirements, obtain the parameters of the image discrimination model at this time, and obtain the standard image generation model according to the parameters;

The original medical image is input to the standard image generation model, a second sample image set is generated, and a final medical image is generated according to the second sample image.

This application also provides a device for generating medical images, which includes:

The image preprocessing module is used to obtain an original medical image, perform conversion processing on the original medical image to obtain an initial medical image, and perform cell enhancement processing on the initial medical image to obtain a standard medical image;

The first sample image generation module is used to obtain the distribution data of each pixel information in the standard medical image, and use an image generation model to generate multiple first sample images similar to the standard medical image according to the distribution data, Obtain the first sample image set;

The effective information calculation module is used to calculate the effective information amount of the first sample image set to obtain the effective information amount of each first sample image in the first sample image set, according to the effective information amount The arrangement sequence selects K first sample images corresponding to the effective amount of information from the first sample image set to obtain an effective image set;

The image discrimination module is used to input the effective image set and the standard medical image into an image discrimination model that has a mutual constraint relationship with the image generation model for image discrimination, obtain the discrimination result, and adjust the image according to the discrimination result. The parameters of the image discrimination model, until the discrimination result meets the preset requirements, the parameters of the image discrimination model at this time are obtained, and the standard image generation model is obtained according to the parameters; the original medical image is input to the standard The image generation model generates a second sample image set, and generates a final medical image according to the second sample image.

This application also provides an electronic device, which includes:

Memory, storing at least one instruction; and

The processor executes the instructions stored in the memory to implement the following steps:

The present application also provides a computer-readable storage medium in which at least one instruction is stored, and the at least one instruction is executed by a processor in an electronic device to implement the following steps:

Description of the drawings

FIG. 1 is a schematic flowchart of a method for generating a medical image provided by an embodiment of the application;

2 is a schematic diagram of modules of a method for generating medical images provided by an embodiment of the application;

3 is a schematic diagram of the internal structure of an electronic device of a method for generating a medical image provided by an embodiment of the application;

The realization, functional characteristics, and advantages of the purpose of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings.

Detailed ways

It should be understood that the specific embodiments described here are only used to explain the present application, and are not used to limit the present application.

This application provides a method for generating medical images. Referring to FIG. 1, it is a schematic flowchart of a method for generating a medical image provided by an embodiment of this application. The method can be executed by a device, and the device can be implemented by software and/or hardware.

In this embodiment, the method for generating medical images includes:

S1. Obtain an original medical image, perform conversion processing on the original medical image, and obtain an initial medical image.

In the embodiment of the present application, the original medical image may be a b-ultrasound image, a color ultrasound image, etc. stored in a hospital.

In the embodiment of the present application, the conversion processing of the original medical image to obtain the initial medical image includes:

Converting the original medical image to a gray value to obtain an original gray image;

Performing noise reduction processing on the original gray image to obtain a noise reduction gray image;

Performing geometric transformation processing on the noise reduction grayscale image to obtain a transformed grayscale image;

The contrast enhancement of the transformed grayscale image is performed to obtain the initial medical image.

Wherein, said converting the original medical image to gray value to obtain the original gray image includes:

All pixels in the original medical image are input into a gray value conversion formula for gray value conversion, and the original gray image is generated according to the converted gray value.

Wherein, the gray value conversion formula is:

Gray＝0.30*R+0.59*G+0.11*B

Wherein R, G, B are the three components of the pixels in the original medical image, and Gray is the gray value obtained by conversion.

Further, in the embodiment of the present application, the performing noise reduction processing on the original gray image to obtain the noise reduction gray image includes:

Replace the pixel value of any pixel in the original grayscale image with the median value of each pixel in a neighborhood of the pixel, so that the pixel value around any pixel is close to the true value , Thereby eliminating isolated noise points.

In detail, the neighborhood may be a two-dimensional sliding template with a preset circular structure, and the pixels in the two-dimensional sliding template are sorted according to the size of the pixel value to generate a monotonically rising (or falling) two-dimensional data sequence , To find the median value of the pixel value of each pixel in the neighborhood.

In detail, the embodiment of the present application uses the following calculation formula to perform noise reduction processing on the original gray image to obtain the noise reduction gray image:

g(x,y)=med{f(x-j,y-k),(j,k∈W)}

Wherein, f(x,y) is the original grayscale image; g(x,y) is the denoising grayscale image, W is a two-dimensional sliding template; j and k are on the boundary of the two-dimensional sliding template The coordinates of the pixel; med is the noise reduction processing operation.

Further, performing geometric transformation processing on the noise reduction grayscale image to obtain a transformed grayscale image includes:

Perform geometric transformation processing such as translation, transposition, mirroring, rotation, and scaling on the denoising grayscale image to correct the system error generated during the acquisition of the original medical image and the random error generated by the position of the instrument; after the geometric transformation processing is completed , To obtain the transformed gray scale image.

The contrast refers to the contrast between the maximum and minimum brightness of pixels in an image.

In this embodiment of the present application, a contrast stretching method may be used to enhance the contrast of the transformed grayscale image.

