A method and system for extracting medical image features

Abstract

The present invention discloses a method and system for extracting medical image features, belonging to the technical field of image processing. First, medical image data is collected and preprocessed to construct a data set. An adaptive feature extraction model is constructed based on the U-Net network. The best model is obtained through the Adam algorithm combined with the cross-entropy loss function. The data set is input into the model to obtain preliminary image features. Then, the uncertainty in the image space and feature space of the preliminary image features is calculated to determine the comprehensive uncertainty. Based on the machine learning model, the feature weights are obtained and weighted calculation is performed to obtain a comprehensive evaluation index. Finally, the top q% features are selected as the preferred image features according to the index ranking. This method improves the accuracy of feature extraction by optimizing the model. The uncertainty perception mechanism can effectively screen features, reduce the influence of redundant and irrelevant features, and comprehensively evaluate features with the comprehensive evaluation index, which helps to improve the efficiency and diagnostic accuracy of medical image analysis and provides a more reliable feature set for medical image processing.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and more specifically, to a method and system for extracting medical image features. Background Art

[0002] In the context of the booming development of intelligent medicine, a large amount of medical image data is continuously generated, such as X-ray films, CT scans, MRI images, etc. Accurate extraction of medical image features can mine valuable information from these complex and large amounts of data, and the medical image feature extraction method has become a core and key link in the field of medical image analysis.

[0003] Traditional medical image feature extraction methods often rely on the experience and knowledge of doctors or experts. However, this method has subjectivity and inconsistency. Different doctors or experts may extract different feature sets, which affects the effectiveness and repeatability of the features. Moreover, in medical images, the number of features may be very large, but not all features are relevant to the diagnosis or analysis task. Existing technologies often have difficulty effectively distinguishing which features are important and which features are redundant or irrelevant. This results in a large number of irrelevant or redundant features being included in the feature extraction process, thereby reducing the effectiveness of the features and leading to low disease diagnosis efficiency.

[0004] Therefore, how to construct an efficient, accurate and reliable medical image feature extraction method and system is an urgent problem to be solved by those skilled in the art. Summary of the Invention

[0005] In view of this, the present invention provides a method and system for extracting medical image features, which effectively improves the accuracy and efficiency of medical image feature extraction and provides strong support for medical image analysis.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] On the one hand, the present invention proposes a method for extracting medical image features, including the following steps:

[0008] Collect medical image data, preprocess the medical image data, and construct a data set;

[0009] Construct and train an adaptive feature extraction model to obtain an optimal feature extraction model;

[0010] Input the data set into the adaptive feature extraction model to obtain preliminary image features;

[0011] Based on an uncertainty perception mechanism, screen the preliminary image features, and determine the preferred image features by combining the importance and uncertainty of the preliminary image features.

[0012] Preferably, an adaptive feature extraction model is constructed based on the U-Net network, and the adaptive feature extraction model is optimized by the Adam algorithm based on the cross-entropy loss function to obtain the optimal feature extraction model.

[0013] Preferably, the adaptive feature extraction model is optimized by the Adam algorithm based on the cross-entropy loss function to obtain the optimal feature extraction model, including:

[0014] Taking the minimization of the loss of the cross-entropy function as the optimization objective, the Adam algorithm optimizes the model parameters according to the optimization objective;

[0015] The BatchNorm technique is used to optimize the data distribution in the training process of the adaptive feature extraction model.

[0016] Preferably, the preliminary image features are screened based on the uncertainty perception mechanism, and the preferred image features are determined by combining the importance and uncertainty of the preliminary image features, including:

[0017] Calculate the image space uncertainty and feature space uncertainty of each preliminary image feature;

[0018] Determine the comprehensive uncertainty of the preliminary image feature according to the image space uncertainty and the feature space uncertainty;

[0019] Based on the machine learning model, obtain the importance of the preliminary image features, and perform weighted calculations with the corresponding comprehensive uncertainties respectively to obtain the comprehensive evaluation index;

[0020] Rank the preliminary image features according to the comprehensive evaluation index, and select the top q% of the preliminary image features as the preferred image features.

