Tumor image processing method and device based on artificial intelligence, and electronic equipment

Abstract

The invention provides a tumor image processing method and device based on artificial intelligence and electronic equipment, and relates to the field of image processing. The method comprises the following steps: acquiring a first tumor image to be detected and a second tumor image of the same body part after a preset time period; performing tumor region segmentation, optimization and feature extraction on the first tumor image to obtain a first tumor form and a first internal structure; performing tumor region segmentation, optimization and feature extraction on the second initial tumor boundary to obtain a second tumor form and a second internal structure; comparing the first tumor morphology with the second tumor morphology to obtain a first comparison result, and comparing the first internal structure with the second internal structure to obtain a second comparison result; fusing the first comparison result and the second comparison result to obtain a third comparison result; and according to the third comparison result, evaluating the body health condition to obtain the health level. The tumor image can be automatically processed and evaluated.

Description

Technical Field

[0001] The present application relates to the field of image processing technology, and specifically to a tumor image processing method, device and electronic device based on artificial intelligence. Background Art

[0002] In the survey report on cancer awareness and early screening, as many as 76.3% of the respondents said that someone among their relatives or friends had suffered from cancer, and 85.8% of the respondents said that they would worry about getting cancer themselves. Cancer is a major health threat. Among people with cancer, treatment is carried out through surgery, chemotherapy, radiotherapy and other means, and regular visits to the hospital for checkups are required to monitor the condition. For some smaller, slow-growing benign tumors that do not cause obvious symptoms to patients, doctors may choose a strategy of observation and monitoring, which requires regular physical examinations and medical imaging examinations. However, doctors have a heavy workload every day and are under great pressure to continuously process each medical image. Summary of the invention

[0003] The present application provides a tumor image processing method, device and electronic device based on artificial intelligence, which can automatically process tumor images and provide evaluations to reduce the pressure on doctors.

[0004] The technical solution of the embodiment of the present application is as follows: In a first aspect, an embodiment of the present application provides a tumor image processing method based on artificial intelligence, the method comprising: Acquire a first tumor image to be detected and a second tumor image of the same body part after a preset time period; Performing tumor region segmentation on the first tumor image to obtain a first initial tumor boundary, optimizing the first initial tumor boundary to obtain a first target tumor boundary, and performing feature extraction on the first tumor image according to the first target tumor boundary to obtain a first tumor morphology and a first internal structure; performing tumor region segmentation on the second tumor image to obtain a second initial tumor boundary, optimizing the second initial tumor boundary to obtain a second target tumor boundary, and performing feature extraction on the second tumor image according to the second target tumor boundary to obtain a second tumor morphology and a second internal structure; Comparing the first tumor morphology with the second tumor morphology to obtain a first comparison result, and comparing the first internal structure with the second internal structure to obtain a second comparison result; Merging the first comparison result and the second comparison result to obtain a third comparison result; According to the third comparison result, the physical health condition is evaluated to obtain a health grade.

[0005] In the above technical solution, a first tumor image to be detected and a second tumor image of the same body part after a preset time period are obtained, and images of the same body part at different times are obtained to provide data support for subsequent image processing to evaluate the health status; the first tumor image is segmented into tumor regions to obtain a first initial tumor boundary, the first initial tumor boundary is optimized to obtain a first target tumor boundary, and based on the first target tumor boundary, features are extracted from the first tumor image to obtain a first tumor morphology and a first internal structure. By automatically performing boundary processing and feature extraction processing on the first tumor image, the accuracy of the processing can be guaranteed, providing a basis for subsequent evaluation; the second tumor image is segmented into tumor regions to obtain a second initial tumor boundary, the second initial tumor boundary is optimized to obtain a second ... According to the boundary of the second target tumor, feature extraction is performed on the second tumor image to obtain the second tumor morphology and the second internal structure. By automatically performing boundary processing and feature extraction processing on the second tumor image, the accuracy of the processing can be guaranteed, providing a basis for subsequent evaluation; the first tumor morphology and the second tumor morphology are compared to obtain a first comparison result, and the first internal structure and the second internal structure are compared to obtain a second comparison result. Through comparison, continuous monitoring is performed to determine whether there are changes, thereby improving the accuracy of the evaluation; the first comparison result and the second comparison result are fused to obtain a third comparison result, and the accuracy is further improved through multi-level comparison and fusion; based on the third comparison result, the physical health status is evaluated to obtain the health level, which can be automatically evaluated so that the doctor can make a diagnosis based on the health level and reduce the pressure on the doctor.

[0006] In some embodiments of the present application, optimizing the first initial tumor boundary to obtain a first target tumor boundary includes: According to the first initial tumor boundary, dividing the first tumor image into tumor regions and normal regions multiple times to obtain respective division boundaries; The first initial tumor boundary is corrected according to the errors between each of the division boundaries and the first initial tumor boundary to obtain the first target tumor boundary.

[0007] In some embodiments of the present application, the first tumor morphology includes a first tumor shape and a first tumor size, and the second tumor morphology includes a second tumor shape and a second tumor size; The comparing the first tumor morphology and the second tumor morphology to obtain a first comparison result includes: comparing the first tumor shape with the second tumor shape to obtain a shape comparison result; Comparing the first tumor size with the second tumor size to obtain a size comparison result; The shape comparison result and the size comparison result are merged to obtain the first comparison result.