The contrast stretching method is also called gray scale stretching. The embodiment of the application uses the piecewise linear transformation function in the contrast stretching method to perform gray-scale stretching for specific regions in the original gray-scale image according to actual needs, thereby enhancing the contrast of the transformed gray-scale image, and obtaining the original medical image. In detail, performing contrast enhancement on the transformed grayscale image to obtain an initial medical image includes:

Use the following piecewise linear transformation function formula to perform contrast enhancement on the transformed grayscale image to obtain the initial medical image:

D _b =f(D _a )=a*D _a +b

Where a is the linear slope, b is _{the intercept of D b} on the Y axis, D _a represents the gray value of the input transformed gray image, and D _b represents the gray value of the output initial medical image.

S2. Perform cell enhancement processing on the initial medical image to obtain a standard medical image.

Taking into account that the cells appear in a linear form, the embodiment of the present application can obtain the standard medical image with different standard offset values by taking different values for the preset standard offset of the Gaussian function. Linear enhancement filtering of cells.

The linear enhancement filtering can be used to perform cell enhancement on the initial medical image.

According to the convolution property of the Gaussian function, the embodiment of the application calculates the convolution of the initial medical image and the second-order Gaussian function to obtain the scale space derivative I _{abc of} the standard medical image:

Where I is the initial medical image; G(x, y, z) is a Gaussian function; x, y, z are the parameters of the Gaussian function; σ is the standard offset of the Gaussian function;

Symbol for reciprocal operation

To find the convolution operation symbol.

Further, in the embodiment of the present application, the matrix H is obtained by taking different values of the standard offset of the Gaussian function:

Wherein, the elements in the matrix are the values of the scale space derivative I _abc under different standard offset values.

It can be known from the linear characteristics of the Gaussian function that if and only when the standard offset σ of the Gaussian function is exactly equal to the actual width of the cell, the best linear enhancement filter can be obtained, and the best linearity can be used. The enhancement filtering performs cell enhancement on the initial medical image.

In summary, in the embodiment of the present application, the initial medical image is calculated as described above, the size of the standard offset is adjusted to be equal to the actual width of the cells in the initial medical image, and the optimal linear enhancement filter is obtained. Furthermore, cell enhancement is performed on the initial medical image to obtain the standard medical image.

S3. Obtain distribution data of each pixel information in the standard medical image, and generate a plurality of first sample images similar to the standard medical image by using an image generation model according to the distribution data to obtain a first sample image set.

In the embodiment of the present application, the existing munpy (Numerical Python, digital python) method can be used to obtain the distribution data of each pixel information in the standard medical image.

In the embodiment of the present application, the image generation model is a sample convolutional neural network constructed with noise that obeys a specific distribution.

The image generation model may use a preset image generation model to generate a plurality of first sample images with the standard medical image according to the distribution data to obtain a first sample image set.

Specifically, the S3 includes:

Construct sample generation loss function F;

Inputting the distribution data of each pixel information in the standard medical image to the image generation model to generate an initial first sample image set;

Inputting the initial first sample image set and the standard medical image to the sample generating loss function F for loss calculation, to obtain a loss function value p;

When the loss function value p is greater than or equal to the preset loss threshold m, adjust the parameters of the image generation model, and regenerate the first sample image;

When the loss function value p is less than the loss threshold m, the first sample image set is obtained.

Wherein, the sample generation loss function F includes:

F=Lc-Ls

Lc=E[logP(C|Xreal)]+E[logP(C|Xfake)]

Ls=E[logP(S|Xreal)]+E[logP(S|Xfake)]

Wherein, E[] is the expected value operation; Lc is the expected value of the similarity between the first sample image and the standard medical image; Ls is the expected value of the effective information in the standard medical image; Xreal is the expected value Standard medical image; Xfake is the first sample image; C is the effective amount of information in the first sample image; S is the effective amount of information in the standard medical image.

In this embodiment of the application, the image generation model is used to generate multiple first sample images to obtain the first sample image set.

In detail, only part of each pixel information of the standard medical image contains effective information helpful to medical treatment, so only part of each pixel information of the generated first sample image also contains the effective information. Therefore, the effective information contained in the first sample image set generated by S3 is still insufficient to meet the requirements of medical research, and the following S4 needs to be further performed for screening.

S4. Calculate the effective information amount of the first sample image set to obtain the effective information amount of each first sample image in the first sample image set, according to the order of the effective information amount from the The K first sample images corresponding to the effective amount of information are selected from the first sample image set to obtain the effective image set.

Specifically, in the embodiment of the present application, the following effective information amount calculation formula is used to calculate the effective information amount R of each first sample image in the first sample image set:

Wherein, b is the number of pixels containing valid information in the first sample image; a is the total number of pixels in the first sample image.

The number of pixels containing valid information in the first sample image can be recognized and acquired by using existing image recognition technology.

Further, after the calculation is completed, the embodiment of the present application arranges the effective amount of information in the first sample image set in order from most to least, and selects the k with the most effective amount of information from the arrangement. The first sample image, the effective image set is obtained.

S5. Input the effective image set and the standard medical image into an image discrimination model that has a mutual constraint relationship with the image generation model for image discrimination, obtain a discrimination result, and adjust the image discrimination model according to the discrimination result Until the discrimination result meets the preset requirements, the parameters of the image discrimination model at this time are obtained, and the standard image generation model is obtained according to the parameters; the original medical image is input to the standard image generation model, A second sample image set is generated, and a final medical image is generated according to the second sample image.

