Focus recognition method, device and equipment and readable storage medium

Abstract

The invention discloses a focus recognition method, device and equipment and a readable storage medium, and the method comprises the steps: obtaining a first medical image of a first tissue organ, and carrying out the focus recognition of the first tissue organ based on the first medical image; if it is determined that the first focus exists in the first tissue organ, predicting a target correlation degree between the first focus and a second focus corresponding to the second tissue organ based on feature information of the first focus and a pre-created prediction model; and identifying a second focus corresponding to the second tissue organ based on the target correlation degree. By identifying the correlation degree between the first focus of the first tissue organ and the second focus of the second tissue organ, whether the second focus exists in the second tissue organ or not is predicted, it can be avoided that other tissue organs with correlation among multiple focuses are independently detected and identified for multiple times, and the identification efficiency of the correlated focuses is improved.

Description

Technical Field

[0001] The present application relates to the field of medical image processing technology, and in particular to a lesion identification method, apparatus, device and readable storage medium. Background Art

[0002] With the development of AI imaging technology, image scanning has become a more common auxiliary diagnosis and treatment method in modern medical diagnosis and treatment.

[0003] Some tissues and organs may not be connected from an organ perspective, but their lesions often correlate with each other and are often detected together. However, the current common identification method involves performing lesion identification on each tissue and organ's corresponding medical image separately. This method results in multiple separate inspections and identifications for other tissues and organs with correlated lesions, reducing the efficiency of identifying related lesions. Summary of the Invention

[0004] The embodiments of the present application provide a lesion identification method, apparatus, device, and readable storage medium, which can improve the efficiency of identifying associated lesions.

[0005] In a first aspect, an embodiment of the present application provides a method for identifying a lesion, the method comprising:

[0006] Acquiring a first medical image of a first tissue organ, and identifying a lesion in the first tissue organ based on the first medical image;

[0007] If it is determined that a first lesion exists in the first tissue organ, predicting a target correlation degree between the first lesion and a second lesion corresponding to a second tissue organ based on the characteristic information of the first lesion and a pre-created prediction model;

[0008] Based on the target correlation degree, a second lesion corresponding to the second tissue organ is identified.

[0009] Optionally, the performing lesion identification on the first tissue organ based on the first medical image includes:

[0010] extracting morphological features, texture features, edge features, and type features of the first tissue organ based on the first medical image;

[0011] Based on the type features, determining a target feature recognition layer in a pre-created recognition model;

[0012] The morphological features, the texture features and the edge features are input into the target feature recognition layer to perform lesion recognition.

[0013] Optionally, the predicting the target correlation between the first lesion and a second lesion corresponding to a second tissue organ based on the feature information of the first lesion and a pre-created prediction model includes:

[0014] determining, based on the first lesion and preset lesion association information, a second lesion corresponding to a second tissue organ associated with the first lesion;

[0015] extracting common feature information from feature information of the first lesion based on the second lesion and the first lesion;

[0016] Based on the common feature information and a pre-created prediction model, the target correlation degree between the first lesion and the second lesion is predicted.

[0017] Optionally, predicting the target correlation degree between the first lesion and the second lesion based on the common feature information and a pre-created prediction model includes:

[0018] Obtain patient medical records;

[0019] The medical record information and the common feature information are input into a pre-created prediction model, and the prediction model is used to predict the target correlation degree between the first lesion and the second lesion based on the medical record information and the common feature information.

[0020] Optionally, predicting the target correlation between the first lesion and the second lesion based on the medical record information and the common feature information by the prediction model includes:

[0021] Predicting a reference correlation degree between the first lesion and the second lesion based on the common feature information using the prediction model;

[0022] The prediction model is used to adjust the reference correlation degree based on the patient metabolic information, patient hormone level information and patient genetic factor information in the medical record information to obtain the target correlation degree between the first lesion and the second lesion.

[0023] Optionally, the identifying the second lesion corresponding to the second tissue organ based on the target correlation degree includes:

[0024] Comparing the target relevance with a preset threshold;

[0025] If the target correlation degree is greater than or equal to a preset threshold, identifying a second lesion corresponding to the second tissue organ;

[0026] If the target correlation degree is less than a preset threshold, it indicates that the second lesion does not exist in the second tissue organ.

[0027] Optionally, identifying the second lesion corresponding to the second tissue organ includes:

[0028] acquiring a second medical image of the second tissue organ;

[0029] Based on the second medical image and a pre-created recognition model, a second lesion corresponding to the second tissue organ is identified.

[0030] In a second aspect, the present invention provides a lesion identification device.

[0031] a first identification unit, configured to acquire a first medical image of a first tissue organ, and identify a first lesion corresponding to the first tissue organ based on the first medical image;

[0032] a prediction unit, configured to, if it is determined that a first lesion exists in the first tissue organ, predict, based on feature information of the first lesion and a pre-created prediction model, a target correlation degree between the first lesion and a second lesion corresponding to a second tissue organ;

[0033] The second identification unit is used to identify the second lesion corresponding to the second tissue organ based on the target correlation degree.