[0008] In some embodiments of the present application, comparing the first tumor shape with the second tumor shape to obtain a shape comparison result includes: Determine whether the first tumor shape and the second tumor shape are consistent, and if they are inconsistent, obtain a shape comparison result indicating that a lesion has occurred, and record the shape comparison result as 1; In the case of consistency, the shape comparison result is obtained as no lesion occurs, and the shape comparison result is recorded as 0.

[0009] In some embodiments of the present application, the first internal structure includes a first tumor density and a first tumor type, and the second internal structure includes a second tumor density and a second tumor type; The comparing the first internal structure and the second internal structure to obtain a second comparison result includes: determining whether the first tumor type is the same as the second tumor type, and if they are the same, converting the first tumor density and the second tumor density into data formats to obtain a first conversion density and a second conversion density; According to the preset evaluation, the first conversion density and the second conversion density, the Kendall coordination coefficient is calculated, and the Kendall coordination coefficient is compared with a preset relevant threshold value to obtain the second comparison result.

[0010] In some embodiments of the present application, the fusing the first comparison result and the second comparison result to obtain a third comparison result includes: Using a preset first machine learning algorithm to extract features from the first comparison result, a first result feature, and using a preset second machine learning algorithm to extract features from the second comparison result, to obtain a second result feature; The first result feature and the second result feature are weightedly fused to obtain the third comparison result.

[0011] In some embodiments of the present application, the step of evaluating the physical health status according to the third comparison result to obtain the health level includes: When the third comparison result is less than or equal to a preset first threshold, the physical health state is evaluated as a first level, where the first level is that the physical health state is good; When the third comparison result is greater than the first threshold value and less than or equal to a preset second threshold value, the physical health state is evaluated as a second level, and the second level is that the physical health state is medium; When the third comparison result is greater than the second threshold, the physical health status is evaluated as a third level, and the third level is that the physical health status is serious.

[0012] In a second aspect, an embodiment of the present application provides a tumor image processing device based on artificial intelligence, the device comprising: A data acquisition module, used to acquire a first tumor image to be detected and a second tumor image of the same body part after a preset time period; a first data processing module, configured to perform tumor region segmentation on the first tumor image to obtain a first initial tumor boundary, optimize the first initial tumor boundary to obtain a first target tumor boundary, and perform feature extraction on the first tumor image according to the first target tumor boundary to obtain a first tumor morphology and a first internal structure; a second data processing module, configured to perform tumor region segmentation on the second tumor image to obtain a second initial tumor boundary, optimize the second initial tumor boundary to obtain a second target tumor boundary, and perform feature extraction on the second tumor image according to the second target tumor boundary to obtain a second tumor morphology and a second internal structure; a data comparison module, configured to compare the first tumor morphology with the second tumor morphology to obtain a first comparison result, and to compare the first internal structure with the second internal structure to obtain a second comparison result; a data fusion module, used for fusing the first comparison result and the second comparison result to obtain a third comparison result; The evaluation module is used to evaluate the physical health condition according to the third comparison result to obtain a health grade.

[0013] In a third aspect, an embodiment of the present application provides an electronic device, comprising a processor, a memory, a user interface, a communication bus and a network interface, wherein the processor, the memory, the user interface and the network interface are respectively connected to the communication bus, the memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes any one of the methods provided in the first aspect.

[0014] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, and when the instructions are executed, any one of the methods provided in the first aspect is executed.

[0015] In summary, one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. By first acquiring a first tumor image to be detected and a second tumor image of the same body part after a preset time period, images of the same body part at different times are acquired to provide data support for subsequent image processing to evaluate the health status; performing tumor region segmentation on the first tumor image to obtain a first initial tumor boundary, optimizing the first initial tumor boundary to obtain a first target tumor boundary, performing feature extraction on the first tumor image according to the first target tumor boundary to obtain a first tumor morphology and a first internal structure, and automatically performing boundary processing and feature extraction processing on the first tumor image to ensure the accuracy of processing, providing a basis for subsequent evaluation; performing tumor region segmentation on the second tumor image to obtain a second initial tumor boundary, optimizing the second initial tumor boundary to obtain a second target tumor boundary, and According to the boundary of the second target tumor, the feature extraction of the second tumor image is performed to obtain the second tumor morphology and the second internal structure. By automatically performing boundary processing and feature extraction processing on the second tumor image, the accuracy of the processing can be guaranteed, providing a basis for subsequent evaluation; the first tumor morphology and the second tumor morphology are compared to obtain the first comparison result, and the first internal structure and the second internal structure are compared to obtain the second comparison result. Through comparison, continuous monitoring is performed to determine whether there is a change, thereby improving the accuracy of the evaluation; the first comparison result and the second comparison result are fused to obtain the third comparison result, and the accuracy is further improved through multi-level comparison fusion; according to the third comparison result, the physical health status is evaluated to obtain the health level, which can be automatically evaluated so that the doctor can make a diagnosis based on the health level and reduce the pressure on the doctor. Therefore, the problem that doctors in related technologies have heavy daily workloads and cannot observe medical images in a timely manner is effectively solved.