In detail, the image discrimination model is a convolutional neural network for image discrimination.

The mutual constraint relationship means that the parameters in the loss function of the image generation model and the image discrimination model are the same and change synchronously. For example, if one of the parameters of the image discrimination model becomes "a", then the image generation model The corresponding parameter also becomes "a".

Specifically, the effective image set and the standard medical image are input into an image discrimination model that has a mutual constraint relationship with the image generation model for image discrimination, and the discrimination result is obtained, and the discrimination result is adjusted according to the discrimination result. The parameters of the image discrimination model, until the discrimination result meets the preset requirements, the parameters of the image discrimination model at this time are obtained, including:

Construct a discriminant loss function Y, and use the discriminant loss function Y to constrain the image discriminant model;

Inputting the standard medical image and the effective image set to the image discrimination model to generate the final medical image set;

Input the final medical image set and the effective image set to the discriminant loss function Y for loss calculation to obtain a loss function value q;

When the loss function value q is greater than or equal to the preset loss threshold n, it means that the final medical image set generated by the image discriminant model is not similar to the effective image set, and the parameters of the image discriminant model are adjusted to regenerate the image. State the final medical image;

When the loss function value q is less than the loss threshold n, it means that the final medical image set generated by the image discrimination model is similar to the effective image set, and training is completed to obtain the image discrimination model parameters at this time.

Wherein, the discriminant loss function Y is as follows:

Y=Lc+Ls

Lc=E[logP(C|Xreal)]+E[logP(C|Xfake)]

Ls=E[logP(S|Xreal)]+E[logP(S|Xfake)]

Further, the original medical image is input to the standard image generation model, a second sample image set is generated, and a final medical image is generated according to the second sample image. According to the distribution data of each pixel information in the standard medical image, the above-mentioned embodiment of the application uses the image generation model to generate multiple first sample images similar to the standard medical image, and realizes the preliminary training of the image generation model. The first sample image can be generated for subsequent use; further, the effective information amount calculation is performed on the first sample image set, and K effective information is selected from the first sample image set according to the order of the effective information amount. The first sample image corresponding to the amount of information obtains an effective image set to filter out medical images containing more effective information, which further guarantees the quality of subsequent image generation, and also avoids manual screening of samples one by one; The image set and the standard medical image are input into the image discrimination model that has a mutual constraint relationship with the image generation model for image discrimination, and the discrimination result is obtained. The parameters of the image discrimination model are adjusted according to the discrimination result, and the image discrimination is used The constraint relationship between the model and the image generation model further adjusts the parameters of the image generation model to realize the retraining of the image generation model and ensure the accuracy of the image generation model during image generation. Therefore, the medical image generation method, device, and computer-readable storage medium proposed in this application can automatically screen the first sample images in batches, and automatically train accurate medical image models to generate the final medical images. Save a lot of human resources.

As shown in Figure 2, it is a functional block diagram of the medical image generation method device of the present application.

The medical image generation method 100 described in this application can be installed in an electronic device. According to the realized functions, the medical image generation method device may include an image preprocessing module 101, a first sample image generation module 102, an effective information calculation module 103, and an image discrimination module 104. The module described in the present invention can also be called a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can complete fixed functions, and are stored in the memory of the electronic device.

In this embodiment, the functions of each module/unit are as follows:

The image preprocessing module 101 is configured to obtain an original medical image, perform conversion processing on the original medical image to obtain an initial medical image, and perform cell enhancement processing on the initial medical image to obtain a standard medical image;

The first sample image generation module 102 is used to obtain distribution data of each pixel information in the standard medical image, and generate multiple first images similar to the standard medical image by using an image generation model according to the distribution data. This image, get the first sample image set;

The effective information calculation module 103 is configured to calculate the effective information amount of the first sample image set to obtain the effective information amount of each first sample image in the first sample image set, according to the effective information amount. The arrangement order of the information amount selects K first sample images corresponding to the effective information amount from the first sample image set to obtain an effective image set;

The image discrimination module 104 is configured to input the effective image set and the standard medical image into an image discrimination model that has a mutual constraint relationship with the image generation model for image discrimination, and obtain a discrimination result, according to the discrimination As a result, the parameters of the image discrimination model are adjusted until the discrimination result meets the preset requirements, the parameters of the image discrimination model at this time are obtained, and the standard image generation model is obtained according to the parameters.

In detail, the specific implementation steps of each module of the medical image generation method device are as follows:

The image preprocessing module 101 obtains an original medical image, performs conversion processing on the original medical image to obtain an original medical image, and performs cell enhancement processing on the original medical image to obtain a standard medical image.

In the embodiment of the present application, the image preprocessing module 101 performs conversion processing on the original medical image to obtain an initial medical image, including:

Wherein, the image preprocessing module 101 converts the gray value of the original medical image to obtain the original gray image, including:

Wherein, the gray value conversion formula is:

Gray＝0.30*R+0.59*G+0.11*B

Further, in the embodiment of the present application, the image preprocessing module 101 performs noise reduction processing on the original gray image to obtain a noise reduction gray image, including:

In detail, the image preprocessing module 101 uses the following calculation formula to perform noise reduction processing on the original gray image to obtain the noise reduction gray image:

g(x,y)=med{f(x-j,y-k),(j,k∈W)}

Further, the image preprocessing module 101 performs geometric transformation processing on the noise reduction grayscale image to obtain a transformed grayscale image, including:

The image preprocessing module 101 may adopt a contrast stretching method to enhance the contrast of the transformed grayscale image.

