Medical image segmentation method based on cross pseudo supervision

Abstract

The present invention proposes a medical image segmentation method based on cross-pseudo-supervision, which is implemented by the following steps: obtaining training and test sample sets; constructing a medical image segmentation network model based on cross-pseudo-supervision with consistency between strong and weak perturbations and performing iterative training on it; and obtaining medical image segmentation results. The present invention uses a first network to strongly perturb the weakly perturbed samples and then perform image-level strong perturbation segmentation. Simultaneously, a second network performs feature-level strong perturbation segmentation on the weakly perturbated samples. The cross-pseudo-supervision module calculates the cross-pseudo-supervision losses of the image-level and feature-level strong perturbation segmentation subnetworks, respectively. This achieves consistency between strong and weak perturbations, avoids the drawback of the prior art that over-reliance on weak perturbations leads to accumulated training bias, and can better learn complex edge features in images. Compared with the prior art, the present invention effectively improves segmentation accuracy.

Description

Technical Field

[0001] The present invention belongs to the technical field of image processing, and relates to a medical image segmentation method, in particular to a medical image segmentation method based on cross pseudo-supervision, which can be applied to the field of medical imaging. Background Art

[0002] In recent years, computer-aided diagnosis (CAD) has made significant progress in medical imaging and diagnostic radiology, driven by advances in deep learning technology. Medical image segmentation, which can identify the pixels of organs or lesions from the background of medical images, is not only a crucial component of CAD but also one of the most challenging tasks in medical image analysis. Accurate medical image segmentation can effectively determine information such as the size and location of organs and lesions, enabling quantitative assessment of pre- and post-treatment outcomes, helping doctors make more accurate clinical decisions quickly and significantly reducing the workload of radiologists.

[0003] However, achieving accurate dense annotation requires the guidance of professional physicians, which requires highly specialized knowledge and experience. Therefore, obtaining large-scale, densely labeled segmentation maps is challenging. Semi-supervised medical image segmentation methods can reduce the cost of medical image annotation by combining a small amount of labeled data with a large amount of unlabeled data. However, most semi-supervised medical image segmentation methods focus on consistency regularization, which is often based on the clustering assumption of semi-supervised learning: samples of the same class will cluster in data clusters separated by low-density regions. In semantic segmentation tasks, the distribution of pixels differs significantly from the distribution of predicted classes in image classification tasks, and there are no obvious low-density regions to distinguish pixel classes. Choosing more meaningful image augmentation methods is key to achieving reliable segmentation performance in such tasks without obvious low-density regions. In the field of semi-supervised image segmentation, data augmentation methods that are applicable to specific image segmentation tasks and can significantly improve semi-supervised image segmentation performance are called strong perturbations. In contrast, classic data augmentation methods such as image flipping, scaling, and rotation are called weak perturbations. Designing strong perturbations that are suitable for specific tasks introduces additional supervisory information, thereby mitigating the confirmation bias that can occur when the model is only subjected to weak perturbations.

[0004] To improve the segmentation performance of medical image segmentation models by introducing cross-pseudo-supervision, a patent application with publication number CN115359029A, entitled "A Semi-Supervised Medical Image Segmentation Method Based on Heterogeneous Cross-Pseudo-Supervised Network," discloses a semi-supervised medical image segmentation method based on a heterogeneous cross-pseudo-supervised network. This invention combines the Unet and Swin-Unet in the HCPS network model for cross-pseudo-supervised learning, improving network training efficiency and segmentation performance. Furthermore, cross-pseudo-supervised learning based on the Unet and Swin-Unet combination enhances both global and local feature capture capabilities, fully utilizing available labeled data to achieve satisfactory segmentation results. However, this method suffers from the drawback of using only weak perturbation for data augmentation, relying heavily on the model's accurate prediction of weakly perturbed images. When there are fewer labeled medical images, the model's confirmation bias accumulates during training, resulting in poor segmentation of edge regions in medical images. Summary of the Invention

[0005] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and propose a cross-pseudo-supervised medical image segmentation method based on the consistency of strong and weak perturbations to solve the technical problem of low segmentation accuracy in the prior art.