[0034] In a third aspect, an embodiment of the present application further provides a lesion identification device, comprising a memory storing a computer program; a processor loading the computer program from the memory to execute the steps of any lesion identification method provided in the embodiment of the present application.

[0035] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, which stores a computer program, and the computer program is suitable for loading by a processor to execute the steps of any lesion identification method provided in the embodiment of the present application.

[0036] In a fifth aspect, an embodiment of the present application further provides a computer program product, comprising a computer program, which, when executed by a processor, implements the steps of any one of the lesion identification methods provided in the embodiments of the present application.

[0037] Using the solution of the embodiment of the application, a first medical image of a first tissue organ is obtained, and lesions of the first tissue organ are identified based on the first medical image; if it is determined that a first lesion exists in the first tissue organ, the target correlation between the first lesion and a second lesion corresponding to the second tissue organ is predicted based on the characteristic information of the first lesion and a pre-created prediction model; based on the target correlation, the second lesion corresponding to the second tissue organ is identified. By identifying the correlation between the first lesion of the first tissue organ and the second lesion of the second tissue organ to predict whether the second lesion exists in the second tissue organ, it is possible to avoid multiple separate detection and identification of other tissue organs where there is correlation between multiple lesions, thereby improving the efficiency of identifying associated lesions. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0039] Figure 1 This is a flowchart of the first embodiment of the lesion identification method provided by the present application; Figure 2 This is a flow chart of a second embodiment of the lesion identification method provided by the present application; Figure 3 This is a flowchart of the third embodiment of the lesion identification method provided by the present application; Figure 4 is a schematic structural diagram of a lesion identification device provided in an embodiment of the present application; Figure 5 It is a structural diagram of the lesion identification device provided in the embodiment of the present application. DETAILED DESCRIPTION

[0040] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without making creative work are within the scope of protection of this application. At the same time, in the description of the embodiments of the present application, the terms "first", "second", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more features. In the description of the embodiments of the present application, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0041] Embodiments of the present application provide a lesion identification method, apparatus, device, and readable storage medium.

[0042] Specifically, this embodiment will be described from the perspective of a lesion identification device, which can be integrated into a lesion identification apparatus, that is, the lesion identification method of the embodiment of the present application can be executed by the lesion identification device.

[0043] The lesion identification method provided in the embodiment of the present application can be applied to a lesion identification device, which can be a smart terminal, a PC terminal, a mobile terminal, or other device.

[0044] The following is a detailed description of each step in conjunction with the accompanying drawings. In this embodiment, the execution subject is a lesion identification device as an example. It should be noted that the order in which the following embodiments are described does not limit the preferred order of the embodiments. Although the flowcharts show a logical order, in some cases, the steps shown or described may be performed in a different order than that shown in the accompanying drawings.

[0045] Please refer to Figure 1 , a first embodiment of the lesion identification method is proposed, and the first embodiment includes the following steps:

[0046] Step 101: Acquire a first medical image of a first tissue organ, and identify a lesion in the first tissue organ based on the first medical image;

[0047] Step 102: If it is determined that a first lesion exists in the first tissue organ, then based on the characteristic information of the first lesion and a pre-created prediction model, predict the target correlation between the first lesion and a second lesion corresponding to a second tissue organ;

[0048] Step 103: Identify the second lesion corresponding to the second tissue organ based on the target correlation degree.

[0049] In this embodiment, when a patient needs to undergo a medical imaging examination of a first tissue organ, the lesion identification device obtains a first medical image of the first tissue organ, extracts characteristic information of the first tissue organ in the first medical image, and identifies the lesion of the first tissue organ based on the characteristic information. If the lesion identification device determines that a first lesion exists in the first tissue organ, the lesion identification device predicts the target correlation between the first lesion and a second lesion corresponding to a second tissue organ based on the characteristic information of the first lesion and a pre-created prediction model. Based on the target correlation, the second lesion corresponding to the second tissue organ is identified, and the likelihood of the second lesion corresponding to the second tissue organ and the type of the second lesion are identified. If the likelihood of the second lesion is high, the doctor may be prompted to undergo a medical imaging examination of the second tissue organ to further determine whether the second lesion exists in the second tissue organ.

[0050] It should be noted that the physiological tissue can be the heart, liver, kidney, breast, thyroid gland, etc.; for example, there is a certain correlation between thyroid nodules and breast nodules, and the two are often detected at the same time. If the lesion recognition device determines that there are nodules in the thyroid gland, it predicts the target correlation degree between the thyroid nodules and the breast nodules based on the characteristic information of the thyroid nodules and the pre-created prediction model, and identifies the breast nodules corresponding to the breast based on the target correlation degree.

[0051] The lesion identification device of this embodiment can improve the efficiency of lesion identification by identifying the correlation between a first lesion in a first tissue organ and a second lesion in a second tissue organ to predict whether the second lesion exists in the second tissue organ.

[0052] Specifically, each step is described in detail below:

[0053] Step 101: Acquire a first medical image of a first tissue organ, and identify lesions of the first tissue organ based on the first medical image.