The contrast stretching method is also called gray scale stretching. The image preprocessing module 101 of the embodiment of the present application uses the piecewise linear transformation function in the contrast stretching method to perform grayscale stretching for a specific region in the original grayscale image according to actual needs, thereby enhancing the transformed grayscale The contrast of the figure, the initial medical image is obtained. In detail, performing contrast enhancement on the transformed grayscale image to obtain an initial medical image includes:

The image preprocessing module 101 uses the following piecewise linear transformation function formula to perform contrast enhancement on the transformed grayscale image to obtain the initial medical image:

D _b =f(D _a )=a*D _a +b

Considering that the cells appear in a linear form, the image preprocessing module 101 described in this embodiment of the application can obtain different standard deviations of the standard medical image by taking different values for the standard deviation of the preset Gaussian function. The linear enhancement filter applied to the cell under the measurement value.

According to the convolution property of the Gaussian function, the image preprocessing module 101 of the embodiment of the present application calculates the convolution of the initial medical image and the second-order Gaussian function to obtain the scale space derivative I _{abc of} the standard medical image:

Symbol for reciprocal operation

To find the convolution operation symbol.

Further, the image preprocessing module 101 of the embodiment of the present application obtains the matrix H by taking different values of the standard offset of the Gaussian function:

In summary, the image preprocessing module 101 of the embodiment of the present application calculates the initial medical image as described above, adjusts the size of the standard offset to be equal to the actual width of the cells in the initial medical image, and obtains the The best linear enhancement filtering is used to achieve cell enhancement of the initial medical image to obtain the standard medical image.

The first sample image generating module 102 obtains the distribution data of each pixel information in the standard medical image, and uses an image generation model to generate a plurality of first sample images similar to the standard medical image according to the distribution data, Get the first sample image set.

The first sample image generating module 102 described in this embodiment of the present application can use existing munpy (Numerical Python, digital python) methods to obtain the distribution data of each pixel information in the standard medical image.

In detail, the first sample image generation module 102 obtains the distribution data of each pixel information in the standard medical image, and generates a plurality of first images similar to the standard medical image by using an image generation model according to the distribution data. In this image, the first sample image set obtained includes:

Construct sample generation loss function F;

Wherein, the sample generation loss function F includes:

F=Lc-Ls

Lc=E[logP(C|Xreal)]+E[logP(C|Xfake)]

Ls=E[logP(S|Xreal)]+E[logP(S|Xfake)]

In detail, only part of each pixel information of the standard medical image contains effective information helpful to medical treatment, so only part of each pixel information of the generated first sample image also contains the effective information. Therefore, the effective information contained in the first sample image set is still insufficient to meet the requirements of medical research, and further screening is required.

The effective information calculation module 103 calculates the effective information amount of the first sample image set to obtain the effective information amount of each first sample image in the first sample image set, according to the effective information amount The arrangement sequence selects K first sample images corresponding to the effective amount of information from the first sample image set to obtain an effective image set.

In detail, the effective information calculation module 103 described in the embodiment of the present application uses the following effective information amount calculation formula to calculate the effective information amount R of each first sample image in the first sample image set:

Further, after the calculation is completed, the effective information calculation module 103 described in this embodiment of the present application arranges the effective information amount in the first sample image set in order from most to least, and selects effective information from the arrangement The k first sample images with the largest amount are obtained to obtain the effective image set.

The image discrimination module 104 inputs the effective image set and the standard medical image into an image discrimination model that has a mutual constraint relationship with the image generation model for image discrimination, obtains a discrimination result, and adjusts the image according to the discrimination result. The parameters of the image discrimination model, until the discrimination result meets the preset requirements, the parameters of the image discrimination model at this time are obtained, and the standard image generation model is obtained according to the parameters; the original medical image is input to the standard The image generation model generates a second sample image set, and generates a final medical image according to the second sample image.

In detail, the image discrimination module 104 inputs the effective image set and the standard medical image into an image discrimination model that has a mutual constraint relationship with the image generation model for image discrimination, and obtains the discrimination result according to the The discrimination result adjusts the parameters of the image discrimination model until the discrimination result meets the preset requirements, and the parameters of the image discrimination model at this time are obtained, including:

Wherein, the discriminant loss function Y is as follows:

Y=Lc+Ls

Lc=E[logP(C|Xreal)]+E[logP(C|Xfake)]

Ls=E[logP(S|Xreal)]+E[logP(S|Xfake)]

Further, the original medical image is input to the standard image generation model, a second sample image set is generated, and a final medical image is generated according to the second sample image.

As shown in FIG. 3, it is a schematic diagram of the structure of an electronic device that implements the method for generating a medical image in the present application.

The electronic device 1 may include a processor 10, a memory 11, and a bus, and may also include a computer program stored in the memory 11 and running on the processor 10, such as a training program 12 of the medical image model. .