[0021] Preferably, for the nth preliminary image feature, the image space uncertainty CV n The calculation formula is as follows:

[0022]

[0023] |N(x, y)| represents the number of pixels in the neighborhood N(x, y) of the pixel (x, y), μ(x, y) is the mean value of all pixel values in the neighborhood N(x, y), and I(i, j) is the pixel value of the image at the position (i, j).

[0024] Preferably, for the m feature points of the nth preliminary image feature, the feature space uncertainty FPD n The calculation formula is as follows:

[0025]

[0026] In the formula, d(p l , p k ) is the distance between feature points, and σ is a parameter for controlling distance sensitivity.

[0027] Preferably, the calculation formula for the comprehensive uncertainty A of the nth preliminary image feature is as follows:

[0028] A = α·CV n + β·FPD n ;

[0029] In the formula, α and β are the uncertainty weights of the image space and the feature space, respectively.

[0030] On the other hand, the present invention also proposes a medical image feature extraction system for implementing the above medical image feature extraction method, including:

[0031] A data preprocessing module for collecting medical image data, preprocessing the medical image data, and constructing a data set;

[0032] A model construction module for constructing and training an adaptive feature extraction model to obtain an optimal feature extraction model;

[0033] A preliminary feature extraction module for inputting the data set into the adaptive feature extraction model to obtain preliminary image features;

[0034] A feature screening module for screening the preliminary image features based on an uncertainty perception mechanism, and determining preferred image features by combining the importance and uncertainty of the preliminary image features.

[0035] Preferably, the model construction module includes:

[0036] A parameter optimization unit for taking the minimization of the loss of the cross-entropy function as the optimization target, and optimizing the model parameters according to the optimization target by the Adam algorithm;

[0037] A data distribution optimization unit for optimizing the data distribution in the training process of the adaptive feature extraction model by using the BatchNorm technique.

[0038] Preferably, the feature screening module includes:

[0039] An uncertainty calculation unit for calculating the image space uncertainty and the feature space uncertainty of each preliminary image feature;

[0040] A comprehensive uncertainty calculation unit for determining the comprehensive uncertainty of the preliminary image features according to the image space uncertainty and the feature space uncertainty;

[0041] A comprehensive evaluation index acquisition unit, which is used to obtain the importance of preliminary image features based on a machine learning model, and perform weighted calculations with the corresponding comprehensive uncertainties respectively to obtain comprehensive evaluation indexes;

[0042] An optimal feature image determination unit, which is used to sort the preliminary image features according to the comprehensive evaluation indexes, and select the top q% of the preliminary image features as the optimal image features.

[0043] As can be seen from the above technical solutions, compared with the prior art, the present invention discloses a medical image feature extraction method and system. By collecting and preprocessing medical image data to construct a data set, using a U-Net network to construct an adaptive feature extraction model, optimizing the model parameters with the goal of minimizing the cross-entropy loss and using BatchNorm to optimize the data distribution to obtain the best model, and then extracting preliminary image features. Subsequently, based on the uncertainty perception mechanism, combining the image space uncertainty and the feature space uncertainty, calculating the weights of the preliminary image features through a machine learning model, comprehensively evaluating and screening out the optimal image features. The present invention improves the accuracy of feature extraction by optimizing the model. The uncertainty perception mechanism can effectively screen features, reduce the influence of redundant and irrelevant features, and the comprehensive evaluation index comprehensively evaluates the features, which helps to improve the efficiency and diagnostic accuracy of medical image analysis and provides a more reliable feature set for medical image processing. Description of the Drawings

[0044] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only the embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to the provided drawings.

[0045] Figure 1 It is a flowchart of the method provided by the present invention;

[0046] Figure 2 It is an adaptive feature extraction model based on the U-Net network provided by the embodiment of the present invention;

[0047] Figure 3 It is a system architecture diagram provided by the present invention. Detailed Embodiments

[0048] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0049] On the one hand, referring to Figure 1 , an embodiment of the present invention provides a medical image feature extraction method, including the following steps:

[0050] S1. Collect medical image data, preprocess the medical image data, and construct a data set.

[0051] Collect data from various medical imaging devices, including but not limited to CT scanners, MRI instruments, X-ray machines, etc., to ensure that images with different imaging principles and resolutions are covered to obtain comprehensive patient body information. For the image formats and data characteristics of different devices, develop corresponding interface programs and conversion tools to enable unified processing of the collected data. For example, for the DICOM format of CT images and the specific format of MRI images, perform format parsing and data extraction, and convert them into a common format suitable for subsequent preprocessing and analysis, such as a common image matrix format.