[0054] In this step, the lesion identification device obtains a first medical image of a first tissue organ and performs lesion identification on the first tissue organ based on the first medical image. Optionally, the lesion identification device extracts characteristic information of the first tissue organ in the first medical image, obtains lesion characteristic information corresponding to each lesion of the first tissue organ from a preset first tissue organ lesion database, compares the characteristic information with the lesion characteristic information, and then performs lesion identification on the first tissue organ. Optionally, the lesion identification device extracts characteristic information of the first tissue organ in the first medical image, inputs the characteristic information into a pre-created recognition model, and uses the recognition model to identify the lesion of the first tissue organ based on the characteristic information; wherein the recognition model is pre-trained using training samples of multiple lesions of multiple different tissue organs.

[0055] Specifically, step 101 includes:

[0056] Step 1011: extracting morphological features, texture features, edge features, and type features of the first tissue organ based on the first medical image;

[0057] In this step, after obtaining the first medical image, the lesion recognition device first preprocesses the first medical image, and the preprocessing includes image standardization, noise reduction, edge detection and other processing to obtain the preprocessed first medical image; the lesion recognition device extracts features from the preprocessed first medical image to extract the morphological features, texture features, edge features and type features of the first tissue organ; wherein the morphological features include: shape: the geometric shape of the tissue or organ, such as circle, ellipse, star, etc., size: the size of the tissue or organ, usually including the measurement of length, width and height, volume: the overall volume of the tissue or organ, position: the spatial position of the tissue or organ in the image, boundary: the edge contour of the tissue or organ, whether it is regular or irregular; texture features include: gray level co-occurrence matrix ( Grayscale gradient oscillator (GLCM): describes the spatial relationship of grayscale values ​​in an image and is often used to extract texture features such as contrast, correlation, and uniformity. Local binary pattern (LBP): used to capture local texture features. Wavelet transform: analyzes multi-scale texture features of an image. Fourier transform: analyzes texture features in the frequency domain. Edge density: the density of texture details in an image. Edge features include: edge detection: using edge detection algorithms (such as Canny and Sobel) to identify the edges of tissues or organs. Edge sharpness: the clarity and sharpness of an edge. Edge roughness: the smoothness or roughness of an edge. Edge shape: the geometric shape of an edge, such as straight, curved, or wavy. Type features are used to indicate whether the first tissue or organ belongs to the heart, liver, kidney, breast, or thyroid gland.

[0058] Step 1012: determining a target feature recognition layer in a pre-created recognition model based on the type feature;

[0059] In this step, after determining the type characteristics of the first tissue organ, the lesion identification device determines the target feature identification layer in the pre-created identification model based on the type characteristics. It should be noted that the pre-created identification model includes multiple feature identification layers, and the feature identification layer is used to identify the characteristics of the first tissue organ to determine whether the first lesion exists in the first tissue organ. For different types of first tissue organs, the corresponding feature identification layers are different. Therefore, it is necessary to determine the target feature identification layer in the pre-created identification model based on the type characteristics; for example, when the type characteristics of the first tissue organ are heart, the feature identification layer for identifying heart lesions in the identification model is determined to be the target feature identification layer, and when the type characteristics of the first tissue organ are breast, the feature identification layer for identifying breast lesions in the identification model is determined to be the target feature identification layer.

[0060] Step 1013: Input the morphological features, the texture features, and the edge features into the target feature recognition layer to perform lesion recognition.

[0061] In this step, after determining the target feature recognition layer in the recognition model, the lesion recognition device inputs the morphological features, texture features and edge features of the first tissue organ into the target feature recognition layer to perform lesion recognition, and inputs information such as whether the first lesion exists in the first tissue organ and the type of the first lesion through the target feature recognition layer.

[0062] Step 102: If it is determined that a first lesion exists in the first tissue organ, then based on the characteristic information of the first lesion and a pre-created prediction model, predict the target correlation between the first lesion and a second lesion corresponding to a second tissue organ;

[0063] In this step, if the lesion identification device determines that a first lesion exists in the first tissue organ, it predicts the target correlation between the first lesion and a second lesion corresponding to the second tissue organ based on the characteristic information of the first lesion and a pre-created prediction model. The prediction model is pre-trained to predict the correlation between different lesions. Optionally, the lesion identification device inputs the characteristic information of the first lesion into the prediction model, and the prediction model outputs the target correlation between the first lesion and each type of second lesion corresponding to the second tissue organ. Optionally, the lesion identification device obtains characteristic information of a second lesion of the second tissue organ to be predicted, combines the characteristic information of the first lesion, determines common characteristic information between the second lesion and the first lesion, inputs the common characteristic information into the prediction model, and the prediction model outputs the target correlation between the first lesion and the second lesion of the second tissue organ. Optionally, the lesion identification device obtains medical records of the patient, inputs the patient's medical records and the common characteristic information between the second lesion and the first lesion into the prediction model, and the prediction model outputs the target correlation between the first lesion and the second lesion of the second tissue organ.

[0064] Step 103: Identify the second lesion corresponding to the second tissue organ based on the target correlation degree.