[0006] To achieve the above object, the technical solution adopted by the present invention includes the following steps:

[0007] (1) Obtain training sample set and test sample set:

[0008] For Z including multiple pixel categories s The 3D medical images with labels are preprocessed and the K s N 2D medical images and their labels in the 2D medical images are used as labeled training sample sets, more than half of the remaining 2D medical images are used as unlabeled training sample sets, and the rest of the 2D medical images are used as test sample sets, where Z s ≥200,K s ≥2000,

[0009] (2) Constructing a medical image segmentation network model based on cross pseudo supervision:

[0010] Construct a medical image segmentation network model O including a weak perturbation module and a first network and a second network cascaded and arranged in parallel therewith, and a cross-pseudo-supervision module connected to the output ends of the two networks; wherein the first network includes a cascaded strong perturbation pool, an image-level strong perturbation segmentation subnetwork, and a perturbation recovery pool; the second network includes a cascaded feature-level strong perturbation segmentation subnetwork and a perturbation redo pool; the output ends of the image-level strong perturbation segmentation subnetwork and the feature-level strong perturbation segmentation subnetwork are also connected to the cross-pseudo-supervision module;

[0011] (3) Iterative training of the medical image segmentation network model:

[0012] The labeled training sample set and the unlabeled training sample set are divided into batches and used as the input of the medical image segmentation network model O for iterative training to obtain the trained medical image segmentation network model O*;

[0013] (4) Obtain the segmentation results of medical images:

[0014] The test sample set is used as the input of the image-level strong perturbation segmentation subnetwork P* in the trained medical image segmentation network model O* for forward inference to obtain the segmentation result map of each test sample.

[0015] Compared with the prior art, the present invention has the following advantages:

[0016] In the process of training the medical image segmentation network model, the present invention performs strong perturbation on the weakly perturbed samples and then performs image-level strong perturbation segmentation. At the same time, the second network performs feature-level strong perturbation segmentation on the weakly perturbated samples. One of the sub-networks can serve as a pseudo-label for the other sub-network for the segmentation of unlabeled data. The cross-pseudo-supervision module calculates the cross-pseudo-supervision losses of the image-level and feature-level strong perturbation segmentation sub-networks respectively, thereby achieving consistency between strong and weak perturbations, avoiding the problem of training bias accumulation caused by inaccurate weak perturbation predictions, and better learning the complex edge features in the image. Compared with the existing technology, the segmentation accuracy is effectively improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 Flowchart for the implementation of the present invention;

[0018] Figure 2 Schematic diagram of the overall structure of the medical image segmentation network model of the present invention;

[0019] Figure 3 This is a schematic diagram of the structure of the segmented network of the present invention;

[0020] Figure 4 Schematic diagram of the structure of the convolution module of the present invention. DETAILED DESCRIPTION

[0021] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0022] Reference Figure 1 , the present invention comprises the following steps:

[0023] Step 1) Obtain training sample set and test sample set:

[0024] Step 1a) Obtain Z from the ICHS intracerebral hemorrhage dataset s The dimension is D s ×H s ×W s 3D cerebral hemorrhage medical images and their corresponding segmentation labels; along each medical image and its corresponding dimension is D s ×H s ×W s The dimension D of the 3D segmentation labels s Extract the vertical s Slice the axis to get K s The dimension is H s ×W s The original image slice is sliced ​​and the skull is removed to keep only the central brain tissue area; s =Z s ×D s In this embodiment, Z s =270,16≤D s ≤24, H s =512,W s =512,K s =4982;

[0025] Step 1b) K s The dimension is H s ×W s The original image slices after skull removal and their corresponding labels are cropped with a center dimension of H×W, and the original image slices after center cropping are normalized to obtain the preprocessed K s 2D medical images of dimension H×W and their corresponding segmentation labels In this embodiment, H=256, W=256;

[0026] Step 1c) K s The N 2D medical images and their labels in the set constitute a labeled training sample set The remaining M preprocessed images are combined into an unlabeled training sample set The remaining U preprocessed target domain images are combined into a test sample set In this embodiment, N=1170, M=2338, and U=1474.