[0016] 2. By comparing the shape of the first tumor with the shape of the second tumor, and by comparing the size of the first tumor with the size of the second tumor, it can be determined whether the tumor has changed, thereby improving the accuracy of the assessment, so as to more effectively assist doctors in diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a flowchart of a tumor image processing method based on artificial intelligence provided by an embodiment of the present application; Figure 2 yes Figure 1 A schematic flow chart of a sub-step of step S200; Figure 3 yes Figure 1 A schematic flow chart of a sub-step of step S400; Figure 4 yes Figure 1 Another sub-step flow chart of step S400; Figure 5is a schematic diagram of the structure of a tumor image processing device based on artificial intelligence provided by an embodiment of the present application; Figure 6 It is a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0018] In order to enable technicians in this field to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments.

[0019] In the description of the embodiments of the present application, words such as "for example" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "for example" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "for example" or "for example" is intended to present related concepts in a specific way.

[0020] In the description of the embodiments of the present application, the meaning of the term "multiple" refers to two or more. For example, multiple systems refer to two or more systems, and multiple screen terminals refer to two or more screen terminals. In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. The terms "include", "comprise", "have" and their variations all mean "including but not limited to", unless otherwise specifically emphasized.

[0021] The embodiments of the present application provide a tumor image processing method, device, electronic device and readable storage medium based on artificial intelligence. The tumor image processing method based on artificial intelligence first obtains a first tumor image to be detected, and a second tumor image of the same body part after a preset time period, and obtains images of the same body part at different times to provide data support for subsequent image processing to evaluate the health status; performs tumor region segmentation on the first tumor image to obtain a first initial tumor boundary, optimizes the first initial tumor boundary to obtain a first target tumor boundary, and extracts features of the first tumor image based on the first target tumor boundary to obtain a first tumor morphology and a first internal structure. By automatically performing boundary processing and feature extraction processing on the first tumor image, the accuracy of the processing can be guaranteed, providing a basis for subsequent evaluation; performs tumor region segmentation on the second tumor image to obtain a second initial tumor boundary, and optimizes the first initial tumor boundary to obtain a first target tumor boundary. The second initial tumor boundary is optimized to obtain a second target tumor boundary. According to the second target tumor boundary, the second tumor image is feature extracted to obtain a second tumor morphology and a second internal structure. By automatically performing boundary processing and feature extraction processing on the second tumor image, the accuracy of the processing can be guaranteed, providing a basis for subsequent evaluation. The first tumor morphology and the second tumor morphology are compared to obtain a first comparison result. The first internal structure and the second internal structure are compared to obtain a second comparison result. Through comparison, continuous monitoring is performed to determine whether there are changes, thereby improving the accuracy of the evaluation. The first comparison result and the second comparison result are fused to obtain a third comparison result. Through multi-level comparison and fusion, the accuracy is further improved. According to the third comparison result, the physical health status is evaluated to obtain a health level, which can be automatically evaluated so that doctors can make diagnoses based on the health level and reduce the pressure on doctors.

[0022] It should be noted that this AI-based tumor image processing method is mainly used to monitor different tumor patients over a long period of time to determine whether the tumor has changed and to provide evaluation suggestions to assist doctors in making diagnoses and reduce their pressure.

[0023] The technical solution provided in the embodiments of the present application is further described below in conjunction with the accompanying drawings.

[0024] Reference Figure 1 , Figure 1 It is a flowchart of the tumor image processing method based on artificial intelligence provided by the embodiment of the present application. The tumor image processing method based on artificial intelligence is applied to a tumor image processing device based on artificial intelligence, and the tumor image processing method based on artificial intelligence is executed by a processor in an electronic device or a readable storage medium. The tumor image processing method based on artificial intelligence includes step S100, step S200, step S300, step S400, step S500 and step S600.

[0025] Step S100, acquiring a first tumor image to be detected and a second tumor image of the same body part after a preset time period.

[0026] In one embodiment, the first tumor image may be an X-ray image or a CT image, wherein the CT image is obtained based on the X-ray image, and the subsequent processing process is described using the CT image as an example in the following of this application. The second tumor image may be an X-ray image or a CT image, which is of the same type as the first tumor image and is not described here. The preset time period may be the time period of the inspection interval, which may be one month, half a month, etc. By archiving and storing the first tumor image, and archiving and storing the second tumor image of the same body part, or uploading it to the image preservation system, wherein the tumor image of the same body part is obtained to ensure monitoring of the tumor. Then, the first tumor image and the second tumor image to be detected are obtained through a preset reading function to provide data support for subsequent image processing to assess the health status, and the preset reading function may be a read() function.

[0027] Step S200, performing tumor region segmentation on the first tumor image to obtain a first initial tumor boundary, optimizing the first initial tumor boundary to obtain a first target tumor boundary, and performing feature extraction on the first tumor image based on the first target tumor boundary to obtain a first tumor morphology and a first internal structure.