[0065] In this step, the lesion identification device determines the likelihood of the second lesion existing in the second tissue or organ, as well as the type of the second lesion, based on the target correlation degree. If the likelihood of the second lesion is high, the doctor may be prompted to perform a medical imaging examination on the second tissue or organ to further determine whether the second lesion exists in the second tissue or organ.

[0066] The lesion identification device of this embodiment obtains a first medical image of a first tissue organ and identifies lesions in the first tissue organ based on the first medical image. If the presence of a first lesion in the first tissue organ is determined, the device predicts the target correlation between the first lesion and a second lesion corresponding to a second tissue organ based on the characteristic information of the first lesion and a pre-created prediction model. Based on the target correlation, the device identifies the second lesion corresponding to the second tissue organ. By identifying the correlation between the first lesion in the first tissue organ and the second lesion in the second tissue organ to predict whether the second lesion exists in the second tissue organ, multiple separate detection and identification procedures can be avoided for other tissue organs where multiple lesions are correlated, thereby improving the efficiency of identifying associated lesions.

[0067] Further, refer to Figure 2 A second embodiment of the lesion identification method is proposed. The difference between the second embodiment and the first embodiment is that, based on the characteristic information of the first lesion and a pre-created prediction model, the target correlation degree between the first lesion and a second lesion corresponding to a second tissue organ is predicted, including:

[0068] Step 1021: determining a second lesion corresponding to a second tissue organ associated with the first lesion based on the first lesion and preset lesion association information;

[0069] In this step, the lesion identification device determines a second lesion corresponding to a second tissue organ associated with the first lesion based on the first lesion and the preset lesion association information. It should be noted that different lesions usually have certain correlations, and based on these correlations, preset lesion association information is generated. After determining the first lesion of the first tissue organ, the lesion identification device can search the preset lesion association information for a second lesion corresponding to the second tissue organ associated with the first lesion. For example, the first tissue organ is the thyroid gland, and the first lesion is a thyroid nodule. Thyroid nodules and breast nodules have a certain correlation and are often detected at the same time. In this case, the lesion identification device determines that the second lesion corresponding to the second tissue organ associated with the first lesion is a breast nodule. For further example, in addition to having a certain correlation with breast nodules, thyroid nodules also have a certain correlation with breast hyperplasia and breast cysts. These correlations are stored in the preset lesion association information. The second lesion corresponding to the second tissue organ associated with the first lesion determined by the lesion identification device may include: breast nodules, breast hyperplasia, and breast cysts.

[0070] Step 1022: extracting common feature information from the feature information of the first lesion based on the second lesion and the first lesion;

[0071] In this step, after determining the second lesion corresponding to the second tissue organ associated with the first lesion, the lesion identification device extracts common feature information from the feature information of the first lesion based on the second lesion and the first lesion. Specifically, the lesion identification device inputs the lesion type of the second lesion and the feature information of the first lesion into a pre-created common feature information extraction model, and extracts the common feature information of the second lesion and the first lesion from the feature information of the first lesion through the common feature information extraction model. Exemplarily, based on the clinical analysis information and diagnosis and treatment guide information of breast nodules and thyroid nodules, a deep learning neural network is used to construct a common feature information extraction model, and the common feature information extraction model is used to extract the lesion feature information in the medical image of the breast nodule to obtain the common feature information of the breast nodule and the thyroid nodule, or the common feature information extraction model is used to extract the lesion feature information in the medical image of the thyroid nodule to obtain the common feature information of the breast nodule and the thyroid nodule.

[0072] Step 1023: Predict the target correlation degree between the first lesion and the second lesion based on the common feature information and a pre-created prediction model.

[0073] In this step, the lesion identification device inputs the common feature information into a pre-created prediction model to predict the target correlation between the first lesion and the second lesion. Optionally, the pre-created prediction model further incorporates the common feature information to predict the target correlation between the first lesion and the second lesion. Optionally, the lesion identification device obtains the patient's medical history information while obtaining the common feature information, and uses the prediction model to predict the target correlation between the first lesion and the second lesion based on the common feature information and the patient's medical history information.

[0074] Specifically, step 1023 includes:

[0075] Step 10231, obtaining the patient's medical record information;

[0076] In this step, the lesion identification device is usually used in a hospital, where the patient's medical records are usually stored. After obtaining the common characteristic information of the patient's first lesion and the second lesion, the lesion identification device obtains the patient's medical records based on the patient's identity information.

[0077] Step 10232: Input the medical record information and the common feature information into a pre-created prediction model, and use the prediction model to predict the target correlation degree between the first lesion and the second lesion based on the medical record information and the common feature information.

[0078] In this step, after obtaining the patient's medical history information, the lesion identification device inputs the medical history information and common feature information into a pre-created prediction model. The prediction model then predicts the target correlation between the first and second lesions based on the medical history information and the common feature information. Optionally, the lesion identification device inputs the medical history information and the common feature information into the prediction model, and then uses the prediction model to predict the target correlation between the first and second lesions. Optionally, the lesion identification device inputs the common feature information into the prediction model to obtain a reference correlation between the first and second lesions, and then adjusts the reference correlation based on the medical history information to obtain the target correlation between the first and second lesions.