[0027] Step 2) Construct a medical image segmentation network model based on cross pseudo supervision:

[0028] Construct a medical image segmentation network model O including a weak perturbation module, a first network and a second network cascaded and arranged in parallel with the weak perturbation module, and a cross pseudo-supervision module connected to the output ends of the two networks. The structure is as follows Figure 2 As shown in FIG, the first network includes a cascaded strong perturbation pool, an image-level strong perturbation segmentation subnetwork, and a perturbation recovery pool; the second network includes a cascaded feature-level strong perturbation segmentation subnetwork and a perturbation redo pool; the output ends of the image-level strong perturbation segmentation subnetwork and the feature-level strong perturbation segmentation subnetwork are also connected to the cross pseudo-supervision module; the image-level strong perturbation segmentation subnetwork and the feature-level strong perturbation segmentation subnetwork both include a cascaded encoder Decoder And the output module Out, its structure is as follows Figure 3 As shown, the encoder It includes a sequentially connected convolution module Conv and a level I convolution Conv-encoding En module; a decoder The network includes sequentially connected I-level convolution Conv-decoding De modules; the input end of the i-th level convolution Conv-encoding En module is connected to the middle position of the I-i+1-th level convolution Conv-decoding De module. In this embodiment, I=4; the encoder in the feature-level strong perturbation segmentation sub-network With decoder There is also a feature perturbation module loaded between

[0029] The structure of the convolution module Conv is as follows Figure 4As shown in the figure: the first convolution layer → the first batch of regularization layers → the first nonlinear function layer → Dropout layer → the second convolution layer → the second batch of regularization layers → the second nonlinear function layer. The first nonlinear function layer and the second nonlinear function layer in all Conv modules use the Leakyrelu nonlinear activation function. The loss probability of the Dropout layer is set to 0.5. The Dropout layer randomly inactivates the neurons in the network model with a probability of 0.5, that is, the output of the neuron is set to 0. The first convolution layer and the second convolution layer have the same structure. The convolution kernel in the first convolution layer and the second convolution layer in the convolution module Conv is set to 3×3, the number is 16, the stride is 1, and the surrounding padding is 1. The first-level encoding submodule En1 includes the maximum pooling layer and the first convolution module Conv1 connected in sequence, where the first convolution layer and the second convolution layer in Conv1 are The convolution kernel in the convolution layer is set to 3×3, the number is 32, the step size is 1, and the surrounding padding is 1; the second-level encoding submodule En2 includes a maximum pooling layer and a second convolution module Conv2 connected in sequence, wherein the convolution kernels in the first convolution layer and the second convolution layer in Conv2 are set to 3×3, the number is 64, the step size is 1, and the surrounding padding is 1; the third-level encoding submodule En3 includes a maximum pooling layer and a third convolution module Conv3 connected in sequence, wherein the convolution kernels in the first convolution layer and the second convolution layer in Conv3 are set to 3×3, the number is 128, the step size is 1, and the surrounding padding is 1; the fourth-level encoding submodule En4 includes a maximum pooling layer and a fourth convolution module Conv4 connected in sequence, wherein the convolution kernels in the first convolution layer and the second convolution layer in Conv4 are set to 3×3, the number is 256, the step size is 1, and the surrounding padding is 1;