[0028] In one embodiment, a preset segmentation algorithm can be used to segment the tumor region of the first tumor image to obtain a first initial tumor boundary. By dividing the boundary, the position of the tumor in the image can be accurately located so as to understand the growth, development and metastasis of the tumor. The preset segmentation algorithm can be a regional segmentation algorithm, a threshold segmentation method, or a boundary-based segmentation method. For the first initial tumor boundary obtained above, the first initial tumor boundary can also be optimized to better observe the development and metastasis of the tumor, and obtain the first target tumor boundary to improve the accuracy of image processing. Then, according to the first target tumor boundary, the tumor position can be accurately divided. For the divided tumor position, the first tumor image is feature extracted using a preset deep learning model to obtain the first tumor morphology and the first internal structure. The first tumor morphology includes tumor shape and tumor size; the first internal structure includes tumor density and tumor cysticity; the preset deep learning model can be a neural network model with an attention mechanism, or a convolutional neural network model with a U-shaped structure. By automatically performing boundary processing and feature extraction processing on the first tumor image, the accuracy of the processing can be guaranteed, providing a basis for subsequent evaluation.

[0029] like Figure 2As shown, the first initial tumor boundary is optimized to obtain the first target tumor boundary, including but not limited to the following steps: Step S210 : dividing the first tumor image into tumor regions and normal regions multiple times according to the first initial tumor boundary to obtain various division boundaries.

[0030] In one embodiment, the first initial tumor boundary is obtained by using a regional segmentation algorithm, and the first tumor image can be divided into tumor regions and normal regions multiple times using the regional segmentation algorithm. Since each division may have errors, the obtained division boundaries may be different. By obtaining each division boundary, the first initial tumor boundary can be corrected later, so that the location of the tumor is more accurate and the processing accuracy is improved.

[0031] In another embodiment, the threshold segmentation method or the boundary-based segmentation method can be used to divide the first tumor image into tumor areas and normal areas multiple times. Since the threshold segmentation method or the boundary-based segmentation method is different from the principle of the regional segmentation algorithm, there is a certain error in the division, and the obtained division boundaries may also be different. By obtaining each division boundary, the first initial tumor boundary can be corrected later, so that the location of the tumor is more accurate and the processing accuracy is improved. It should be noted that the first initial tumor boundary can also be obtained by using the threshold segmentation method or the boundary-based segmentation method, and then a single segmentation method or three mixed segmentation methods are used for segmentation to obtain each division boundary, which is similar to the above-mentioned use of the regional segmentation method or the mixed three segmentation methods to obtain each division boundary, and will not be repeated here.

[0032] Step S220: According to the errors between each segmentation boundary and the first initial tumor boundary, the first initial tumor boundary is corrected to obtain a first target tumor boundary.

[0033] In one embodiment, the error between each dividing boundary and the first initial tumor boundary is calculated, specifically: first, the first initial tumor boundary is represented by pixel coordinate points, and each dividing boundary is also represented by pixel coordinate points, and then the first initial tumor boundary and the boundary of each dividing boundary are aligned, and then the points on each dividing boundary are matched with the points on the first initial tumor boundary, and the error between the boundaries is calculated by calculating the distance between the points. According to the calculated error, in each dividing boundary, the error with the first initial tumor boundary is kept within a certain range, the first initial tumor boundary is modified, and the first initial tumor boundary and the error value are calculated to obtain the first target tumor boundary. By correcting the first initial tumor boundary, the location of the tumor can be accurately assessed, providing support for subsequent accurate processing.

[0034] Step S300, performing tumor region segmentation on the second tumor image to obtain a second initial tumor boundary, optimizing the second initial tumor boundary to obtain a second target tumor boundary, and performing feature extraction on the second tumor image based on the second target tumor boundary to obtain a second tumor morphology and a second internal structure.

[0035] In one embodiment, a preset segmentation algorithm can be used to segment the tumor region of the second tumor image to obtain a second initial tumor boundary. By dividing the boundary, the position of the tumor in the image can be accurately located so as to understand the growth, development and metastasis of the tumor. The preset segmentation algorithm can be a regional segmentation algorithm, a threshold segmentation method, or a boundary-based segmentation method. For the second initial tumor boundary obtained above, the second initial tumor boundary can also be optimized to better observe the development and metastasis of the tumor, and obtain the second target tumor boundary to improve the accuracy of image processing. Then, according to the second target tumor boundary, the tumor position can be accurately divided. For the divided tumor position, the second tumor image is feature extracted using a preset deep learning model to obtain the second tumor morphology and the second internal structure. The second tumor morphology includes tumor shape and tumor size; the second internal structure includes tumor density and tumor cysticity; the preset deep learning model can be a neural network model with an attention mechanism, or a convolutional neural network model with a U-shaped structure. By automatically performing boundary processing and feature extraction processing on the second tumor image, the accuracy of the processing can be guaranteed, providing a basis for subsequent evaluation. It should be noted that the processing method of the second tumor image is similar to the specific processing optimization method of the first tumor image, which will not be described in detail here.

[0036] Step S400: compare the first tumor morphology and the second tumor morphology to obtain a first comparison result, and compare the first internal structure and the second internal structure to obtain a second comparison result.