Wherein, the memory 11 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, mobile hard disk, multimedia card, card-type memory (for example: SD or DX memory, etc.), magnetic memory, magnetic disk, CD etc. The memory 11 may be an internal storage unit of the electronic device 1 in some embodiments, for example, a mobile hard disk of the electronic device 1. In other embodiments, the memory 11 may also be an external storage device of the electronic device 1, such as a plug-in mobile hard disk, a smart media card (SMC), and a secure digital (Secure Digital) equipped on the electronic device 1. , SD) card, flash card (Flash Card), etc. Further, the memory 11 may also include both an internal storage unit of the electronic device 1 and an external storage device. The memory 11 can be used not only to store application software and various data installed in the electronic device 1, such as the code of the training program 12 of the medical image model, etc., but also to temporarily store data that has been output or will be output.

The processor 10 may be composed of integrated circuits in some embodiments, for example, may be composed of a single packaged integrated circuit, or may be composed of multiple integrated circuits with the same function or different functions, including one or more Combinations of central processing unit (CPU), microprocessor, digital processing chip, graphics processor, and various control chips, etc. The processor 10 is the control unit of the electronic device, which uses various interfaces and lines to connect the various components of the entire electronic device, and runs or executes programs or modules stored in the memory 11 (such as executing The medical image generation program 12, etc.), and call the data stored in the memory 11 to execute various functions of the electronic device 1 and process data.

The bus may be a peripheral component interconnect standard (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus can be divided into address bus, data bus, control bus and so on. The bus is configured to implement connection and communication between the memory 11 and at least one processor 10 and the like.

FIG. 3 only shows an electronic device with components. Those skilled in the art can understand that the structure shown in FIG. 3 does not constitute a limitation on the electronic device 1, and may include fewer or more components than shown in the figure. Components, or a combination of certain components, or different component arrangements.

For example, although not shown, the electronic device 1 may also include a power source (such as a battery) for supplying power to various components. Preferably, the power source may be logically connected to the at least one processor 10 through a power management device, thereby controlling power The device implements functions such as charge management, discharge management, and power consumption management. The power supply may also include any components such as one or more DC or AC power supplies, recharging devices, power failure detection circuits, power converters or inverters, and power status indicators. The electronic device 1 may also include various sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be repeated here.

Further, the electronic device 1 may also include a network interface. Optionally, the network interface may include a wired interface and/or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is usually used in the electronic device 1 Establish a communication connection with other electronic devices.

Optionally, the electronic device 1 may also include a user interface. The user interface may be a display (Display) and an input unit (such as a keyboard (Keyboard)). Optionally, the user interface may also be a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode, organic light-emitting diode) touch device, etc. Among them, the display can also be appropriately called a display screen or a display unit, which is used to display the information processed in the electronic device 1 and to display a visualized user interface.

It should be understood that the embodiments are only for illustrative purposes, and are not limited by this structure in the scope of the patent application.

The medical image generation method program 12 stored in the memory 11 in the electronic device 1 is a combination of multiple instructions. When running in the processor 10, it can realize:

Specifically, for the specific implementation method of the above-mentioned instructions by the processor 10, reference may be made to the description of the relevant steps in the embodiment corresponding to FIG. 2, which will not be repeated here.

Further, if the integrated module/unit of the electronic device 1 is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer readable storage medium. The computer-readable storage medium may be non-volatile or volatile. The computer-readable storage medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile Hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory).

The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are implemented:

Step 1: Obtain an original medical image, and perform conversion processing on the original medical image to obtain an initial medical image.

Wherein, the gray value conversion formula is:

Gray＝0.30*R+0.59*G+0.11*B

g(x,y)=med{f(x-j,y-k),(j,k∈W)}

D _b =f(D _a )=a*D _a +b

Step 2: Perform cell enhancement processing on the initial medical image to obtain a standard medical image.

Considering that the cells appear in a linear form, the embodiment of the present application takes different values for the standard offset of the preset Gaussian function to obtain the standard medical image suitable for different standard offset values. Linear enhancement filtering of cells.

Symbol for reciprocal operation

To find the convolution operation symbol.

Step 3: Obtain the distribution data of each pixel information in the standard medical image, and use the image generation model to generate multiple first sample images similar to the standard medical image according to the distribution data to obtain a first sample image set .

Specifically, the step three includes:

Construct sample generation loss function F;

Wherein, the sample generation loss function F includes:

F=Lc-Ls

Lc=E[logP(C|Xreal)]+E[logP(C|Xfake)]

Ls=E[logP(S|Xreal)]+E[logP(S|Xfake)]

In detail, only part of each pixel information of the standard medical image contains effective information helpful to medical treatment, so only part of each pixel information of the generated first sample image also contains the effective information. Therefore, the effective information contained in the first sample image set generated in step 3 is still insufficient to meet the requirements of medical research, and the following step 4 needs to be further performed for screening.

Step 4: Calculate the effective information amount of the first sample image set to obtain the effective information amount of each first sample image in the first sample image set. In the first sample image set, K first sample images corresponding to the effective amount of information are selected to obtain an effective image set.

Step 5. Input the effective image set and the standard medical image into an image discrimination model that has a mutual constraint relationship with the image generation model for image discrimination, obtain a discrimination result, and adjust the image discrimination according to the discrimination result The parameters of the model, until the discrimination result meets the preset requirements, the parameters of the image discrimination model at this time are obtained, and the standard image generation model is obtained according to the parameters; the original medical image is input to the standard image generation model To generate a second sample image set, and generate a final medical image according to the second sample image.