[0030] Decoder It includes 4 levels of decoding submodules De connected in sequence j, where the first-level decoding submodule De1 includes the first deconvolution layer and the fifth-level convolution module Conv5 connected in sequence, where the convolution kernel of the first deconvolution layer is set to 2×2, the number is 128, the step size is 2, and the surrounding padding is 0, the convolution kernels in the first convolution layer and the second convolution layer in Conv5 are set to 3×3, the number is 128, the step size is 1, and the surrounding padding is 1; the second-level decoding submodule De2 includes the second deconvolution layer and the sixth-level convolution module Conv6 connected in sequence, where the convolution kernel of the second deconvolution layer is set to 2×2, the number is 64, the step size is 2, and the surrounding padding is 0, the convolution kernels in the first convolution layer and the second convolution layer in Conv6 are set to 3×3, the number is 64, the step size is 1, and the surrounding padding is 1. is 1; the third-level decoding submodule De3 includes a third deconvolution layer and a seventh-level convolution module Conv7 connected in sequence, wherein the convolution kernel of the third deconvolution layer is set to 2×2, the number is 32, the step size is 2, and the surrounding padding is 0, and the convolution kernels in the first convolution layer and the second convolution layer in Conv7 are set to 3×3, the number is 32, the step size is 1, and the surrounding padding is 1; the fourth-level decoding submodule De4 includes a fourth deconvolution layer and an eighth-level convolution module Conv8 connected in sequence, wherein the convolution kernel of the fourth deconvolution layer is set to 2×2, the number is 16, the step size is 2, and the surrounding padding is 0, and the convolution kernels in the first convolution layer and the second convolution layer in Conv8 are set to 3×3, the number is 16, the step size is 1, and the surrounding padding is 1.

[0031] The output layer Out includes the output convolution layer and the softmax layer connected in sequence, where the convolution kernel in the output convolution layer is set to 3×3, the number is 2, the stride is 1, and the surrounding padding is 1.

[0032] Step 3) Iteratively train the medical image segmentation network model:

[0033] Step 3a) Initialize the number of iterations to r, the maximum number of iterations to R, R>10000, and the weight parameters of the current image-level strong perturbation segmentation subnetwork and feature-level strong perturbation segmentation subnetwork are θ respectively. Pr and θ Tr , and let r = 1. In this embodiment, R = 30000;

[0034] Step 3b) The weak perturbation module performs a perturbation on each training sample in each batch of input with probability p r The rotation operation f(θ) is performed under the condition that the rotation angle θ is randomly selected; with probability p x Flip the image up and down f(h); with probability p y Flip the image left and right f(ω); with probability p n In this embodiment, the image is not weakly disturbed. r =p x=p y =p n =0.5;

[0035] Step 3c) Strong perturbation pool Aug in the first network s (x) For a batch of even numbers B after weak perturbation L labeled training samples and an even number B U For the data of unlabeled training samples, a rectangular mask M(h,w) with a length and width of h and w is randomly generated. The mask parts of the b1th and b2th images in the labeled data and the unlabeled data are exchanged to obtain each labeled image x after strong perturbation. L S , the label y corresponding to the labeled image S and unlabeled images The way to generate the rectangular mask and the calculation process of the strong perturbation pool is:

[0036] M(h,w)= ((h,w)∈[λ1H:λ2H,λ3W:λ4W])

[0037]

[0038] in, is the indicator function; λ1<λ2,|λ1-λ2|>0.15, λ3<λ4,|λ3-λ4|>0.15. This hyperparameter is a random value, and its range can be set according to the specific dataset; In this embodiment, B L =4, B U =4;

[0039] Encoder in the image-level strong perturbation segmentation subnetwork right Encode and decode separately right The encoded features are decoded respectively, and the output module Out Decoding gets decoding features and predicts them separately to get labeled images Prediction probability plot of Unlabeled images Prediction probability plot of Disturbance Recovery Pool UnAug s (p S ) uses the rectangular mask M(h,w) generated by the strong perturbation pool The predicted probability map of the b1th and b2th images in The part corresponding to the mask part of is exchanged to obtain the cross pseudo label y' corresponding to the feature-level strong perturbation segmentation sub-network Un ; The calculation process of the disturbance recovery pool is:

[0040]