[0037] In one embodiment, the first tumor morphology includes a first tumor shape and a first tumor size, and the second tumor morphology includes a second tumor shape and a second tumor size. Figure 3 As shown, comparing the first tumor morphology and the second tumor morphology to obtain a first comparison result includes but is not limited to the following steps: Step S410: Compare the first tumor shape with the second tumor shape to obtain a shape comparison result.

[0038] In one embodiment, the first tumor shape is compared with the second tumor shape to obtain a shape comparison result, including but not limited to: determining whether the first tumor shape and the second tumor shape are consistent. If they are inconsistent, the shape comparison result is that a lesion has occurred, and the shape comparison result is recorded as 1; if they are consistent, the shape comparison result is that no lesion has occurred, and the shape comparison result is recorded as 0.

[0039] Specifically, firstly, whether the first tumor shape and the second tumor shape are consistent can be determined by shape translation and rotation, and the first tumor shape and the second tumor shape are compared to see whether they overlap. If they are inconsistent, it means that the malignant tumor has spread, or a benign tumor has become a malignant tumor and a lesion has occurred. The shape comparison result is that a lesion has occurred. For subsequent calculations, the shape comparison result is represented as 1. If they are consistent, it means that the benign tumor has not undergone a lesion or the malignant tumor has been controlled through treatment. The shape comparison result is that no lesion has occurred. For subsequent calculations, the shape comparison result is represented as 0. By comparing the shapes of tumors at different time periods, it is possible to observe whether the tumor has changed, thereby achieving automatic monitoring.

[0040] Step S420: Compare the first tumor size with the second tumor size to obtain a size comparison result.

[0041] In one embodiment, the first tumor size is represented by an area, and the second tumor size is also represented by an area. The first tumor size and the second tumor size are compared by calculating the areas of the two. If the area difference is within a specified threshold error range, it means that the tumor has not undergone a lesion, and the size comparison result is represented as 0. If the area difference is not within the specified threshold error range, it means that the tumor has undergone a lesion, and the size comparison result is represented as 1. By comparing the tumor sizes at different time periods, it is possible to observe whether the tumor has changed, thereby achieving automatic monitoring.

[0042] Step S430: Merge the shape comparison result and the size comparison result to obtain a first comparison result.

[0043] In one embodiment, the shape comparison result and the size comparison result are fused in a matrix manner to obtain a first comparison result, so that the physical condition can be automatically evaluated based on the first comparison result to assist doctors in diagnosis.

[0044] In another embodiment, the first internal structure includes a first tumor density and a first tumor type, and the second internal structure includes a second tumor density and a second tumor type. Figure 4 As shown, comparing the first internal structure and the second internal structure to obtain a second comparison result includes but is not limited to the following steps: Step S440, determining whether the first tumor type is the same as the second tumor type. If they are the same, converting the data formats of the first tumor density and the second tumor density to obtain a first conversion density and a second conversion density.

[0045] In some possible embodiments of the present application, the first tumor type and the second tumor type are breast tumors, lung cancer, etc. First, it is determined whether the first tumor type is the same as the second tumor type. By determining the type, it can be determined whether there are other cancerous changes, and it can also be guaranteed that the same tumor type is monitored. In the case of the same type, the first tumor density and the second tumor density are converted into data formats. Specifically, the density can be calculated using statistical methods such as mean value and standard deviation, and the first tumor density and the second tumor density are represented by unified data to obtain a first conversion density corresponding to the first tumor density and a second conversion density corresponding to the second tumor density, so as to facilitate subsequent comparison and obtain a second comparison result.

[0046] In some other possible embodiments of the present application, when the types are different, it indicates that other cancerous factors have appeared in the body part, and an alarm message is output, which prompts to seek medical treatment as soon as possible.

[0047] Step S450, calculating the Kendall coordination coefficient according to the preset evaluation, the first conversion density and the second conversion density, and comparing the Kendall coordination coefficient with a preset correlation threshold to obtain a second comparison result.

[0048] In some possible embodiments of the present application, the preset evaluation is an evaluation value of different tumor densities by a professional doctor. The Kendall coordination coefficient is calculated for the preset evaluation, the first conversion density, and the second conversion density to obtain a coordination coefficient value, which can reflect whether the first tumor density is consistent with the second tumor density, thereby determining whether the tumor has changed. The coordination coefficient is then compared with a preset correlation threshold. When the coordination coefficient is less than the preset correlation threshold, it indicates that the tumor has not changed, and a second comparison result is obtained, and the second comparison result is set to the coordination coefficient. When the coordination coefficient is greater than or equal to the preset correlation threshold, it indicates that the tumor has changed, and a second comparison result is obtained, and the second comparison result is set to 1. By comparing tumor densities in different time periods, it is possible to observe whether the tumor has changed, thereby achieving automatic monitoring.

[0049] Step S500: merge the first comparison result and the second comparison result to obtain a third comparison result.

[0050] Specifically, the first comparison result and the second comparison result are fused to obtain a third comparison result, including but not limited to: using a preset first machine learning algorithm to extract features from the first comparison result, a first result feature, and using a preset second machine learning algorithm to extract features from the second comparison result to obtain a second result feature; weighted fusion of the first result feature and the second result feature to obtain the third comparison result.