[0052] Noise removal: Adopt various filtering algorithms, such as Gaussian filtering, median filtering, etc., and select appropriate filter parameters according to the type of image noise. For images with more Gaussian noise, use a Gaussian filter, and smooth the image by setting an appropriate kernel size (such as 3×3 or 5×5) to reduce the interference of noise on image details; for salt-and-pepper noise, use a median filter and select an appropriate neighborhood size (such as 3×3) to replace the noise pixel values.

[0053] Gray normalization: Calculate the gray histogram of the image, determine the gray range of the image, and normalize it to a specific gray interval, such as [0, 255]. By counting the number of pixels at each gray level in the image, find the minimum and maximum values of the gray level, and then map the gray value of each pixel to the target interval according to the normalization formula to enhance the contrast and brightness consistency of the image, facilitating subsequent feature extraction.

[0054] Image enhancement: Apply techniques such as histogram equalization and contrast stretching. Histogram equalization redistributes the gray histogram of the image to make the gray distribution of the image more uniform and improve the contrast of the image; contrast stretching selects an appropriate stretching coefficient according to the gray range of the image to expand the gray dynamic range of the image and highlight the detail information in the image.

[0055] Image Cropping and Scaling: Automatically crop images based on regions of interest (ROIs), removing irrelevant background information and reducing data volume and computational burden. For example, in lung CT images, the system automatically identifies and crops the lung region, retaining only the lung and surrounding areas. Furthermore, the system scales images based on subsequent model input requirements, such as scaling images of varying resolutions to a standard size such as 256×256 or 512×512, to ensure data consistency.

[0056] Data Labeling and Classification: We invite professional medical imaging doctors to annotate some images, including lesion areas, organ outlines, abnormal structures, etc. After labeling, the images are classified according to the annotated categories, and further subdivided into multiple categories based on specific disease types, providing accurate label information for subsequent model training.

[0057] S2. Build and train an adaptive feature extraction model to obtain the optimal feature extraction model.

[0058] Build an adaptive feature extraction model based on the U-Net network, refer to Figure 2 The network is an encoder-decoder structure, where the encoder is used to extract contextual information of the image and the decoder is used to restore image information.

[0059] U-Net uses the same-level convolution between the upsampling and downsampling layers. The corresponding upsampling and downsampling layers are spliced in the form of skip connections, rather than pixel-by-pixel splicing layer by layer. This skip connection structure helps to transfer information from the shallow layers of the network to the deep layers, thereby obtaining better segmentation results.

[0060] The U-Net network structure is shaped like the Latin letter U and mainly includes two paths, namely the contraction path and the expansion path.

[0061] The contraction path consists of two repeated 3×3 convolutions, followed by a convolution (RELU) and a maximum pooling (maxpooling with a stride of 2) for downsampling. Each downsampling reduces the feature channel by half. The expansion path consists of an upsampling (2×2 upconvolution), followed by a feature map of the corresponding contraction path, and then two 3×3 convolutions, each followed by a RELU. Because each convolution loses the image edge and causes the image size to change, which is not conducive to the subsequent propagation of information, it is cropped and followed by a 1×1 convolution. This network structure does not have a fully connected layer. The deeper layers are used to solve the problem of pixel classification, and the shallower high-resolution layers are used to solve the problem of pixel positioning.

[0062] The adaptive feature extraction model is optimized using the Adam algorithm and the cross-entropy loss function to obtain the optimal feature extraction model, including:

[0063] Taking the minimization of the loss of the cross - entropy function as the optimization objective, the Adam algorithm optimizes the model parameters according to the optimization objective.

[0064] The BatchNorm technique is used to optimize the data distribution during the training of the adaptive feature extraction model.

[0065] S3. Input the data set into the adaptive feature extraction model to obtain preliminary image features.

[0066] S4. Screen the preliminary image features based on the uncertainty - aware mechanism, and determine the preferred image features by combining the importance and uncertainty of the preliminary image features, including:

[0067] S41. Calculate the image - space uncertainty and feature - space uncertainty of each preliminary image feature.