[0041] Step 3d) Encoder in the feature-level strong perturbation segmentation subnetwork in the second network Encode the labeled training samples and unlabeled training samples after weak perturbation respectively; Feature perturbation module Two-dimensional Dropout is used to strongly perturb the bottleneck features after passing through the encoder, and the bottleneck features after strong perturbation are connected to the original bottleneck features through channels, and pass through the encoder together to obtain the overall prediction result. Then the feature dimension is divided equally to obtain the prediction results without strong feature perturbation And the prediction results after strong feature perturbation Output module Out Decode the features to make predictions and get the predicted probability map of the labeled image And the predicted probability map of the unlabeled image Disturb the redo pool ReAug s (y' W ) uses the rectangular mask M(h,w) generated by the strong perturbation pool The parts corresponding to the mask parts of the b1th and b2th predicted probability maps are exchanged to obtain the cross pseudo label y' corresponding to the feature-level strong perturbation segmentation subnetwork Re ;

[0042] For the input weak perturbation image, the overall calculation process of the feature-level strong perturbation segmentation model is:

[0043]

[0044] Among them, e W represents the original bottleneck feature after the encoder, Represents the bottleneck feature after strong perturbation. In this chapter, the strong perturbation module Feat(·) uses two-dimensional Dropout in verification mode. concate(·,·) means concatenating two features, and chunk(·) means dividing the features into equal parts.

[0045] The specific calculation process of disturbing the redo pool is:

[0046]

[0047] Step 3e) The cross-pseudo-supervision module uses consistency regularization and entropy minimization methods to pass the strongly perturbed unlabeled image Prediction probability plot of and cross pseudo-label y' Re Calculate the cross pseudo-supervision loss L of the image-level strongly perturbed segmentation subnetwork unsup1 , while passing the predicted probability map of the unlabeled image and cross pseudo-label y' Un Calculate the cross pseudo-supervision loss L of the feature-level strongly perturbed segmentation subnetwork unsup2 , the calculation process is:

[0048]

[0049]

[0050] Among them, λ cons and λ ent They represent the weighted coefficients of the balanced mean square error loss and the cross entropy loss, onehot(·) represents the operation of converting features into one-hot encoding, and L cons , L ent Represents the mean square error loss function, cross entropy loss function, B u Represents the number of unlabeled images in each batch, p and p' respectively represent the feature map output by the encoder and the probability map obtained by the activation function for the data after different perturbations.

[0051] Step 3f) Segment all predicted probability maps p corresponding to the sub-network through image-level strong perturbation S , all predicted probability maps p corresponding to the feature-level strong perturbation segmentation sub-network W , the true labels y of all labeled training samples in each batch S ,by

[0052] and cross pseudo-label y' Un and y' Re ; Calculate the current loss value d of O r :

[0053] d r =L sup +λL unsup

[0054] Among them L sup , L unsup Represents supervised loss function and unsupervised loss function respectively, and λ represents L sup With L unsup Balance coefficient; supervised loss function, unsupervised loss function L sup , L unsup , whose expressions are:

[0055] L sup =λ sup1 L sup1 +λ sup2 L sup2

[0056]

[0057] L unsup =λ un (L unsup1 +L unsup2 )+λ fp L fp

[0058]

[0059] Among them, B L Indicates the number of labeled images in each batch, B u represents the number of unlabeled images in each batch, p and p' respectively represent the feature map output by the encoder and the probability map obtained by the activation function for the data after different perturbations;

[0060] Step 3g) Calculate the loss value d respectively r θ Pr and θ Tr The partial derivative g Pr and g Tr ; Using stochastic gradient descent method, through the partial derivative g Pr and g Tr θ Pr and θ Tr Update, the update formula is:

[0061]

[0062] in, Represent weight parameters and The update result of α r represents the learning rate. In this embodiment, α1=0.01, α r =α r-1 (1.0-r / R) 0.9 ; r>1, the semi-supervised medical image segmentation model O of this iteration is obtained r ;

[0063] Step 3h) Determine whether r=1 is true. If so, obtain the trained medical image segmentation network model O*. Otherwise, set r=r+1, O r =O, and execute step (3b).

[0064] Step 4) Obtain the segmentation results of the medical image:

[0065] The test sample set is used as the input of the image-level strong perturbation segmentation subnetwork P* in the trained medical image segmentation network model O* for forward inference to obtain the segmentation result map of each test sample.