[0051] In some possible embodiments of the present application, a large number of tumor images to be detected and the comparison results of tumor images of the same body part after a preset time period are used to train a machine learning algorithm to learn the mutual influence relationship between the comparison results. For the mutual influence relationship, the preset first machine learning algorithm and the preset second machine learning algorithm can be the same model or different models. The preset first machine learning algorithm is used to extract features from the first comparison result, the first result feature, and the preset second machine learning algorithm is used to extract features from the second comparison result to obtain the second result feature, which can extract important information reflecting tumor changes. The first result feature and the second result feature are then weighted fused. The specific weighted fusion method is feature splicing, which can distinguish which features after fusion come from the first result feature and which come from the second result feature through the feature dimension to obtain the third comparison result. By retaining the difference between the first result feature and the second result feature, the third comparison result can ultimately reflect whether the tumor has changed, so that subsequent evaluation can be performed based on the third comparison result. Among them, the preset first machine learning algorithm and the preset second machine learning algorithm can be a support vector machine, or a convolutional neural network, etc.

[0052] Step S600: Evaluate the health condition of the body according to the third comparison result to obtain a health level.

[0053] Specifically, based on the third comparison result, the physical health status is evaluated to obtain a health level, including but not limited to: when the third comparison result is less than or equal to a preset first threshold, the physical health status is evaluated as a first level, and the first level is a good physical health status; when the third comparison result is greater than the first threshold and less than or equal to a preset second threshold, the physical health status is evaluated as a second level, and the second level is a medium physical health status; when the third comparison result is greater than the second threshold, the physical health status is evaluated as a third level, and the third level is a serious physical health status.

[0054] In some possible embodiments of the present application, the preset first threshold and the preset second threshold are both evaluated by professionals based on experience, and can be used to evaluate physical health. Exemplarily, the preset first threshold can be 0.45, the preset second threshold can be 0.75, and the smaller the value of the third comparison result, the smaller the physical health. When the third comparison result is less than or equal to the preset first threshold, it indicates that the physical health state is good and is at the first level, and the evaluation result can be given for the tumor image, and the evaluation result has a high accuracy. The user doctor at the first level does not need to spend too much energy, so as to help screen a part of the cases and reduce pressure. When the third comparison result is greater than the first threshold and less than or equal to the preset second threshold, it indicates that the physical health state is medium and is at the second level. Users at the second level need to consult a doctor, diagnose through a doctor, and reasonably control. When the third comparison result is greater than the second threshold, the physical health state is evaluated to be serious and at the third level. Users at the third level need to consult a doctor and are key users. By performing grade division, it automatically helps doctors to classify, thereby clarifying the use of main energy and reducing doctor pressure.

[0055] like Figure 5As shown, the embodiment of the present application provides an artificial intelligence-based tumor image processing device 100. The artificial intelligence-based tumor image processing device 100 acquires a first tumor image to be detected and a second tumor image of the same body part after a preset time period through a data acquisition module 110, and acquires images of the same body part at different times, so as to provide data support for subsequent image processing to evaluate the health status; uses a first data processing module 120 to perform tumor region segmentation on the first tumor image to obtain a first initial tumor boundary, optimizes the first initial tumor boundary to obtain a first target tumor boundary, and performs feature extraction on the first tumor image according to the first target tumor boundary to obtain a first tumor morphology and a first internal structure. By automatically performing boundary processing and feature extraction processing on the first tumor image, the accuracy of the processing can be guaranteed, providing a basis for subsequent evaluation; uses a second data processing module 130 to perform tumor region segmentation on the second tumor image to obtain a second initial tumor boundary, and optimizes the first initial tumor boundary to obtain a first target tumor boundary. The first initial tumor boundary is optimized to obtain a second target tumor boundary. According to the second target tumor boundary, the second tumor image is feature extracted to obtain a second tumor morphology and a second internal structure. By automatically performing boundary processing and feature extraction processing on the second tumor image, the accuracy of the processing can be guaranteed, providing a basis for subsequent evaluation. The data comparison module 140 is then used to compare the first tumor morphology and the second tumor morphology to obtain a first comparison result. The first internal structure and the second internal structure are compared to obtain a second comparison result. Through comparison, continuous monitoring is performed to determine whether there is a change, thereby improving the accuracy of the evaluation. The data fusion module 150 is used to fuse the first comparison result and the second comparison result to obtain a third comparison result. Through multi-level comparison and fusion, the accuracy is further improved. The evaluation module 160 is used to evaluate the physical health status according to the third comparison result to obtain the health level. The evaluation can be performed automatically so that the doctor can make a diagnosis based on the health level and reduce the pressure on the doctor.