[0079] Specifically, step 10232 includes:

[0080] Step 102321: predicting a reference correlation degree between the first lesion and the second lesion based on the common feature information using the prediction model;

[0081] In this step, the lesion identification device inputs the common feature information into the prediction model, and the prediction model predicts the reference correlation degree between the first lesion and the second lesion based on the common feature information. Specifically, the lesion identification device first processes the common feature information, and the common feature information needs to be encoded in a way that can be understood and processed by a machine learning model to obtain a common feature representation. The encoding method includes: feature vector: combining the features of each nodule into a vector, for example, combining feature values ​​such as shape, texture, and color into a multidimensional vector; feature mapping: mapping image features into a high-dimensional space so that the machine learning algorithm can better understand and process it. After determining the common feature representation, the lesion identification device inputs the common feature representation into the prediction model and outputs the reference correlation degree between the first lesion and the second lesion, wherein the prediction model includes the use of supervised or unsupervised learning models, such as support vector machines, neural networks, clustering algorithms, etc.; applying similarity measurement methods (such as Euclidean distance, Manhattan distance, correlation coefficient, etc.) to calculate the reference correlation degree between the first lesion and the second lesion.

[0082] Step 102322: Adjust the reference correlation degree based on the patient metabolic information, patient hormone level information, and patient genetic factor information in the medical record information through the prediction model to obtain the target correlation degree between the first lesion and the second lesion.

[0083] In this step, the lesion identification device extracts the patient's metabolic information, hormone level information, and genetic factor information from the patient's medical records, determines a target weight based on the patient's metabolic information, hormone level information, and genetic factor information, and then adjusts the reference correlation degree based on the target weight to obtain the target correlation degree between the first lesion and the second lesion. Specifically, the lesion identification device constructs a regression model in advance, using the target weight as the dependent variable and the patient's metabolic information, hormone level information, and genetic factor information as independent variables. The model can be constructed using linear regression, ridge regression, or other regression techniques. The patient's metabolic information, hormone level information, and genetic factor information are input into the regression model to obtain target weights corresponding to the patient's metabolic information, hormone level information, and genetic factor information, respectively. The lesion identification device determines a metabolic characteristic value based on the patient's metabolic information, a hormone level characteristic value based on the patient's hormone level information, and a genetic factor characteristic value based on the patient's genetic factor information. Then, based on the metabolic characteristic value, hormone level characteristic value, genetic factor characteristic value, and the corresponding target weight, combined with the reference correlation degree, the target correlation degree between the first lesion and the second lesion is calculated. The specific calculation formula is: target correlation degree = 0.5*[reference correlation degree + (metabolic characteristic value*target weight of metabolic characteristic value+hormone level value*target weight of hormone level value+genetic factor value*target weight of genetic factor value)].

[0084] The lesion identification device of this embodiment determines the target correlation degree based on the common characteristic information of the first lesion and the second lesion and the patient's medical history information, and combines the actual situation of the patient himself to further improve the accuracy of the determined target correlation degree; at the same time, it helps to subsequently identify the second lesion corresponding to the second tissue organ based on the target correlation degree, which can improve the accuracy of identifying related lesions.

[0085] Further, refer to Figure 3 A third embodiment of the lesion identification method is proposed. The difference between the third embodiment and the first embodiment and the second embodiment is that the second lesion corresponding to the second tissue organ is identified based on the target correlation degree, including:

[0086] Step 1031, comparing the target relevance with a preset threshold;

[0087] Step 1032: If the target correlation degree is greater than or equal to a preset threshold, identifying a second lesion corresponding to the second tissue organ;

[0088] Step 1033: If the target correlation degree is less than a preset threshold, it is indicated that the second lesion does not exist in the second tissue organ.

[0089] In steps 1031 to 1033, the lesion identification device compares the target correlation degree with a preset threshold when determining the target correlation degree. If the target correlation degree is greater than or equal to the preset threshold, it is determined that the second tissue organ may have a second lesion, and the second lesion corresponding to the second tissue organ is identified. If the target correlation degree is less than the preset threshold, it is indicated that the second tissue organ does not have a second lesion.

[0090] Exemplarily, the first tissue organ is the thyroid gland, the first lesion is a thyroid nodule, and there is a certain correlation between thyroid nodules and breast nodules, breast hyperplasia, and breast cysts. The lesion identification device determines the first target correlation degree between thyroid nodules and breast nodules, the second target correlation degree between thyroid nodules and breast hyperplasia, and the third target correlation degree between thyroid nodules and breast cysts; the lesion identification device compares the first target correlation degree, the second target correlation degree, and the third target correlation degree with the preset threshold value. If it is determined that the first target correlation degree is greater than or equal to the preset threshold value, the breast nodule corresponding to the breast is identified.

[0091] Specifically, identifying the second lesion corresponding to the second tissue organ includes:

[0092] Step 10321, obtaining a second medical image of the second tissue organ;

[0093] Step 10322: Identify the second lesion corresponding to the second tissue organ based on the second medical image and the pre-created recognition model.