Wherein, the discriminant loss function Y is as follows:

Y=Lc+Ls

Lc=E[logP(C|Xreal)]+E[logP(C|Xfake)]

Ls=E[logP(S|Xreal)]+E[logP(S|Xfake)]

Specifically, for the method steps implemented when the computer program is executed by the processor, reference may be made to the description of the relevant steps in the embodiment corresponding to FIG. 2, which will not be repeated here.

In the several embodiments provided in this application, it should be understood that the disclosed equipment, device, and method may be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.

The modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the objectives of the solutions of the embodiments.

In addition, the functional modules in the various embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware, or may be implemented in the form of hardware plus software functional modules.

For those skilled in the art, it is obvious that the present application is not limited to the details of the foregoing exemplary embodiments, and the present application can be implemented in other specific forms without departing from the spirit or basic characteristics of the application.

Therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of this application is defined by the appended claims rather than the above description, and therefore it is intended to fall into the claims. All changes in the meaning and scope of the equivalent elements of are included in this application. Any associated diagram marks in the claims should not be regarded as limiting the claims involved.

In addition, it is obvious that the word "including" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in the system claims can also be implemented by one unit or device through software or hardware. The second class words are used to indicate names, and do not indicate any specific order.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the application and not to limit them. Although the application has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the application can be Make modifications or equivalent replacements without departing from the spirit and scope of the technical solution of the present application.

Claims

A method for generating medical images, wherein the method includes:

Acquiring an original medical image, performing conversion processing on the original medical image, to obtain an initial medical image;

Performing cell enhancement processing on the initial medical image to obtain a standard medical image;

Acquiring distribution data of each pixel information in the standard medical image, and generating a plurality of first sample images similar to the standard medical image by using an image generation model according to the distribution data to obtain a first sample image set;

The effective information amount calculation is performed on the first sample image set to obtain the effective information amount of each first sample image in the first sample image set. Select K first sample images corresponding to the effective amount of information from the sample image set to obtain an effective image set;

The effective image set and the standard medical image are input into an image discrimination model that has a mutual constraint relationship with the image generation model for image discrimination, and the discrimination result is obtained, and the parameters of the image discrimination model are adjusted according to the discrimination result , Until the discrimination result meets the preset requirements, obtain the parameters of the image discrimination model at this time, and obtain the standard image generation model according to the parameters;

The original medical image is input to the standard image generation model, a second sample image set is generated, and a final medical image is generated according to the second sample image.
5. The method for generating medical images according to claim 1, wherein said converting the original medical image to obtain the initial medical image comprises:

Converting the original medical image to a gray value to obtain an original gray image;

Performing noise reduction processing on the original gray image to obtain a noise reduction gray image;

Performing geometric transformation processing on the noise reduction grayscale image to obtain a transformed grayscale image;

The contrast enhancement of the transformed grayscale image is performed to obtain the initial medical image.
The method for generating medical images according to claim 1, wherein said performing cell enhancement processing on the initial medical image to obtain a standard medical image comprises:

The following calculation formula is used to calculate the convolution of the initial medical image and the second-order Gaussian function to obtain the scale space derivative I abc of the initial medical image:

Where I is the initial medical image; G(x, y, z) is a Gaussian function; x, y, z are the parameters of the Gaussian function; σ is the standard offset of the Gaussian function;
Symbol for reciprocal operation
To find the convolution operation symbol;

Perform cell enhancement on the initial medical image according to the scale space derivative I abc to obtain the standard medical image.
The method for generating a medical image according to claim 1, wherein said acquiring distribution data of each pixel information in said standard medical image, and generating a plurality of images similar to said standard medical image using an image generation model according to said distribution data The first sample image of, the first sample image set is obtained, including:

Construct sample generation loss function F;

Inputting the distribution data of each pixel information in the standard medical image to the image generation model to generate an initial first sample image set;

Inputting the initial first sample image set and the standard medical image to the sample generating loss function F for loss calculation, to obtain a loss function value p;

When the loss function value p is greater than or equal to the preset loss threshold m, adjust the parameters of the image generation model, and regenerate the first sample image;

When the loss function value p is less than the loss threshold m, the first sample image set is obtained.
5. The medical image generation method according to claim 4, wherein the sample generation loss function F comprises:

F=Lc-Ls

Lc=E[logP(C|Xreal)]+E[logP(C|Xfake)]

Ls=E[logP(S|Xreal)]+E[logP(S|Xfake)]

Wherein, E[] is the expected value operation; Lc is the expected value of the similarity between the first sample image and the standard medical image; Ls is the expected value of the effective information in the standard medical image; Xreal is the expected value Standard medical image; Xfake is the first sample image; C is the effective amount of information in the first sample image; S is the effective amount of information in the standard medical image.
5. The medical image generation method according to claim 1, wherein said inputting said effective image set and said standard medical image into an image discrimination model which has a mutual constraint relationship with said image generation model for image discrimination, Obtain the discrimination result, adjust the parameters of the image discrimination model according to the discrimination result, until the discrimination result meets the preset requirements, obtain the parameters of the image discrimination model at this time, including:

