Medical image automatic classification method and system based on machine learning

Abstract

The invention relates to the technical field of medical image classification, in particular to a medical image automatic classification method and system based on machine learning, and the method comprises the steps: collecting medical image data, and enhancing the number of the medical image data; carrying out dimension reduction processing on the enhanced medical image data; and classifying the dimensionality-reduced features to obtain medical image data categories. A generative adversarial network based on quantum interference is used for data expansion, complex modes and textures of real medical images are simulated through multi-dimensional modulation and control, the number and diversity of medical image data are increased, and meanwhile the detail quality of the images is maintained. Dimensionality reduction intensity is adjusted in a self-adaptive mode, and information storage is optimized. The classifier uses a fractional order neural network based on a dynamic penalty term, combines the dynamic penalty term and a gating vector regularization technology, and optimizes the classification precision and the generalization ability of the model. The defects in the prior art are overcome.

Description

Technical Field

[0001] The present invention relates to the technical field of medical image classification, and in particular to a method and system for automatic medical image classification based on machine learning. Background Art

[0002] With the development of medical technology, medical imaging plays a vital role in clinical practice. However, traditional image classification relies on the professional judgment of radiologists, which is not only time-consuming and labor-intensive, but also prone to human errors when faced with large amounts of imaging data.

[0003] The invention patent with the publication number CN113744845A proposes a medical image processing method, device, electronic device and medium based on artificial intelligence. The method includes: in response to the processing instructions of multiple medical images of the target patient, obtaining the recognition task corresponding to each medical image; determining the image factors of each recognition task; constructing the image factor hierarchy model according to the hierarchical relationship corresponding to the image factors; obtaining the image preset judgment matrix, and calculating the combination weight corresponding to the recognition task based on the judgment matrix and the image factor hierarchy model; generating a task scheduling set of multiple recognition tasks according to the combination weight corresponding to the recognition task; sequentially obtaining the target medical image corresponding to each recognition task in the task scheduling set in the preset storage space; calling the AI ​​edge computing device to process the sequentially obtained target medical images. The processing efficiency of medical images is improved. The invention patent with publication number CN111210414B proposes a medical image analysis method, a computer device and a readable storage medium. The method includes: obtaining a medical image to be analyzed; inputting the medical image into a preset classification model to obtain a classification result of the medical image; performing a reduction operation: performing a reduction operation on the medical image to obtain a reduced medical image; inputting the reduced medical image into the classification model to obtain a classification result of the reduced medical image; if the accuracy of the classification result of the reduced medical image is greater than the accuracy of the classification result of the medical image, the reduced medical image is determined as the region of interest of the medical image; if the accuracy of the classification result of the reduced medical image is not greater than the accuracy of the classification result of the medical image, the direction of the reduction operation is changed, and the reduction operation is repeated. The method improves the accuracy of the determined region of interest.

[0004] However, the above patent introduces machine learning methods into medical image analysis, which improves processing speed and accuracy, but there are still some key problems. First, due to the high dimensionality and complexity of medical images, the existing machine learning models need to be improved in feature processing and classification accuracy. In addition, the existing technology still has deficiencies in data expansion, model training stability, and generalization ability. Summary of the invention

[0005] The purpose of the present invention is to provide a method and system for automatic classification of medical images based on machine learning, which solves the problems that the prior art still has deficiencies.

[0006] To achieve the above objectives,

[0007] In a first aspect, the present invention provides a method for automatic classification of medical images based on machine learning, comprising the following steps:

[0008] Collect medical imaging data and increase the amount of medical imaging data, wherein the amount of medical imaging data is increased through a generative adversarial network algorithm based on quantum interference;

[0009] Performing dimensionality reduction processing on the quantity of the extracted and enhanced medical image data; wherein an autoencoder algorithm based on a relaxation factor is used as a feature dimensionality reduction model;

[0010] Classify the features after dimensionality reduction to obtain the categories of medical imaging data.

[0011] Furthermore, the training step of the quantum interference-based generative adversarial network algorithm includes:

[0012] Initialize the generator and discriminator of the generative adversarial network;

[0013] The generator simulates a plurality of medical image states through a quantum state preparation stage, wherein the medical image states include images of health conditions or pathological characteristics; through a quantum interference simulation strategy, the medical image states are coherently superimposed to generate first medical image data with new characteristics;

[0014] The generator outputs first medical image data; the first medical image data is slightly disturbed by the pixels of the image through the chaotic sequence generation module to increase the diversity of the first medical image data;

[0015] The discriminator learns to identify the difference between the first medical image data and the real medical image data, adjusts parameters according to the difference and improves recognition accuracy;

[0016] The iteration is repeated until the first medical image data generated by the generator can make it difficult for the discriminator to distinguish the authenticity.

[0017] Furthermore, in the repeated iterative process, the back propagation algorithm is used to update the generator parameters and the discriminator parameters, and the updating method is expressed as:

[0018]

[0019]

[0020] In the formula, is the generator parameter updated for this iteration, is the discriminator parameter updated in this iteration, is the learning rate of the generative adversarial network, and They represent the gradients of the loss function of the generative adversarial network with respect to the generator parameters and the discriminator parameters, respectively. is the learning rate of the discriminator, and are the weight parameters of the generator and the discriminator respectively;

[0021] Furthermore, a dynamic adjustment strategy is adopted to adjust the learning rate of the discriminator Make adjustments, set To generate the gradient of the loss function of the adversarial network with respect to the discriminator parameters, it is expressed as:

[0022]

[0023] Among them, the calculation method of the discriminator's learning rate is expressed as:

[0024]

[0025] In the formula, represents the rate of change of the discriminator's historical gradient, is the adjustment sensitivity parameter of the discriminator, which controls the response speed of the adjustment factor, and e is a natural number;

[0026] During the training process, the quality and diversity of the generated first medical image data are continuously evaluated until the iteration is stopped when the evaluation function value is greater than a preset value.