[0094] In steps 10321 and 10322, the lesion identification device, after determining that a second lesion may exist in the second tissue organ, obtains a second medical image of the second tissue organ, and identifies the second lesion corresponding to the second tissue organ based on the second medical image and a pre-established recognition model. Specifically, the lesion identification device extracts morphological features, texture features, edge features, and type features of the second tissue organ based on the second medical image; based on the type features, the lesion identification device determines a target feature recognition layer in the pre-established recognition model; and the lesion identification device inputs the morphological features, texture features, and edge features into the target feature recognition layer to identify the second lesion corresponding to the second tissue organ.

[0095] The lesion identification device in this embodiment compares the target correlation degree with a preset threshold; if the target correlation degree is greater than or equal to the preset threshold, the second lesion corresponding to the second tissue organ is identified; if the target correlation degree is less than the preset threshold, a prompt indicates that the second lesion does not exist in the second tissue organ. By comparing the target correlation degree with the preset threshold and only identifying the second lesion corresponding to the second tissue organ after determining that the target correlation degree is greater than or equal to the preset threshold, it can avoid multiple separate detection and identification of other tissue organs where multiple lesions are correlated, thereby improving the efficiency of identifying related lesions.

[0096] This embodiment also provides a lesion identification device, which can be integrated into lesion identification devices such as smart terminals, PC terminals, and mobile terminals. Figure 4 As shown, the lesion identification device may include:

[0097] A first identification unit 1001 is configured to obtain a first medical image of a first tissue organ, and identify a first lesion corresponding to the first tissue organ based on the first medical image;

[0098] The prediction unit 1002 is configured to predict, if it is determined that the first tissue organ has a first lesion, a target correlation between the first lesion and a second lesion corresponding to a second tissue organ based on feature information of the first lesion and a pre-created prediction model;

[0099] The second identification unit 1003 is configured to identify a second lesion corresponding to the second tissue organ based on the target correlation degree.

[0100] In an optional example, the first identification unit is further configured to:

[0101] extracting morphological features, texture features, edge features, and type features of the first tissue organ based on the first medical image;

[0102] Based on the type features, determining a target feature recognition layer in a pre-created recognition model;

[0103] The morphological features, the texture features and the edge features are input into the target feature recognition layer to perform lesion recognition.

[0104] In an optional example, the prediction unit is further configured to:

[0105] determining, based on the first lesion and preset lesion association information, a second lesion corresponding to a second tissue organ associated with the first lesion;

[0106] extracting common feature information from feature information of the first lesion based on the second lesion and the first lesion;

[0107] Based on the common feature information and a pre-created prediction model, the target correlation degree between the first lesion and the second lesion is predicted.

[0108] In an optional example, the prediction unit is further configured to:

[0109] Obtain patient medical records;

[0110] The medical record information and the common feature information are input into a pre-created prediction model, and the prediction model is used to predict the target correlation degree between the first lesion and the second lesion based on the medical record information and the common feature information.

[0111] In an optional example, the prediction unit is further configured to:

[0112] Predicting a reference correlation degree between the first lesion and the second lesion based on the common feature information using the prediction model;

[0113] The prediction model is used to adjust the reference correlation degree based on the patient metabolic information, patient hormone level information and patient genetic factor information in the medical record information to obtain the target correlation degree between the first lesion and the second lesion.

[0114] In an optional example, the second identification unit is further configured to:

[0115] Comparing the target relevance with a preset threshold;

[0116] If the target correlation degree is greater than or equal to a preset threshold, identifying a second lesion corresponding to the second tissue organ;

[0117] If the target correlation degree is less than a preset threshold, it indicates that the second lesion does not exist in the second tissue organ.

[0118] In an optional example, the second identification unit is further configured to:

[0119] acquiring a second medical image of the second tissue organ;

[0120] Based on the second medical image and a pre-created recognition model, a second lesion corresponding to the second tissue organ is identified.

[0121] Using the solution of this embodiment, a first medical image of a first tissue organ is acquired, and lesions are identified in the first tissue organ based on the first medical image. If the presence of a first lesion in the first tissue organ is determined, the target correlation between the first lesion and a second lesion corresponding to a second tissue organ is predicted based on the characteristic information of the first lesion and a pre-created prediction model. Based on the target correlation, the second lesion corresponding to the second tissue organ is identified. By identifying the correlation between the first lesion in the first tissue organ and the second lesion in the second tissue organ to predict whether the second lesion exists in the second tissue organ, multiple separate detection and identification procedures can be avoided for other tissue organs where multiple lesions are correlated, thereby improving the efficiency of identifying associated lesions.

[0122] Accordingly, the present application also provides a lesion identification device, such as Figure 5 As shown, Figure 5 This is a schematic diagram of the structure of a lesion identification device provided in an embodiment of the present application. The lesion identification device 1100 includes a processor 1101 having one or more processing cores, a memory 1102 having one or more computer-readable storage media, and a computer program stored in the memory 1102 and executable on the processor. The processor 1101 is electrically connected to the memory 1102. Those skilled in the art will understand that the structure of the lesion identification device shown in the figure does not constitute a limitation of the lesion identification device, and may include more or fewer components than shown, or combine certain components, or arrange the components differently.