Construct the discriminative loss function Y;

Inputting the standard medical image and the effective image set to the image discrimination model to generate the final medical image set;

Input the final medical image set and the effective image set to the discriminant loss function Y for loss calculation to obtain a loss function value q;

When the loss function value q is greater than or equal to the preset loss threshold n, adjust the parameters of the image discrimination model to regenerate the final medical image;

When the loss function value q is less than the loss threshold n, the parameters of the image discrimination model at this time are obtained.
8. The method for generating medical images according to claim 6, wherein the discriminant loss function Y comprises:

Y=Lc+Ls

Lc=E[logP(C|Xreal)]+E[logP(C|Xfake)]

Ls=E[logP(S|Xreal)]+E[logP(S|Xfake)]

Wherein, E[] is the expected value operation; Lc is the expected value of the similarity between the first sample image and the standard medical image; Ls is the expected value of the effective information in the standard medical image; Xreal is the expected value Standard medical image; Xfake is the first sample image; C is the effective amount of information in the first sample image; S is the effective amount of information in the standard medical image.
A medical image generation device, wherein the device includes:

The image preprocessing module is used to obtain an original medical image, perform conversion processing on the original medical image to obtain an initial medical image, and perform cell enhancement processing on the initial medical image to obtain a standard medical image;

The first sample image generation module is used to obtain the distribution data of each pixel information in the standard medical image, and use an image generation model to generate multiple first sample images similar to the standard medical image according to the distribution data, Obtain the first sample image set;

The effective information calculation module is used to calculate the effective information amount of the first sample image set to obtain the effective information amount of each first sample image in the first sample image set, according to the effective information amount The arrangement sequence selects K first sample images corresponding to the effective amount of information from the first sample image set to obtain an effective image set;

The image discrimination module is used to input the effective image set and the standard medical image into an image discrimination model that has a mutual constraint relationship with the image generation model for image discrimination, obtain the discrimination result, and adjust the image according to the discrimination result. The parameters of the image discrimination model, until the discrimination result meets the preset requirements, the parameters of the image discrimination model at this time are obtained, and the standard image generation model is obtained according to the parameters; the original medical image is input to the standard The image generation model generates a second sample image set, and generates a final medical image according to the second sample image.
An electronic device, wherein the electronic device includes:

At least one processor; and,

A memory communicatively connected with the at least one processor; wherein,

The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor, so that the at least one processor can execute the following steps:

Acquiring an original medical image, performing conversion processing on the original medical image, to obtain an initial medical image;

Performing cell enhancement processing on the initial medical image to obtain a standard medical image;

Acquiring distribution data of each pixel information in the standard medical image, and generating a plurality of first sample images similar to the standard medical image by using an image generation model according to the distribution data to obtain a first sample image set;

The effective information amount calculation is performed on the first sample image set to obtain the effective information amount of each first sample image in the first sample image set. Select K first sample images corresponding to the effective amount of information from the sample image set to obtain an effective image set;

The effective image set and the standard medical image are input into an image discrimination model that has a mutual constraint relationship with the image generation model for image discrimination, and the discrimination result is obtained, and the parameters of the image discrimination model are adjusted according to the discrimination result , Until the discrimination result meets the preset requirements, obtain the parameters of the image discrimination model at this time, and obtain the standard image generation model according to the parameters;

The original medical image is input to the standard image generation model, a second sample image set is generated, and a final medical image is generated according to the second sample image.
9. The electronic device according to claim 9, wherein said performing conversion processing on said original medical image to obtain an initial medical image comprises:

Converting the original medical image to a gray value to obtain an original gray image;

Performing noise reduction processing on the original gray image to obtain a noise reduction gray image;

Performing geometric transformation processing on the noise reduction grayscale image to obtain a transformed grayscale image;

The contrast enhancement of the transformed grayscale image is performed to obtain the initial medical image.
9. The electronic device according to claim 9, wherein said performing cell enhancement processing on said initial medical image to obtain a standard medical image comprises:

The following calculation formula is used to calculate the convolution of the initial medical image and the second-order Gaussian function to obtain the scale space derivative I abc of the initial medical image:

Where I is the initial medical image; G(x, y, z) is a Gaussian function; x, y, z are the parameters of the Gaussian function; σ is the standard offset of the Gaussian function;
Symbol for reciprocal operation
To find the convolution operation symbol;

Perform cell enhancement on the initial medical image according to the scale space derivative I abc to obtain the standard medical image.
The electronic device according to claim 9, wherein the acquiring distribution data of each pixel information in the standard medical image, and generating a plurality of first images similar to the standard medical image using an image generation model according to the distribution data Sample images to obtain the first sample image set, including:

Construct sample generation loss function F;

Inputting the distribution data of each pixel information in the standard medical image to the image generation model to generate an initial first sample image set;

Inputting the initial first sample image set and the standard medical image to the sample generating loss function F for loss calculation, to obtain a loss function value p;

When the loss function value p is greater than or equal to the preset loss threshold m, adjust the parameters of the image generation model, and regenerate the first sample image;

When the loss function value p is less than the loss threshold m, the first sample image set is obtained.
The electronic device according to claim 12, wherein the sample generation loss function F comprises:

F=Lc-Ls

Lc=E[logP(C|Xreal)]+E[logP(C|Xfake)]