[0027] Furthermore, the first medical image data is diversified; specifically, it is expressed as follows:

[0028]

[0029]

[0030] In the formula, is the generator function; is the output image of the generator; represents the chaotic perturbation function applied to the image, is the coefficient that adjusts the intensity of chaotic disturbance; The final image generated by the generator in this iteration; It is the quantum state after quantum interference simulation.

[0031] Furthermore, the calculation of the chaotic perturbation function applied to the image utilizes a combination of the logical mapping and the hyperbolic tangent function in chaos theory.

[0032] Furthermore, based on the relaxation factor autoencoder algorithm model, the training steps include:

[0033] Initializing an autoencoder network structure, the autoencoder includes an encoder and a decoder, wherein the encoder maps a high-dimensional medical image data feature vector to a low-dimensional representation space, and the decoder reconstructs high-dimensional input data from the low-dimensional space;

[0034] Setting an initial value of a dynamic relaxation factor and its adjustment strategy, wherein the relaxation factor is adjusted during the training process according to error feedback;

[0035] The input medical image data feature vector is converted into a representation of a low-dimensional space through an encoder, and then reconstructed into an output close to the high-dimensional input data through a decoder;

[0036] Calculate the loss function of the autoencoder, including reconstruction error, dimensionality reduction error, and regularization term;

[0037] Perform back propagation according to the loss function to update the weights and bias parameters in the autoencoder network;

[0038] Repeat the above steps until the preset stop iteration condition is met.

[0039] Furthermore, based on the relaxation factor autoencoder algorithm model, the training steps also include:

[0040] The calculation method of the regularization term of the loss function is expressed as:

[0041]

[0042] In the formula, and Represent the Frobenius norm and L1 norm, respectively, for the regularization of weights and biases, is the first regularization coefficient of the autoencoder, For encoder The weight matrix of the layer; For encoder The bias of the layer, is the second regularization coefficient of the autoencoder, is the total number of layers of the encoder, MRr represents the cumulative error through all layers of the encoder, and Rr is the regularization term of the autoencoder;

[0043] The calculation method of the accumulated error through all layers of the encoder is expressed as:

[0044]

[0045]

[0046] In the formula, Indicates The forward propagation function of the layer, For encoder The output of the hidden layer, For encoder The output of the hidden layer, For encoder The activation function of the layer, is the weight coefficient of the local reconstruction error, It is a local mapping of low-dimensional space, an adaptive low-dimensional representation, expressed as:

[0047]

[0048] In the formula, is a local mapping of the low-dimensional space, is the weight matrix of the encoder; is the encoder bias, is the sample set input into the autoencoder for the current batch, For the Neighborhood points, For encoder The activation function of the layer, For the Medical image data points for The weighted coefficients of neighbor points satisfy , the weighted coefficient calculation method is expressed as:

[0049]

[0050] In the formula, It is the scale parameter of the distance, which controls the image range of the weighted coefficient of the point pairs within the neighborhood.

[0051] Furthermore, the features after dimensionality reduction are classified to obtain medical image data categories, and the steps further include:

[0052] Initialize the parameters of the fractional neural network;

[0053] The medical image data features that have undergone dimension reduction processing are input into the fractional-order neural network, processed by the fractional-order activation function, and transmitted to the output layer layer by layer;

[0054] Calculate the loss function based on the actual output and the expected output;

[0055] The gradient of the loss function for each parameter is calculated through the back-propagation algorithm, and the weights and biases of the fractional neural network are updated;

[0056] The fractional order parameters are adjusted according to the loss function to optimize the learning efficiency and accuracy of the model. The adjustment method is expressed as:

[0057]

[0058] In the formula, is the fractional-order adjustment coefficient, which is used to control the update speed and amplitude of the fractional-order parameters. is the gradient of the loss function with respect to the fractional-order parameter, which is used to guide the adaptive adjustment of the fractional-order parameter. It is a fractional order neural network. The fractional order of the layer activation function;

[0059] Repeat the above steps until the preset stop iteration condition is met.

[0060] The second aspect is the automatic classification system of medical images based on machine learning, including:

[0061] A medical image data acquisition unit, the medical image data acquisition unit is used to collect medical image data and increase the amount of medical image data;

[0062] A medical image data mining unit, the medical image data mining unit is used to perform dimensionality reduction processing on the enhanced medical image data;

[0063] A medical image data classification unit; the medical image data classification unit is used to classify the features after dimensionality reduction to obtain medical image data categories.