Claims

1. A medical image segmentation method based on cross pseudo supervision, characterized in that: The steps include: (1) Obtain training sample set and test sample set: For Z including multiple pixel categories s The 3D medical images with labels are preprocessed and the K s N 2D medical images and their labels in the 2D medical images are used as labeled training sample sets, more than half of the remaining 2D medical images are used as unlabeled training sample sets, and the rest of the 2D medical images are used as test sample sets, where Z s ≥200,K s ≥2000, (2) Constructing a medical image segmentation network model based on cross pseudo supervision: Construct a medical image segmentation network model O including a weak perturbation module and a first network and a second network cascaded and arranged in parallel therewith, and a cross-pseudo-supervision module connected to the output ends of the two networks; wherein the first network includes a cascaded strong perturbation pool, an image-level strong perturbation segmentation subnetwork, and a perturbation recovery pool; the second network includes a cascaded feature-level strong perturbation segmentation subnetwork and a perturbation redo pool; the output ends of the image-level strong perturbation segmentation subnetwork and the feature-level strong perturbation segmentation subnetwork are also connected to the cross-pseudo-supervision module; (3) Iterative training of the medical image segmentation network model: The labeled training sample set and the unlabeled training sample set are divided into batches and used as the input of the medical image segmentation network model O for iterative training to obtain the trained medical image segmentation network model O*; (4) Obtain the segmentation results of medical images: The test sample set is used as the input of the image-level strong perturbation segmentation subnetwork P* in the trained medical image segmentation network model O* for forward inference to obtain the segmentation result map of each test sample.

2. The method according to claim 1, characterized in that The step (1) is to include multiple pixel categories Z s The preprocessing steps for the labeled 3D medical image are as follows: Each labeled 3D medical image is sliced ​​along the depth direction, and each 2D medical image slice after slicing is center-cropped, and then the center-cropped image slices are normalized to obtain the preprocessed K s 2D medical images.

3. The method according to claim 1, characterized in that The medical image segmentation network model O described in step (2), wherein: Both the image-level strong perturbation segmentation subnetwork and the feature-level strong perturbation segmentation subnetwork include cascaded encoders Decoder And output module Out, encoder in feature-level strong perturbation segmentation subnetwork With decoder There is also a feature perturbation module loaded between The encoder It includes a cascaded convolution module Conv and a level I convolution Conv-encoding En module; the convolution module Conv contains multiple stacked convolution layers, normalization layers and activation function layers; the Conv-encoding En module contains multiple stacked maximum pooling layers and multiple convolution layers; the decoder It includes a cascaded I-level convolution Conv-decoding De module; the output end of the i-th level convolution Conv-encoding En module is connected to the input end of the I-i+1-th level convolution Conv-decoding De module.