[0068] For the nth preliminary image feature, the image - space uncertainty CV n The calculation formula is as follows:

[0069]

[0070] |N(x,y)| represents the number of pixels in the neighborhood N(x,y) of the pixel (x,y), μ(x,y) is the mean value of all pixel values in the neighborhood N(x,y), and I(i,j) is the pixel value of the image at the position (i,j).

[0071] For the m feature points of the nth preliminary image feature, the feature - space uncertainty FPD n The calculation formula is as follows:

[0072]

[0073] In the formula, d(p l ,p k ) is the distance between feature points, and σ is the parameter controlling the distance sensitivity.

[0074] S42. Determine the comprehensive uncertainty of the preliminary image features according to the image - space uncertainty and feature - space uncertainty.

[0075] The calculation formula for the comprehensive uncertainty A of the nth preliminary image feature is as follows:

[0076] A = α·CV n +β·FPD n ;

[0077] In the formula, α and β are the weights of the image - space uncertainty and feature - space uncertainty respectively, and α + β = 1.

[0078] S43. Obtain the importance of the preliminary image features based on the machine learning model, and perform weighted calculations with the corresponding comprehensive uncertainties respectively to obtain the comprehensive evaluation index.

[0079] In this embodiment, a machine learning model (such as random forest, support vector machine, etc.) is used for training, and the importance of the features is evaluated according to the output of the model (such as accuracy, F1 score, etc.) or the internal mechanism of the model (such as the branch structure of the decision tree, the weight coefficient of the support vector machine, etc.).

[0080] Set a weight coefficient, and perform weighted summation of the uncertainty and importance to obtain the comprehensive evaluation index. The selection of the weight can be adjusted according to the specific situation of the problem.

[0081] S44. Sort the preliminary image features according to the comprehensive evaluation index, and select the top q% of the preliminary image features as the preferred image features.

[0082] On the other hand, as Figure 3 shown, the present invention also proposes a medical image feature extraction system for implementing the above medical image feature extraction method, including:

[0083] A data preprocessing module for collecting medical image data, preprocessing the medical image data, and constructing a data set;

[0084] A model construction module for constructing and training an adaptive feature extraction model to obtain the best feature extraction model;

[0085] A preliminary feature extraction module for inputting the data set into the adaptive feature extraction model to obtain preliminary image features;

[0086] A feature screening module for screening the preliminary image features based on the uncertainty perception mechanism, and determining the preferred image features by combining the importance and uncertainty of the preliminary image features.

[0087] Preferably, the model construction module includes:

[0088] A parameter optimization unit for taking the minimization of the loss of the cross-entropy function as the optimization goal, and optimizing the model parameters according to the optimization goal by the Adam algorithm;

[0089] A data distribution optimization unit for optimizing the data distribution in the training process of the adaptive feature extraction model by using the BatchNorm technology.

[0090] Preferably, the feature screening module includes:

[0091] An uncertainty calculation unit for calculating the image space uncertainty and the feature space uncertainty of each preliminary image feature;

[0092] A comprehensive uncertainty calculation unit is configured to determine the comprehensive uncertainty of the preliminary image features based on the image space uncertainty and the feature space uncertainty;

[0093] A comprehensive evaluation index acquisition unit is configured to obtain the importance of the preliminary image features based on a machine learning model, and perform weighted calculations with the corresponding comprehensive uncertainties respectively to obtain a comprehensive evaluation index;

[0094] An optimal feature image determination unit is configured to rank the preliminary image features according to the comprehensive evaluation index, and select the top q% of the preliminary image features as the optimal image features.

[0095] In the present specification, the various embodiments are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The same or similar parts among the various embodiments can be referred to each other. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple. For the relevant parts, reference can be made to the description in the method part.