[0064] The present invention discloses a method and system for automatic classification of medical images based on machine learning. The data expansion uses a generative adversarial network based on quantum interference, which simulates the complex patterns and textures of real medical images through modulation and control in multiple dimensions, increases the quantity and diversity of medical image data, and maintains the quality of image details. The feature extraction uses a neural network based on dynamic adaptive oscillation, which optimizes the exploration of high-dimensional parameter space by simulating nonlinear oscillation behavior in physical phenomena, allowing the neural network to effectively utilize local extreme values. The feature dimensionality reduction uses an autoencoder based on a relaxation factor, which adjusts the training intensity of the autoencoder to meet the training requirements of different stages, adaptively adjusts the dimensionality reduction intensity, and optimizes the preservation of information. The classifier uses a fractional-order neural network based on a dynamic penalty term, combined with a dynamic penalty term and a gated vector regularization technique, to optimize the classification accuracy and the generalization ability of the model. The shortcomings of the prior art are solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art are briefly introduced below.

[0066] Figure 1 is a flow chart of a method for automatic classification of medical images based on machine learning according to a first embodiment of the present invention;

[0067] Figure 2 is a principle block diagram of a medical image automatic classification system based on machine learning according to a second embodiment of the present invention;

[0068] Figure 3 It is an experimental effect diagram that reveals the coupling effect between parameters by scanning the three-dimensional parameter space and verifies the robustness of the algorithm to key hyperparameters.

[0069] In the figure: 201 - medical image data acquisition unit, 202 - medical image data mining unit, 203 - medical image data classification unit. DETAILED DESCRIPTION

[0070] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0071] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0072] The first embodiment of the present application is:

[0073] See also Figure 1 and Figure 2 ,in, Figure 1 It is a training flow chart of a neural network algorithm based on dynamic adaptive oscillation according to the first embodiment of the present invention. Figure 2 It is a flowchart of the steps of the automatic classification method of medical images based on machine learning according to the first embodiment of the present invention.

[0074] The present invention provides a method for automatic classification of medical images based on machine learning, comprising the following steps:

[0075] S101: Collect medical imaging data and increase the amount of medical imaging data;

[0076] Specifically, the source of the medical image data collection of the present invention is a medical image database, including but not limited to CT, MRI and X-ray images, and the medical image data collection complies with the privacy protection standards of the Health Insurance Portability and Accountability Act to ensure the anonymization of patient information;

[0077] The medical image data acquisition method of the present invention includes exporting medical image data directly from the hospital's image storage and transmission system, and collecting through a cooperative medical team in specific diagnosis and treatment activities, and all medical image data are stored in DICOM (Digital Imaging and Communications Standard) format;

[0078] The medical imaging data acquisition type of the present invention is grayscale image, with a pixel depth of 16 bits and a resolution of not less than 256x256 pixels. The acquisition content focuses on common disease hallmarks, such as tumors, inflammation and other signs;

[0079] The medical image data annotation categories of the present invention cover multi-level classification from general anatomical structures to specific pathological changes, specifically organ types under different pathological conditions. The annotation process is completed by professional doctors to ensure the accuracy of the annotation;

[0080] It is understandable that in the task of the present invention, the collection, acquisition, annotation and preprocessing of medical image training data are time-consuming and labor-intensive, and insufficient training samples can easily lead to poor generalization ability of the model and affect the accuracy of the model. The present invention uses a generative adversarial network algorithm based on quantum interference to generate samples, thereby realizing medical image data expansion;

[0081] Quantum interference modulates and controls medical imaging data in multiple dimensions to mimic the complex patterns and textures in real medical images, so that the generated images can be closer to the statistical characteristics of real images in detail.

[0082] Specifically, the training process of the quantum interference-based generative adversarial network algorithm is as follows:

[0083] 1. Initialize the generator and discriminator of the generative adversarial network. The initialization method is expressed as:

[0084] ,

[0085] In the formula, To obey a specific distribution; and are the weight parameters of the generator and the discriminator respectively, represents the initialization variance of the weight; It means the mean is 0 and the variance is Normal distribution of Represents a normal distribution.

[0086] 2. The generator first simulates a variety of possible medical imaging states through the quantum state preparation stage. Each state represents an image of a specific health condition or pathological characteristics. Based on the quantum interference strategy, these states will be coherently superimposed to generate medical images with new characteristics. The quantum interference simulation method is expressed as:

[0087] , In the formula, Representing quantum states, specifically, expressing different possible states of medical images through quantum bits; is the quantum state before quantum interference simulation; It is the quantum state after quantum interference simulation; represents the operating unit of the quantum interferometer; is the control parameter of the quantum interferometer; and are the amplitude modulation and phase modulation parameters respectively, is the index of the characteristic dimension of the quantum state; Is an imaginary unit.

[0088] Furthermore, the calculation method of the amplitude modulation and phase modulation parameters is expressed as:

[0089] , In the formula, It is the angle parameter dynamically generated by the generator network based on the input medical imaging data; From a uniform distribution The values ​​randomly sampled in simulate the randomness of the quantum phase and increase the detail richness of the generated image.

[0090] 3. Based on the results of quantum interference, the generator outputs preliminary medical image data, which then passes through a chaotic sequence generation module, which slightly perturbs the pixels of the image to increase the diversity of the image, expressed as:

[0091] , In the formula, is the generator function; is the output image of the generator; represents the chaotic perturbation function applied to the image, is the coefficient that adjusts the intensity of chaotic disturbance; This is the final image generated by the generator in this iteration.

[0092] Furthermore, the calculation of the chaotic perturbation function utilizes the combination of the logical mapping and the hyperbolic tangent function in chaos theory to enhance the randomness and complexity of the generated image. The calculation method is expressed as: In the formula, is the disturbance amplitude control coefficient, is a parameter that adjusts the intensity of nonlinear perturbations, represents element-wise multiplication; is a hyperbolic tangent function. Preferably, Set to 0.2.