4. The method according to claim 1, wherein The iterative training of the medical image segmentation network model O described in step (3) is implemented as follows: (3a) The number of initial iterations is r, the maximum number of iterations is R, R>10000, and the weight parameters of the current image-level strong perturbation segmentation subnetwork and feature-level strong perturbation segmentation subnetwork are θ respectively. Pr and θ Tr , and let r = 1; (3b) The weak perturbation module rotates, flips upside down, or flips left to right each labeled training sample and unlabeled training sample in each batch according to a pre-set probability to obtain the weakly perturbed labeled training samples and unlabeled training samples; (3c) The strong perturbation pool in the first network performs strong perturbations on the labeled training samples and unlabeled training samples after weak perturbation; the image-level strong perturbation segmentation subnetwork performs strong perturbation on each labeled image obtained and each unlabeled image Perform feature learning and prediction to obtain Prediction probability plot of and Prediction probability plot of Perturbation recovery pool pair prediction probability map Perform perturbation recovery to obtain the cross pseudo label y' corresponding to the feature-level strong perturbation segmentation subnetwork Un ; (3d) The feature-level strong perturbation segmentation sub-network in the second network performs feature learning and prediction on the weakly perturbed labeled training samples and unlabeled training samples, and obtains the predicted probability map of the labeled image And the predicted probability map of the unlabeled image Disturb the redo pool Perform perturbation and redo to obtain the cross pseudo label y' corresponding to the image-level strong perturbation segmentation subnetwork Re ; (3e) The cross-pseudo-supervision module uses consistency regularization and entropy minimization methods to obtain the unlabeled images after strong perturbation. Prediction probability plot of and cross pseudo-label y' Re Calculate the cross pseudo-supervision loss L of the image-level strongly perturbed segmentation subnetwork unsup1 , while passing the predicted probability map of the unlabeled image and cross pseudo-label y' Un Calculate the cross pseudo-supervision loss L of the feature-level strongly perturbed segmentation subnetwork unsup2 ; (3f) All predicted probability maps p corresponding to the image-level strong perturbation segmentation sub-network S , all predicted probability maps p corresponding to the feature-level strong perturbation segmentation sub-network W , the true and cross-pseudo labels y' of all labeled training samples in each batch Un and y' Re ; Calculate the current loss value d of O r , and adopt the stochastic gradient descent method, through d r The weight parameters θ are Pr and θ Tr Update and obtain the semi-supervised medical image segmentation model O of this iteration r ; (3g) Determine whether r = 1. If so, obtain the trained medical image segmentation network model O*. Otherwise, let r = r + 1, O r =O, and execute step (3b).

5. The method according to claim 4, characterized in that The step (3c) Prediction probability plot of and Prediction probability plot of The steps to achieve the acquisition process are: (3c1) Strong perturbation pooling randomly generates a rectangular mask M(h,w) with a length and width of h and w respectively for a batch of data containing an even number of labeled training samples and an even number of unlabeled training samples after weak perturbation, and exchanges the mask parts corresponding to the b1th and b2th images in the labeled data and the unlabeled data respectively, to obtain each labeled image after strong perturbation. and each unlabeled image (3c2) Encoder in the image-level strong perturbation segmentation subnetwork right Encode and decode separately right The encoded features are decoded respectively, and the output module Out Decoding gets decoding features and predicts them separately to get labeled images Prediction probability plot of Unlabeled images Prediction probability plot of 6. The method according to claim 4, characterized in that The feature-level strong perturbation segmentation sub-network described in step (3d) performs feature learning and prediction on the labeled training samples and unlabeled training samples after weak perturbation respectively. The implementation steps are as follows: encoder Encode the labeled training samples and unlabeled training samples after weak perturbation respectively; Feature perturbation module Two-dimensional Dropout is used to strongly perturb some of the bottleneck features obtained by encoding; Decoder The strongly disturbed features and the unperturbed features are concatenated and decoded; the output module Out Decode the features to make predictions and get the predicted probability map of the labeled image And the predicted probability map of the unlabeled image 7. The method according to claim 4, characterized in that The cross-pseudo-supervision loss L described in step (3e) unsup1 , and the cross pseudo-supervision loss L unsup2 , the calculation formulas are: Among them, λ cons and λ ent They represent the weighted coefficients of the balanced mean square error loss and the cross entropy loss, onehot(·) represents the operation of converting features into one-hot encoding, and L cons , L ent Represents the mean square error loss function, cross entropy loss function, B u represents the number of unlabeled images in each batch, p i and p' i They respectively represent the feature maps output by the encoder and the probability maps obtained by activation function processing for data after different perturbations.

8. The method according to claim 4, characterized in that The current loss value d of O described in step (3f) r , the calculation formula is: d r =L sup +λL unsup Among them L sup , L unsup Represents supervised loss function and unsupervised loss function respectively, and λ represents L sup With L unsup The balance coefficient.

9. The method according to claim 4, characterized in that The step (3f) described in d r The weight parameters θ are Pr and θ Tr To update, the update formulas are: in, Represent the weight parameters θ respectively Pr and θ Tr The update result of α r represents the learning rate, g Pr 、g Tr Respectively represent d r θ Pr and θ Tr The partial derivative of .

Images