[0056] It should be noted that the data acquisition module 110 is connected to the first data processing module 120 , the first data processing module 120 is connected to the second data processing module 130 , the second data processing module 130 is connected to the data comparison module 140 , the data comparison module 140 is connected to the data fusion module 150 , and the data fusion module 150 is connected to the evaluation module 160 . The above-mentioned artificial intelligence-based tumor image processing method is applied to an artificial intelligence-based tumor image processing device 100. The artificial intelligence-based tumor image processing device 100 obtains images of the same body part at different times by obtaining a first tumor image to be detected and a second tumor image of the same body part after a preset time period, so as to provide data support for subsequent image processing to evaluate the health status; performs tumor region segmentation on the first tumor image to obtain a first initial tumor boundary, optimizes the first initial tumor boundary to obtain a first target tumor boundary, and extracts features of the first tumor image according to the first target tumor boundary to obtain a first tumor morphology and a first internal structure. By automatically performing boundary processing and feature extraction processing on the first tumor image, the accuracy of the processing can be guaranteed, providing a basis for subsequent evaluation; performs tumor region segmentation on the second tumor image to obtain a second initial tumor boundary, The second initial tumor boundary is optimized to obtain a second target tumor boundary. According to the second target tumor boundary, the second tumor image is feature extracted to obtain a second tumor morphology and a second internal structure. By automatically performing boundary processing and feature extraction processing on the second tumor image, the accuracy of the processing can be guaranteed, providing a basis for subsequent evaluation. The first tumor morphology and the second tumor morphology are compared to obtain a first comparison result. The first internal structure and the second internal structure are compared to obtain a second comparison result. Through comparison, continuous monitoring is performed to determine whether there are changes, thereby improving the accuracy of the evaluation. The first comparison result and the second comparison result are fused to obtain a third comparison result. Through multi-level comparison and fusion, the accuracy is further improved. According to the third comparison result, the physical health status is evaluated to obtain a health level, which can be automatically evaluated so that doctors can make diagnoses based on the health level and reduce the pressure on doctors.

[0057] It should also be noted that: when the device provided in the above embodiment realizes its function, only the division of the above functional modules is used as an example. In actual application, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the device and method embodiments provided in the above embodiment belong to the same concept, and the specific implementation process is detailed in the method embodiment, which will not be repeated here.

[0058] The present application also discloses an electronic device. Figure 6 , Figure 6The electronic device 500 may include: at least one processor 501 , at least one network interface 504 , a user interface 503 , a memory 505 , and at least one communication bus 502 .

[0059] The communication bus 502 is used to realize the connection and communication between these components.

[0060] The user interface 503 may include a display screen (Display) and a camera (Camera). Optionally, the user interface 503 may also include a standard wired interface and a wireless interface.

[0061] The network interface 504 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).

[0062] Among them, the processor 501 may include one or more processing cores. The processor 501 uses various interfaces and lines to connect various parts in the entire server, and executes various functions of the server and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 505, and calling data stored in the memory 505. Optionally, the processor 501 can be implemented in at least one hardware form of digital signal processing (Digital Signal Processing, DSP), field programmable gate array (Field-Programmable Gate Array, FPGA), and programmable logic array (Programmable Logic Array, PLA). The processor 501 can integrate one or a combination of a central processing unit (Central Processing Unit, CPU), a graphics processing unit (Graphics Processing Unit, GPU) and a modem. Among them, the CPU mainly processes the operating system, user interface and application programs; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 501, and it can be implemented separately through a chip.

[0063] Among them, the memory 505 may include a random access memory (Random Access Memory, RAM) and may also include a read-only memory (Read-Only Memory). Optionally, the memory 505 includes a non-transitory computer-readable storage medium. The memory 505 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 505 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 505 may also be optionally at least one storage device located away from the aforementioned processor 501. Refer to Figure 6 , the memory 505 as a computer storage medium may include an operating system, a network communication module, a user interface module and an application program of a tumor image processing method based on artificial intelligence.

[0064] exist Figure 6 In the electronic device 500 shown, the user interface 503 is mainly used to provide an input interface for the user and obtain the data input by the user; and the processor 501 can be used to call the application program storing an artificial intelligence-based tumor image processing method in the memory 505, and when executed by one or more processors 501, the electronic device 500 executes one or more methods in the above-mentioned embodiments. It should be noted that for the aforementioned method embodiments, for the sake of simple description, they are all expressed as a series of action combinations, but those skilled in the art should know that the present application is not limited to the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required for the present application.

[0065] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0066] In the several embodiments provided in this application, it should be understood that the disclosed devices can be implemented in other ways. For example, the device embodiments described above are only schematic, such as the division of units, which is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some service interfaces, and the indirect coupling or communication connection of devices or units can be electrical or other forms.

[0067] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0068] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, 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 in the form of software functional units.

[0069] If the integrated unit 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 memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a memory and includes several instructions for a computer device (which can be a personal computer, server or network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned memory includes: various media that can store program codes, such as USB flash drives, mobile hard drives, magnetic disks or optical disks.

[0070] The above are only exemplary embodiments of the present disclosure and cannot be used to limit the scope of the present disclosure. That is, any equivalent changes and modifications made according to the teachings of the present disclosure are still within the scope of the present disclosure. After considering the disclosure of the specification, those skilled in the art will easily think of other embodiments of the present disclosure.