[0093] 4. The task of the discriminator is to distinguish between the generated images and the real images. It learns to identify the subtle differences between the generated images and the real images and continuously adjusts its parameters to improve the recognition accuracy. In the adversarial training process, the loss function is calculated as follows: In the formula, is the loss function for generating adversarial networks; express expectations; is the real medical imaging data distribution, is to generate the medical imaging data distribution, Represents a real medical image; Medical imaging data distribution representing real medical images; Represents the medical imaging data distribution of the final image generated by the generator in this iteration; is the discriminator function.

[0094] 5. In each iteration, the performance of the generator and the discriminator will be evaluated, and the parameters will be adjusted according to the feedback of the discriminator. This process is repeated until the generator can produce images of sufficient quality so that the discriminator cannot distinguish between true and false. The back propagation algorithm is used to update the parameters of the generator and the discriminator. The update method is expressed as:

[0095] ,

[0096] In the formula, The generator parameters updated for this iteration; The discriminator parameters updated for this iteration; is the learning rate of the generative adversarial network, and Respectively represent the gradients of the loss function of the generative adversarial network with respect to the generator parameters and the discriminator parameters; is the learning rate of the discriminator.

[0097] Furthermore, a dynamic adjustment strategy is used to adjust the learning rate of the discriminator, and the speed of gradient descent is adaptively adjusted according to the characteristics of the input medical imaging data in each iteration, thereby improving the stability and efficiency of training. To generate the gradient of the loss function of the adversarial network with respect to the discriminator parameters, it is expressed as:

[0098]

[0099] Furthermore, the calculation method of the discriminator's learning rate is expressed as:

[0100]

[0101] In the formula, Represents the rate of change of the discriminator's historical gradient; is the adjustment sensitivity parameter of the discriminator, controlling the response speed of the adjustment factor. Preferably, Set to 2.

[0102] Furthermore, the calculation method of the rate of change of the historical gradient of the discriminator is expressed as:

[0103] In the formula, For the The gradient of the loss function of the generative adversarial network with respect to the discriminator parameters at this iteration; For the The gradient of the loss function of the generative adversarial network with respect to the discriminator parameters at this iteration.

[0104] 6. During the training process, the quality and diversity of the generated images are continuously evaluated to ensure that the newly generated medical image data matches the real medical image data in terms of visual and statistical characteristics. Finally, the quality of the generated images is evaluated to ensure that the classification quality standards are met. The evaluation method is expressed as:

[0105]

[0106] In the formula, To generate an image quality evaluation function, when the evaluation function value is greater than the preset value, it means that the generated medical image data meets the classification quality standard. Then the iteration is stopped, otherwise it continues to iterate; and are the pixel values ​​of the generated image and the reference real image respectively, is the total number of pixels in the image; is the index of the image pixel.

[0107] After the training of the medical image data expansion model is completed, the trained medical image data expansion model is used to increase the number of samples. In one embodiment, assuming that the original collected samples are 800, and the medical image data expansion model expands and generates 200 samples, the expanded medical image data set contains 1000 samples.

[0108] S102: performing dimensionality reduction processing on the enhanced medical imaging data;

[0109] Specifically, the medical imaging data is further compressed through feature dimensionality reduction to achieve full mining of the medical imaging data.

[0110] The present invention adopts an autoencoder based on a relaxation factor as a feature dimensionality reduction model, wherein the autoencoder realizes compression and reconstruction of medical image data through an encoder and a decoder, and aims to learn the optimal representation of medical image data in a low-dimensional space;

[0111] Different from traditional autoencoders, the present invention adopts a dynamic relaxation factor in the structure of the autoencoder, allowing the autoencoder model to adaptively adjust the dimensionality reduction strength during the training process to ensure maximum information retention, optimize the generalization ability of the autoencoder model, and reduce the risk of overfitting.

[0112] Specifically, the training process of the autoencoder algorithm based on the relaxation factor is as follows:

[0113] 1. Initialize the network structure of the autoencoder. The autoencoder includes two parts: the encoder and the decoder. The encoder is responsible for mapping the high-dimensional input data to a low-dimensional representation space. The high-dimensional input data is the feature vector of medical image data. The decoder reconstructs the high-dimensional input data from this low-dimensional space. The parameter update method of the encoder is expressed as follows:

[0114] In the formula, For encoder The weight matrix of the layer; For encoder Bias of the layer; Indicates compliance with a specific distribution; represents normal distribution; represents the initialization variance.

[0115] 2. Before model training, set the initial value of the dynamic relaxation factor and its adjustment strategy. The dynamic relaxation factor is adjusted according to error feedback during the training process to adapt to the training needs of different stages. The adjustment method is expressed as:

[0116] In the formula, For the relaxation factor for each training cycle; is the initial relaxation factor; is the relaxation factor decay rate; Indicates the current iteration number. Preferably, Set to 1, Set to 0.5.

[0117] 3. The input medical image data feature vector is converted into a low-dimensional space representation through the encoder, and then these low-dimensional representations are reconstructed into outputs close to the high-dimensional input data through the decoder. The conversion process from the input layer to the hidden layer can be expressed as:

[0118] In the formula, For encoder The output of the hidden layer; The feature vector of medical imaging data input to the encoder; For encoder The activation function of the layer.