[0096] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for extracting medical image features, characterized in that, Including the following steps: Collect medical image data, preprocess the medical image data, and construct a data set; Construct and train an adaptive feature extraction model to obtain an optimal feature extraction model, including: Construct an adaptive feature extraction model based on the U-Net network, and optimize the adaptive feature extraction model through the Adam algorithm and based on the cross-entropy loss function to obtain an optimal feature extraction model; Input the data set into the adaptive feature extraction model to obtain preliminary image features; Screen the preliminary image features based on the uncertainty perception mechanism, and determine the preferred image features by combining the importance and uncertainty of the preliminary image features, including: Calculate the image space uncertainty and feature space uncertainty of each preliminary image feature; For the n-th preliminary image feature, the image space uncertainty CV n The calculation formula is as follows: |N(x,y)| represents the number of pixels in the neighborhood N(x,y) of the pixel (x,y), μ(x,y) is the mean value of all pixel values in the neighborhood N(x,y), and I(i,j) is the pixel value of the image at the position (i,j); For the m feature points of the nth preliminary image feature, the feature space uncertainty FPD n is calculated as follows: where d(p l , p k ) is the distance between feature points, and σ is a parameter for controlling the distance sensitivity; Determine the comprehensive uncertainty of the preliminary image features according to the image space uncertainty and the feature space uncertainty; The calculation formula of the comprehensive uncertainty A of the nth preliminary image feature is as follows: A = α·CV n + β·FPD n ; In the formula, α and β are the image space uncertainty weight and feature space uncertainty weight respectively; Obtain the importance of the preliminary image features based on the machine learning model, and perform weighted calculations with the corresponding comprehensive uncertainties respectively to obtain a comprehensive evaluation index; Sort the preliminary image features according to the comprehensive evaluation index, and select the top q% of the preliminary image features as the preferred image features.

2. The medical image feature extraction method according to claim 1, characterized in that Optimize the adaptive feature extraction model through the Adam algorithm and based on the cross-entropy loss function to obtain an optimal feature extraction model, including: Taking the minimization of the loss of the cross-entropy function as the optimization goal, the Adam algorithm optimizes the model parameters according to the optimization goal; Use the BatchNorm technique to optimize the data distribution during the training process of the adaptive feature extraction model.

3. A medical image feature extraction system, characterized in that, Including: A data preprocessing module for collecting medical image data, preprocessing the medical image data, and constructing a data set; A model construction module for constructing and training an adaptive feature extraction model to obtain an optimal feature extraction model, including: Construct an adaptive feature extraction model based on the U-Net network, and optimize the adaptive feature extraction model through the Adam algorithm and based on the cross-entropy loss function to obtain an optimal feature extraction model; A preliminary feature extraction module for inputting the data set into the adaptive feature extraction model to obtain preliminary image features; A feature screening module for screening the preliminary image features based on the uncertainty perception mechanism, and determining the preferred image features by combining the importance and uncertainty of the preliminary image features; The feature screening module includes: An uncertainty calculation unit for calculating the image space uncertainty and feature space uncertainty of each preliminary image feature; For the nth preliminary image feature, the image spatial uncertainty CV n The calculation formula is as follows: |N(x, y)| represents the number of pixels in the neighborhood N(x, y) of the pixel (x, y), μ(x, y) is the mean value of all pixel values in the neighborhood N(x, y), and I(i, j) is the pixel value of the image at the position (i, j); For the m feature points of the nth preliminary image feature, the feature space uncertainty FPD n is calculated as follows: where d(p l , p k ) is the distance between feature points, and σ is a parameter for controlling the distance sensitivity; A comprehensive uncertainty calculation unit, configured to determine the comprehensive uncertainty of the preliminary image features according to the image space uncertainty and the feature space uncertainty; The calculation formula for the comprehensive uncertainty A of the nth preliminary image feature is as follows: A = α·CV n + β·FPD n ; In the formula, α and β are the image space uncertainty weight and the feature space uncertainty weight respectively; A comprehensive evaluation index acquisition unit, configured to obtain the importance of the preliminary image features based on a machine learning model, and perform weighted calculations with the corresponding comprehensive uncertainties respectively to obtain a comprehensive evaluation index; A preferred feature image determination unit, configured to rank the preliminary image features according to the comprehensive evaluation index, and select the top q% of the preliminary image features as the preferred image features.

4. A medical image feature extraction system according to claim 3, wherein, The model construction module includes: A parameter optimization unit, configured to use the minimization of the loss of the cross-entropy function as the optimization target, and the Adam algorithm optimizes the model parameters according to the optimization target; A data distribution optimization unit, configured to optimize the data distribution in the training process of the adaptive feature extraction model by using the BatchNorm technology.

Images