[0071] This application is intended to cover any variation, use or adaptation of the present disclosure, which follows the general principles of the present disclosure and includes common knowledge or customary technical means in the art not described in the present disclosure. The description and examples are to be regarded as exemplary only, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. A tumor image processing method based on artificial intelligence, characterized in that: The method comprises: Acquire a first tumor image to be detected and a second tumor image of the same body part after a preset time period; Performing tumor region segmentation on the first tumor image to obtain a first initial tumor boundary, optimizing the first initial tumor boundary to obtain a first target tumor boundary, and performing feature extraction on the first tumor image according to the first target tumor boundary to obtain a first tumor morphology and a first internal structure; performing tumor region segmentation on the second tumor image to obtain a second initial tumor boundary, optimizing the second initial tumor boundary to obtain a second target tumor boundary, and performing feature extraction on the second tumor image according to the second target tumor boundary to obtain a second tumor morphology and a second internal structure; Comparing the first tumor morphology with the second tumor morphology to obtain a first comparison result, and comparing the first internal structure with the second internal structure to obtain a second comparison result; Merging the first comparison result and the second comparison result to obtain a third comparison result; According to the third comparison result, the physical health condition is evaluated to obtain a health grade.

2. The method according to claim 1, characterized in that The step of optimizing the first initial tumor boundary to obtain a first target tumor boundary includes: According to the first initial tumor boundary, dividing the first tumor image into tumor regions and normal regions multiple times to obtain respective division boundaries; The first initial tumor boundary is corrected according to the errors between each of the division boundaries and the first initial tumor boundary to obtain the first target tumor boundary.

3. The method according to claim 1, characterized in that The first tumor morphology includes a first tumor shape and a first tumor size, and the second tumor morphology includes a second tumor shape and a second tumor size; The comparing the first tumor morphology and the second tumor morphology to obtain a first comparison result includes: comparing the first tumor shape with the second tumor shape to obtain a shape comparison result; Comparing the first tumor size with the second tumor size to obtain a size comparison result; The shape comparison result and the size comparison result are merged to obtain the first comparison result.

4. The method according to claim 3, characterized in that: The step of comparing the first tumor shape with the second tumor shape to obtain a shape comparison result includes: Determine whether the first tumor shape and the second tumor shape are consistent, and if they are inconsistent, obtain a shape comparison result indicating that a lesion has occurred, and record the shape comparison result as 1; In the case of consistency, the shape comparison result is obtained as no lesion occurs, and the shape comparison result is recorded as 0.

5. The method according to claim 1, characterized in that: The first internal structure includes a first tumor density and a first tumor type, and the second internal structure includes a second tumor density and a second tumor type; The comparing the first internal structure and the second internal structure to obtain a second comparison result includes: determining whether the first tumor type is the same as the second tumor type, and if they are the same, converting the first tumor density and the second tumor density into data formats to obtain a first conversion density and a second conversion density; According to the preset evaluation, the first conversion density and the second conversion density, the Kendall coordination coefficient is calculated, and the Kendall coordination coefficient is compared with a preset relevant threshold value to obtain the second comparison result.

6. The method according to claim 1, characterized in that The fusing the first comparison result and the second comparison result to obtain a third comparison result includes: Using a preset first machine learning algorithm to extract features from the first comparison result, a first result feature, and using a preset second machine learning algorithm to extract features from the second comparison result, to obtain a second result feature; The first result feature and the second result feature are weightedly fused to obtain the third comparison result.

7. The method according to claim 1, characterized in that The step of evaluating the health status of the patient according to the third comparison result to obtain a health level includes: When the third comparison result is less than or equal to a preset first threshold, the physical health state is evaluated as a first level, where the first level is that the physical health state is good; When the third comparison result is greater than the first threshold value and less than or equal to a preset second threshold value, the physical health state is evaluated as a second level, and the second level is that the physical health state is medium; When the third comparison result is greater than the second threshold, the physical health status is evaluated as a third level, and the third level is that the physical health status is serious.

8. A tumor image processing device based on artificial intelligence, characterized in that: The device comprises: A data acquisition module (110) is used to acquire a first tumor image to be detected and a second tumor image of the same body part after a preset time period; A first data processing module (120) is used to perform tumor region segmentation on the first tumor image to obtain a first initial tumor boundary, optimize the first initial tumor boundary to obtain a first target tumor boundary, and perform feature extraction on the first tumor image according to the first target tumor boundary to obtain a first tumor morphology and a first internal structure; A second data processing module (130) is used to perform tumor region segmentation on the second tumor image to obtain a second initial tumor boundary, optimize the second initial tumor boundary to obtain a second target tumor boundary, and perform feature extraction on the second tumor image according to the second target tumor boundary to obtain a second tumor morphology and a second internal structure; A data comparison module (140) is used to compare the first tumor morphology with the second tumor morphology to obtain a first comparison result, and to compare the first internal structure with the second internal structure to obtain a second comparison result; A data fusion module (150), configured to fuse the first comparison result and the second comparison result to obtain a third comparison result; The evaluation module (160) is used to evaluate the physical health condition according to the third comparison result to obtain a health grade.

9. An electronic device, characterized in that: The electronic device (500) comprises a processor (501), a memory (505), a user interface (503), a communication bus (502) and a network interface (504), wherein the processor (501), the memory (505), the user interface (503) and the network interface (504) are respectively connected to the communication bus (502), the memory (505) is used to store instructions, the user interface (503) and the network interface (504) are used to communicate with other devices, and the processor (501) is used to execute the instructions stored in the memory (505) so that the electronic device (500) executes the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions, and when the instructions are executed, the method according to any one of claims 1 to 7 is performed.

Images