[0119] 4. Calculate the loss function of the autoencoder to evaluate the performance of the current autoencoder network. The loss function of the autoencoder includes reconstruction error, dimensionality reduction error and regularization term. The calculation method is expressed as:

[0120] In the formula, is the loss function of the autoencoder; is the reconstructed data output by the decoder, is the feature vector of medical imaging data input to the encoder, For the The relaxation factor for each training cycle, represents the deviation between the encoder output and its ideal low-dimensional representation; is the L2 norm; is the regularization term of the autoencoder.

[0121] Furthermore, the calculation method of the regularization term of the loss function is expressed as:

[0122]

[0123] In the formula, and Represent the Frobenius norm and L1 norm, respectively, which are used for regularization of weights and biases; is the first regularization coefficient of the autoencoder, For encoder The weight matrix of the layer; For encoder The bias of the layer, is the second regularization coefficient of the autoencoder; is the total number of layers of the encoder; represents the accumulated error through all layers of the encoder.

[0124] Furthermore, the calculation method of the accumulated error through all layers of the encoder is expressed as:

[0125]

[0126]

[0127] In the formula, Indicates The forward propagation function of the layer; For encoder The output of the hidden layer, For encoder The output of the hidden layer, For encoder The activation function of the layer, is the weight coefficient of local reconstruction error; is a local mapping of the low-dimensional space. Preferably, Set to 0.2.

[0128] Furthermore, since the traditional autoencoder does not fully consider the local characteristics of different medical image data points in the low-dimensional space, the medical image data points are samples corresponding to the medical image data. The present invention adopts a local adaptive reconstruction method based on the low-dimensional subspace. In the process of dimensionality reduction, the local geometric features of each medical image data point are adaptively adjusted to represent it in the low-dimensional space, thereby strengthening the model's recognition and retention of the local structure of the medical image data and improving the accuracy and flexibility of feature dimensionality reduction. Specifically, the local mapping of the medical image data point in the low-dimensional space is defined as , this mapping is not only generated by a global dimensionality reduction strategy, but also by weighted reconstruction of the features of all medical imaging data points in the sample set input into the autoencoder in the current batch to obtain an adaptive low-dimensional representation, which is expressed as:

[0129]

[0130] In the formula, is a local mapping of the low-dimensional space, is the weight matrix of the encoder; is the encoder bias, The sample set input into the autoencoder for the current batch; For the Neighbor points; For encoder The activation function of the layer; For the Medical image data points for The weighted coefficients of neighbor points satisfy .

[0131] Furthermore, the weighting coefficient adopts a distance-based weighting strategy, so that the neighbor medical image data points closer to the medical image data point contribute more to its local reconstruction, and the neighbor medical image data points farther away contribute less. The calculation method is expressed as:

[0132]

[0133] In the formula, is the scale parameter of the distance, which controls the influence range of the points in the neighborhood on the weighted coefficient; exp() is the natural exponential function. Preferably, Set to 0.1.

[0134] 5. Perform back propagation according to the loss function to update the weights and bias parameters in the autoencoder network. The gradient descent update method for the weights and biases of the encoder is expressed as:

[0135]

[0136]

[0137] In the formula, is the learning rate of the autoencoder; and are the gradients of the autoencoder’s loss function with respect to weights and biases, respectively.

[0138] 6. Repeat the above steps until the preset stop iteration condition is met, indicating that the model training is completed. In one embodiment, the preset stop iteration condition is reaching a preset maximum number of iterations. Preferably, the preset maximum number of iterations is set to 1000 times.

[0139] S103: Classify the features after dimensionality reduction to obtain medical impact data categories.

[0140] Specifically, the medical imaging data after feature dimension reduction is classified through the classifier model to obtain the medical impact data category;

[0141] The present invention adopts a fractional-order neural network based on a dynamic penalty term to classify the medical imaging data after feature dimensionality reduction. The dynamic penalty term is to impose penalty constraints on the weights of the fractional-order neural network during the training process of the fractional-order neural network, so as to adjust the fractional-order parameters and optimize the classification accuracy.

[0142] Specifically, the training process of the fractional-order neural network based on dynamic penalty term is as follows:

[0143] 1. Initialize the parameters of the fractional-order neural network. The initialization method is expressed as:

[0144]

[0145]

[0146] In the formula, represents the fractional-order neural network The weight of the layer; represents the fractional-order neural network Bias of the layer; is the variance of the weight initialization of the fractional neural network; Indicates compliance with a specific distribution; represents a normal distribution. Preferably, Set to 0.01.

[0147] 2. Input the features of medical image data after dimensionality reduction into the fractional-order neural network, process it through the fractional-order activation function, and pass it to the output layer layer by layer, which is expressed as:

[0148]

[0149] In the formula, It is a fractional order neural network. The output of the layer; It is a fractional order neural network. The output of the layer; is a fractional-order activation function; It is a fractional order neural network. The fractional order of the layer activation function; represents element-wise multiplication; The fractional-order neural network The gating vector for the layer.

[0150] Furthermore, an adaptive feature recognition gating mechanism is used to enhance the feature selection capability in the classifier model training process. Specifically, a gating vector is calculated for each layer in the classifier fractional-order neural network, and its element value determines the transmission strength of the corresponding feature. The calculation method is expressed as:

[0151]

[0152] In the formula, is the Sigmoid function; and They are the gate weight parameter and the gate bias parameter, which are training parameters used to learn which features are most important for the current task.