Ls=E[logP(S|Xreal)]+E[logP(S|Xfake)]

Wherein, E[] is the expected value operation; Lc is the expected value of the similarity between the first sample image and the standard medical image; Ls is the expected value of the effective information in the standard medical image; Xreal is the expected value Standard medical image; Xfake is the first sample image; C is the effective amount of information in the first sample image; S is the effective amount of information in the standard medical image.
The electronic device according to claim 9, wherein said inputting said effective image set and said standard medical image into an image discrimination model which has a mutual constraint relationship with said image generation model for image discrimination, and obtaining a discrimination result , Adjusting the parameters of the image discrimination model according to the discrimination result, until the discrimination result meets the preset requirements, obtaining the parameters of the image discrimination model at this time, including:

Construct the discriminative loss function Y;

Inputting the standard medical image and the effective image set to the image discrimination model to generate the final medical image set;

Input the final medical image set and the effective image set to the discriminant loss function Y for loss calculation to obtain a loss function value q;

When the loss function value q is greater than or equal to the preset loss threshold n, adjust the parameters of the image discrimination model to regenerate the final medical image;

When the loss function value q is less than the loss threshold n, the parameters of the image discrimination model at this time are obtained.
The electronic device according to claim 14, wherein the discriminant loss function Y comprises:

Y=Lc+Ls

Lc=E[logP(C|Xreal)]+E[logP(C|Xfake)]

Ls=E[logP(S|Xreal)]+E[logP(S|Xfake)]

Wherein, E[] is the expected value operation; Lc is the expected value of the similarity between the first sample image and the standard medical image; Ls is the expected value of the effective information in the standard medical image; Xreal is the expected value Standard medical image; Xfake is the first sample image; C is the effective amount of information in the first sample image; S is the effective amount of information in the standard medical image.
A computer-readable storage medium storing a computer program, wherein the computer program is executed by a processor to implement the following steps:

Acquiring an original medical image, performing conversion processing on the original medical image, to obtain an initial medical image;

Performing cell enhancement processing on the initial medical image to obtain a standard medical image;

Acquiring distribution data of each pixel information in the standard medical image, and generating a plurality of first sample images similar to the standard medical image by using an image generation model according to the distribution data to obtain a first sample image set;

The effective information amount calculation is performed on the first sample image set to obtain the effective information amount of each first sample image in the first sample image set. Select K first sample images corresponding to the effective amount of information from the sample image set to obtain an effective image set;

The effective image set and the standard medical image are input into an image discrimination model that has a mutual constraint relationship with the image generation model for image discrimination, and the discrimination result is obtained, and the parameters of the image discrimination model are adjusted according to the discrimination result , Until the discrimination result meets the preset requirements, obtain the parameters of the image discrimination model at this time, and obtain the standard image generation model according to the parameters;

The original medical image is input to the standard image generation model, a second sample image set is generated, and a final medical image is generated according to the second sample image.
15. The computer-readable storage medium according to claim 16, wherein said converting the original medical image to obtain the initial medical image comprises:

Converting the original medical image to a gray value to obtain an original gray image;

Performing noise reduction processing on the original gray image to obtain a noise reduction gray image;

Performing geometric transformation processing on the noise reduction grayscale image to obtain a transformed grayscale image;

The contrast enhancement of the transformed grayscale image is performed to obtain the initial medical image.
16. The computer-readable storage medium of claim 16, wherein said performing cell enhancement processing on said initial medical image to obtain a standard medical image comprises:

The following calculation formula is used to calculate the convolution of the initial medical image and the second-order Gaussian function to obtain the scale space derivative I abc of the initial medical image:

Where I is the initial medical image; G(x, y, z) is a Gaussian function; x, y, z are the parameters of the Gaussian function; σ is the standard offset of the Gaussian function;
Symbol for reciprocal operation
To find the convolution operation symbol;

Perform cell enhancement on the initial medical image according to the scale space derivative I abc to obtain the standard medical image.
The computer-readable storage medium according to claim 16, wherein the acquiring distribution data of each pixel information in the standard medical image, and generating a plurality of images similar to the standard medical image using an image generation model according to the distribution data The first sample image of, the first sample image set is obtained, including:

Construct sample generation loss function F;

Inputting the distribution data of each pixel information in the standard medical image to the image generation model to generate an initial first sample image set;

Inputting the initial first sample image set and the standard medical image to the sample generating loss function F for loss calculation, to obtain a loss function value p;

When the loss function value p is greater than or equal to the preset loss threshold m, adjust the parameters of the image generation model, and regenerate the first sample image;

When the loss function value p is less than the loss threshold m, the first sample image set is obtained.
The computer-readable storage medium of claim 19, wherein the sample generation loss function F comprises:

F=Lc-Ls

Lc=E[logP(C|Xreal)]+E[logP(C|Xfake)]

Ls=E[logP(S|Xreal)]+E[logP(S|Xfake)]

Wherein, E[] is the expected value operation; Lc is the expected value of the similarity between the first sample image and the standard medical image; Ls is the expected value of the effective information in the standard medical image; Xreal is the expected value Standard medical image; Xfake is the first sample image; C is the effective amount of information in the first sample image; S is the effective amount of information in the standard medical image.