[0123] Processor 1101 is the control center of lesion identification device 1100. It connects the various components of lesion identification device 1100 using various interfaces and lines. By running or loading software programs and / or units stored in memory 1102 and accessing data stored in memory 1102, it executes various functions of lesion identification device 1100 and processes data, thereby monitoring lesion identification device 1100 as a whole. Processor 1101 can be a processor CPU, graphics processor GPU, network processor (NP), etc., and can implement or execute the various methods, steps, and logic blocks disclosed in the embodiments of this application.

[0124] In an embodiment of the present application, the processor 1101 in the lesion identification device 1100 will load the instructions corresponding to the processes of one or more applications into the memory 1102 in accordance with the following steps, and the processor 1101 will run the applications stored in the memory 1102 to implement various functions. For specific implementation, please refer to the previous embodiments and will not be repeated here.

[0125] Optional, such as Figure 5As shown, the lesion identification device 1100 further includes: a touch screen 1103, a radio frequency circuit 1104, an audio circuit 1105, an input unit 1106, and a power supply 1107. Among them, the processor 1101 is electrically connected to the touch screen 1103, the radio frequency circuit 1104, the audio circuit 1105, the input unit 1106, and the power supply 1107 respectively. It can be understood by those skilled in the art that Figure 5 The structure of the lesion identification device shown in the figure does not constitute a limitation to the lesion identification device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0126] The touch screen 1103 can be used for displaying a graphical user interface and receiving the operation instructions that the user acts on the graphical user interface. The touch screen 1103 can include a display panel and a touch panel. Wherein, the display panel can be used for displaying the information input by the user or the information provided to the user and various graphical user interfaces of the lesion identification device, and these graphical user interfaces can be composed of graphics, text, icons, videos and any combination thereof. Optionally, the display panel can be configured in the form of a liquid crystal display (LCD, Liquid Crystal Display), an organic light emitting diode (OLED, Organic Light-Emitting Diode) or the like. The touch panel can be used for collecting the user's touch operation thereon or near it (such as the user uses any suitable object or accessory such as a finger, a stylus on the touch panel or near the touch panel), and generates corresponding operation instructions, and the operation instructions execute corresponding programs. Optionally, the touch panel can include two parts: a touch detection device and a touch controller. Among them, the touch detection device detects the user's touch direction, detects the signal brought by the touch operation, and transmits the signal to the touch controller; the touch controller receives the touch information from the touch detection device, converts it into the touch point coordinates, and then sends it to the processor 1101, and can receive the command sent by the processor 1101 and execute it. The touch panel can cover the display panel. When the touch panel detects a touch operation on or near it, it is transmitted to the processor 1101 to determine the type of touch event. Then the processor 1101 provides a corresponding visual output on the display panel according to the type of touch event. In an embodiment of the present application, the touch panel and the display panel can be integrated into the touch display screen 1103 to realize the input and output functions. However, in some embodiments, the touch panel and the touch panel can be used as two independent components to realize the input and output functions. That is, the touch display screen 1103 can also be used as part of the input unit 1106 to realize the input function.

[0127] The radio frequency circuit 1104 may be used to transmit and receive radio frequency signals, so as to establish wireless communication with a network device or other lesion identification device through wireless communication, and to transmit and receive signals between the network device or other lesion identification device.

[0128] The audio circuit 1105 can be used to provide an audio interface between the user and the lesion identification device through a speaker and a microphone. The audio circuit 1105 can convert the received audio data into an electrical signal and transmit it to the speaker, which converts it into a sound signal for output. On the other hand, the microphone converts the collected sound signal into an electrical signal, which is received by the audio circuit 1105 and converted into audio data. The audio data is then output to the processor 1101 for processing, and then sent to another lesion identification device through the radio frequency circuit 1104, or the audio data is output to the memory 1102 for further processing. The audio circuit 1105 may also include an earphone jack to provide communication between external headphones and the lesion identification device.

[0129] The input unit 1106 may be configured to receive input digital, character information, or user feature information (such as fingerprint, iris, or facial information), and to generate keyboard, mouse, joystick, optical, or trackball signal input related to user settings and function control.

[0130] Power supply 1107 is used to power the various components of lesion identification device 1100. Optionally, power supply 1107 can be logically connected to processor 1101 via a power management device, thereby enabling the power management device to manage charging, discharging, and power consumption. Power supply 1107 can also include one or more DC or AC power supplies, recharging devices, power failure detection circuits, power converters or inverters, power status indicators, and other arbitrary components.

[0131] although Figure 5 Not shown in the figure, the lesion identification device 1100 may also include a camera, a sensor, a wireless fidelity module, a Bluetooth module, etc., which will not be repeated here.

[0132] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0133] Those skilled in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be accomplished by instructions, or by controlling related hardware through instructions. The instructions may be stored in a computer-readable storage medium and loaded and executed by a processor.

[0134] To this end, embodiments of the present application provide a computer-readable storage medium storing a plurality of computer programs capable of being loaded by a processor to execute any of the lesion identification methods provided in embodiments of the present application. The computer program can execute the lesion identification method. The specific implementation thereof can be found in the previous embodiments and will not be further described here.