[0153] 3. The loss function is calculated according to the actual output and expected output of the fractional-order neural network. The present invention adopts a loss function containing a dynamic penalty term to adjust the fractional-order parameters and optimize the classification accuracy. At the same time, the L1 regularization term of the gate vector is used to promote the generalization ability of the model. The calculation method is expressed as:

[0154]

[0155] In the formula, is the loss function of the fractional-order neural network; It is One-hot encoding of the true labels of the samples; is a fractional order neural network. Classification output of samples; is the number of layers of the fractional-order neural network; It is the regularization parameter of the fractional-order neural network, which controls the weight decay to prevent overfitting; Indicates The square of the Frobenius norm of the layer weights; The weight of the gate vector regularization is used to balance the model complexity and performance; is the L1 norm. Preferably, Set to 0.2, Set to 0.3.

[0156] 4. Through the back-propagation algorithm, the gradient of the loss function for each parameter is calculated, and the weights and biases of the fractional-order neural network are updated, which can be expressed as:

[0157]

[0158]

[0159] In the formula, It is the parameter update operation; is the learning rate of the fractional-order neural network; and is the gradient of the loss function of the fractional-order neural network with respect to weights and biases.

[0160] 5. Adjust the fractional order parameters according to the loss function to optimize the learning efficiency and accuracy of the model. The adjustment method is expressed as:

[0161]

[0162] In the formula, is the fractional-order adjustment coefficient, which is used to control the update speed and amplitude of the fractional-order parameters; is the gradient of the loss function with respect to the fractional-order parameter, which is used to guide the adaptive adjustment of the fractional-order parameter. Preferably, Set to 0.2.

[0163] 6. Repeat the above steps until the preset stop iteration condition is met, indicating that the model training is completed. In one embodiment, the preset stop iteration condition is reaching a preset maximum number of iterations. Preferably, the preset maximum number of iterations is set to 1000 times.

[0164] After the training is completed, the trained machine learning model is used to process new medical image samples to achieve automatic classification. The classification result is the output category of the medical image data classification unit.

[0165] Data augmentation uses a generative adversarial network based on quantum interference to simulate the complex patterns and textures of real medical images through modulation and control in multiple dimensions, increasing the quantity and diversity of medical imaging data while maintaining the quality of image details. Feature dimensionality reduction uses an autoencoder based on a relaxation factor. By adjusting the training intensity of the autoencoder to meet the training needs of different stages, the dimensionality reduction intensity is adaptively adjusted to optimize information preservation. The classifier uses a fractional-order neural network based on a dynamic penalty term, combined with a dynamic penalty term and gated vector regularization technology to optimize classification accuracy and the generalization ability of the model. It solves the shortcomings of existing technologies.

[0166] The second embodiment of the present application is:

[0167] Based on the first embodiment, please refer to Figure 2 ,in, Figure 2 It is a principle block diagram of a medical image automatic classification system based on machine learning according to the second embodiment of the present invention.

[0168] The medical image automatic classification system based on machine learning in this embodiment includes a medical image data acquisition unit 201 , a medical image data mining unit 202 and a medical image data classification unit 203 .

[0169] According to this specific implementation, the medical image data mining unit 202 is connected to the medical image data acquisition unit 201 and the medical image data classification unit 203 respectively;

[0170] The medical image data acquisition unit 201 is used to collect medical image data and increase the amount of medical image data;

[0171] The medical image data mining unit 202 is used to extract features from the enhanced medical image data and perform dimensionality reduction processing on the extracted features;

[0172] The medical image data classification unit 203 is used to classify the features after dimensionality reduction to obtain disease categories.

[0173] In the present embodiment, a medical image automatic classification system based on machine learning is used. The medical image data acquisition unit 201 mainly collects and enhances medical image data. The enhancement method of medical image data is to increase the number of medical image data by sample expansion; and further compress the medical image data to fully mine the medical image data. The medical image data classification unit 203 classifies the medical image data after feature dimension reduction through a classifier model, and then obtains the medical image data category. The medical image data acquisition unit 201, the medical image data mining unit 202, and the medical image data classification unit 203 perform model training according to the above-mentioned medical image automatic classification method based on machine learning. After the model training is completed, the trained machine learning model is used to process new medical image samples to achieve automatic classification, and the classification result is the output category of the medical image data classification unit 203. Data expansion uses a generative adversarial network based on quantum interference to simulate the complex patterns and textures of real medical images through modulation and control in multiple dimensions, increase the quantity and diversity of medical image data, and maintain the quality of image details. The feature dimensionality reduction uses an autoencoder based on a relaxation factor. By adjusting the training intensity of the autoencoder to meet the training requirements of different stages, the dimensionality reduction intensity is adaptively adjusted to optimize the preservation of information. The classifier uses a fractional-order neural network based on a dynamic penalty term, combined with a dynamic penalty term and gated vector regularization technology to optimize the classification accuracy and generalization ability of the model. This solves the shortcomings of the existing technology.

[0174] Experimental example

[0175] like Figure 3 As mentioned above, in order to deeply study the effectiveness of joint optimization of hyperparameters of the autoencoder algorithm based on relaxation factor, in the dynamic relaxation factor autoencoder, the relaxation factor decay rate and the learning rate are the core hyperparameters, which jointly control the parameter update amplitude (learning rate directly affects the step size of gradient descent), the change of regularization intensity (relaxation factor decay rate determines the time-varying characteristics of the regularization term), and the exploration and development balance (dynamic relaxation factor adjustment model strategy during training). Therefore, in this experiment, the coupling effect between parameters is revealed by scanning the three-dimensional parameter space, the robustness of the algorithm to key hyperparameters is verified, and the optimal parameter combination is located.

[0176] In this experiment, the range of parameters is chosen as:

[0177] Relaxation factor decay rate ∈ [0.1, 1.0]: covers the full scene from slow decay to fast decay;

[0178] Learning rate∈[0.001,0.01]: contains the typical learning rate range for deep learning;

[0179] Sampling point 50×50: ensure parameter space resolution.

[0180] The loss surface modeling is expressed as:

[0181] Theoretical terms: The first two terms reflect the independent effects of parameters (increasing learning rate reduces loss, and increasing learning rate reduces loss);

[0182] Coupling term: nonlinear interaction between simulation parameters;

[0183] Noise term: simulates actual training fluctuations.

[0184] In this experiment, the low-high loss areas are visually displayed using a blue-red gradient through hot-cold color mapping, and the global minimum point is experimentally highlighted.

[0185] In terms of parameter coupling effect, in the area of ​​high relaxation factor decay rate + low learning rate (upper right), the loss surface rises steeply, indicating that when the relaxation factor decays rapidly, appropriate learning rate compensation is required; in the area of ​​low relaxation factor decay rate + high learning rate (lower left), periodic fluctuations appear, reflecting that the parameter combination causes gradient oscillation; for the optimal channel, a low loss corridor is formed along the relaxation factor decay rate = 0.45~0.55, indicating that there is an optimal matching relationship between the parameters.

[0186] To verify the robustness of the autoencoder algorithm, for the relaxation factor decay rate robust band, in the relaxation factor decay rate range of 0.4~0.6, the loss value remains below 0.6 (cold color area); for the learning rate tolerance, when the learning rate>0.005, the loss surface shows a bifurcation mutation, indicating that the learning rate needs to be precisely controlled.

[0187] Experiments show that the theoretical optimal point is (relaxation factor decay rate = 0.5, learning rate = 0.005). The gradient along the relaxation factor decay rate axis is greater than the learning rate axis, proving that the relaxation factor strategy is more sensitive to performance.

[0188] Furthermore, a parameter adjustment strategy is given according to the surface characteristics, as shown in Table 1: Table 1 Optimization suggestions for parameter adjustment strategies based on surface characteristics

[0189] Furthermore, to verify the necessity of the dynamic relaxation factor, experimental data show that the optimal relaxation factor decay rate = 0.5 corresponds to a relaxation factor of Attenuation, satisfying the requirement of decaying to 0.67% of the initial value at t=10, ensuring that the regularization strength is reduced in the later stages of training.

[0190] Furthermore, a learning rate sensitivity analysis was performed, and the calculation of the second-order derivative of the loss with respect to the learning rate showed that the curvature suddenly changed when the learning rate was > 0.008, corresponding to the surface bifurcation phenomenon.

[0191] Compared with the traditional method, the traditional fixed relaxation factor (relaxation factor decay rate = 0) is used as a control, and the results are shown in Table 2: Table 2 Comparison results of dynamic relaxation factor attenuation rate and fixed relaxation factor attenuation rate = 0

[0192] The experimental results show that there is a significant coupling effect between the relaxation factor decay rate and the learning rate, and they need to be optimized jointly. The dynamic relaxation factor mechanism makes the loss surface smoother and the optimal area wider. In the range of relaxation factor decay rate = 0.450.55 and learning rate = 0.0040.006, the algorithm shows strong robustness.

[0193] What is disclosed above is only one or more preferred embodiments of the present application, and cannot be used to limit the scope of rights of the present application. Ordinary technicians in this field can understand that all or part of the processes of implementing the above embodiments and equivalent changes made according to the claims of the present application are still within the scope covered by the present application.

[0194] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.

Claims

1. A method for automatic classification of medical images based on machine learning, characterized in that: The following steps are involved: Collect medical imaging data and increase the amount of medical imaging data, wherein the amount of medical imaging data is increased through a generative adversarial network algorithm based on quantum interference; Performing dimensionality reduction processing on the quantity of the extracted and enhanced medical image data; wherein an autoencoder algorithm based on a relaxation factor is used as a feature dimensionality reduction model; Classify the features after dimensionality reduction to obtain the categories of medical imaging data.

2. The method for automatic classification of medical images based on machine learning according to claim 1, characterized in that: The training steps of the quantum interference-based generative adversarial network algorithm include: Initialize the generator and discriminator of the generative adversarial network; The generator simulates a plurality of medical image states through a quantum state preparation stage, wherein the medical image states include images of health conditions or pathological characteristics; through a quantum interference simulation strategy, the medical image states are coherently superimposed to generate first medical image data with new characteristics; The generator outputs first medical image data; the first medical image data is slightly disturbed by the pixels of the image through the chaotic sequence generation module to increase the diversity of the first medical image data; The discriminator learns to identify the difference between the first medical image data and the real medical image data, adjusts parameters according to the difference and improves recognition accuracy; The iteration is repeated until the first medical image data generated by the generator can make it difficult for the discriminator to distinguish the authenticity.