[0135] The computer-readable storage medium may include a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0136] Since the computer program stored in the computer-readable storage medium can execute any of the lesion identification methods provided in the embodiments of the present application, the beneficial effects that can be achieved by any of the lesion identification methods provided in the embodiments of the present application can be achieved. Please refer to the previous embodiments for details and will not be repeated here.

[0137] According to one aspect of the present application, a computer program product or computer program is also provided, comprising computer instructions stored in a computer-readable storage medium. A processor of a lesion identification device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the lesion identification device to perform the methods provided in various optional implementations of the above-described embodiments.

[0138] In the above-mentioned embodiments of the lesion identification device, computer-readable storage medium, lesion identification equipment, and computer program product, the descriptions of each embodiment have different focuses. For parts not described in detail in a particular embodiment, reference can be made to the relevant descriptions of other embodiments. Those skilled in the art will clearly understand that for the convenience and brevity of description, the specific working processes and beneficial effects of the lesion identification device, computer-readable storage medium, computer program product, lesion identification equipment, and their corresponding units described above can be referred to the description of the lesion identification method in the above embodiments, and the details will not be repeated here.

[0139] The above is a detailed introduction to a lesion identification method, device, equipment, readable storage medium and computer program product provided in the embodiments of the present application. Specific examples are used herein to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for technical personnel in this field, based on the ideas of the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A lesion identification method, characterized in that: The lesion identification method comprises: Acquiring a first medical image of a first tissue organ, and identifying a lesion in the first tissue organ based on the first medical image; If it is determined that a first lesion exists in the first tissue organ, predicting a target correlation degree between the first lesion and a second lesion corresponding to a second tissue organ based on the characteristic information of the first lesion and a pre-created prediction model; Based on the target correlation degree, a second lesion corresponding to the second tissue organ is identified.

2. The lesion identification method according to claim 1, wherein The performing lesion identification on the first tissue organ based on the first medical image includes: extracting morphological features, texture features, edge features, and type features of the first tissue organ based on the first medical image; Based on the type features, determining a target feature recognition layer in a pre-created recognition model; The morphological features, the texture features and the edge features are input into the target feature recognition layer to perform lesion recognition.

3. The lesion identification method according to claim 1, wherein The predicting, based on the feature information of the first lesion and a pre-created prediction model, the target correlation degree between the first lesion and a second lesion corresponding to a second tissue organ includes: determining, based on the first lesion and preset lesion association information, a second lesion corresponding to a second tissue organ associated with the first lesion; extracting common feature information from feature information of the first lesion based on the second lesion and the first lesion; Based on the common feature information and a pre-created prediction model, the target correlation degree between the first lesion and the second lesion is predicted.

4. The method for identifying lesions according to claim 3, wherein: The predicting the target correlation between the first lesion and the second lesion based on the common feature information and a pre-created prediction model includes: Obtain patient medical records; The medical record information and the common feature information are input into a pre-created prediction model, and the prediction model is used to predict the target correlation degree between the first lesion and the second lesion based on the medical record information and the common feature information.

5. The method for identifying lesions according to claim 4, wherein: The predicting the target correlation between the first lesion and the second lesion based on the medical record information and the common feature information by the prediction model includes: Predicting a reference correlation degree between the first lesion and the second lesion based on the common feature information using the prediction model; The prediction model is used to adjust the reference correlation degree based on the patient metabolic information, patient hormone level information and patient genetic factor information in the medical record information to obtain the target correlation degree between the first lesion and the second lesion.

6. The method for identifying lesions according to claim 1, wherein: The identifying, based on the target correlation degree, a second lesion corresponding to the second tissue organ, includes: Comparing the target relevance with a preset threshold; If the target correlation degree is greater than or equal to a preset threshold, identifying a second lesion corresponding to the second tissue organ; If the target correlation degree is less than a preset threshold, it indicates that the second lesion does not exist in the second tissue organ.

7. The method for identifying lesions according to claim 6, wherein: The identifying the second lesion corresponding to the second tissue organ includes: acquiring a second medical image of the second tissue organ; Based on the second medical image and a pre-created recognition model, a second lesion corresponding to the second tissue organ is identified.

8. A lesion identification device, characterized in that: The lesion identification device comprises: a first identification unit, configured to acquire a first medical image of a first tissue organ, and identify a first lesion corresponding to the first tissue organ based on the first medical image; a prediction unit, configured to, if it is determined that a first lesion exists in the first tissue organ, predict, based on feature information of the first lesion and a pre-created prediction model, a target correlation degree between the first lesion and a second lesion corresponding to a second tissue organ; The second identification unit is used to identify the second lesion corresponding to the second tissue organ based on the target correlation degree.

9. An electronic device, characterized in that: It comprises a processor and a memory, wherein the memory stores a computer program; the processor loads the computer program from the memory to execute the steps of the lesion identification 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 a computer program, and the computer program is suitable for being loaded by a processor to execute the steps of the lesion identification method according to any one of claims 1 to 7.

Images