3. The method for automatic classification of medical images based on machine learning according to claim 2, characterized in that: In the repeated iterative process, the back propagation algorithm is used to update the generator parameters and the discriminator parameters, and the updating method is expressed as: In the formula, is the generator parameter updated for this iteration, is the discriminator parameter updated in this iteration, is the learning rate of the generative adversarial network, and They represent the gradients of the loss function of the generative adversarial network with respect to the generator parameters and the discriminator parameters, respectively. is the learning rate of the discriminator, and are the weight parameters of the generator and the discriminator respectively.

4. The method for automatic classification of medical images based on machine learning according to claim 2, characterized in that: Adopt a dynamic adjustment strategy to the learning rate of the discriminator Make adjustments, set To generate the gradient of the loss function of the adversarial network with respect to the discriminator parameters, it is expressed as: Among them, the calculation method of the discriminator's learning rate is expressed as: In the formula, represents the rate of change of the discriminator's historical gradient, is the adjustment sensitivity parameter of the discriminator, which controls the response speed of the adjustment factor, and e is a natural number; During the training process, the quality and diversity of the generated first medical image data are continuously evaluated until the iteration is stopped when the evaluation function value is greater than a preset value.

5. The method for automatic classification of medical images based on machine learning according to claim 2, characterized in that: Increase the diversity of the first medical image data; specifically expressed as: In the formula, is the generator function; is the output image of the generator; represents the chaotic perturbation function applied to the image, is the coefficient that adjusts the intensity of chaotic disturbance; The final image generated by the generator in this iteration; It is the quantum state after quantum interference simulation.

6. The method for automatic classification of medical images based on machine learning according to claim 5, characterized in that: The calculation of the chaotic perturbation function applied to the image utilizes the combination of the logical mapping and the hyperbolic tangent function in the chaos theory.

7. The method for automatic classification of medical images based on machine learning according to claim 1, characterized in that: Based on the relaxation factor autoencoder algorithm model, the training steps include: Initializing an autoencoder network structure, the autoencoder includes an encoder and a decoder, wherein the encoder maps a high-dimensional medical image data feature vector to a low-dimensional representation space, and the decoder reconstructs high-dimensional input data from the low-dimensional space; Setting an initial value of a dynamic relaxation factor and its adjustment strategy, wherein the relaxation factor is adjusted during the training process according to error feedback; The input medical image data feature vector is converted into a representation of a low-dimensional space through an encoder, and then reconstructed into an output close to the high-dimensional input data through a decoder; Calculate the loss function of the autoencoder, including reconstruction error, dimensionality reduction error, and regularization term; Perform back propagation according to the loss function to update the weights and bias parameters in the autoencoder network; Repeat the above steps until the preset stop iteration condition is met.

8. The method for automatic classification of medical images based on machine learning according to claim 7, characterized in that: Based on the relaxation factor autoencoder algorithm model, the training steps also include: The calculation method of the regularization term of the loss function is expressed as: In the formula, and Represent the Frobenius norm and L1 norm, respectively, for the regularization of weights and biases, is the first regularization coefficient of the autoencoder, For encoder The weight matrix of the layer; For encoder The bias of the layer, is the second regularization coefficient of the autoencoder, is the total number of layers of the encoder, MRr represents the cumulative error through all layers of the encoder, and Rr is the regularization term of the autoencoder; The calculation method of the accumulated error through all layers of the encoder is expressed as: In the formula, Indicates The forward propagation function of the layer, For encoder The output of the hidden layer, For encoder The output of the hidden layer, For encoder The activation function of the layer, is the weight coefficient of the local reconstruction error, is a local mapping of the low-dimensional space, is the feature vector of medical imaging data input to the encoder, The reconstructed data output by the decoder; is the L2 norm; The adaptive low-dimensional representation is expressed as: In the formula, is a local mapping of the low-dimensional space, is the weight matrix of the encoder; is the encoder bias, is the sample set input into the autoencoder for the current batch, For the Neighborhood points, For encoder The activation function of the layer, For the Medical image data points for The weighted coefficients of neighbor points satisfy , the weighted coefficient calculation method is expressed as: In the formula, It is the scale parameter of the distance, which controls the image range of the weighted coefficient of the point pairs within the neighborhood.

9. The method for automatic classification of medical images based on machine learning according to claim 1, characterized in that: Classifying the features after dimensionality reduction to obtain medical image data categories, the steps further include: Initialize the parameters of the fractional neural network; The medical image data features that have undergone dimension reduction processing are input into the fractional-order neural network, processed by the fractional-order activation function, and transmitted to the output layer layer by layer; Calculate the loss function based on the actual output and the expected output; The gradient of the loss function for each parameter is calculated through the back-propagation algorithm, and the weights and biases of the fractional neural network are updated; The fractional order parameters are adjusted according to the loss function to optimize the learning efficiency and accuracy of the model. The adjustment method is expressed as: In the formula, is the fractional-order adjustment coefficient, which is used to control the update speed and amplitude of the fractional-order parameters. is the gradient of the loss function with respect to the fractional-order parameter, which is used to guide the adaptive adjustment of the fractional-order parameter. It is a fractional order neural network. The fractional order of the layer activation function; Repeat the above steps until the preset stop iteration condition is met.

10. The automatic classification system of medical images based on machine learning is characterized by: include: A medical image data acquisition unit, the medical image data acquisition unit is used to collect medical image data and increase the amount of medical image data; A medical image data mining unit, the medical image data mining unit is used to perform dimensionality reduction processing on the enhanced medical image data; A medical image data classification unit; the medical image data classification unit is used to classify the features after dimensionality reduction to obtain medical image data categories.

Images