A Generative Adversarial Network Fused with Dual-Attention Mechanism and Its Application in Multitarget Image Fine Segmentation

2.1. The Generative Adversarial Network Basic Model

The generative adversarial network (GAN) basic model in this paper includes two parts: generator and discriminator. The basic structure and information processing flow are shown in Figure 1.

In Figure 1, the generative network module in GAN is mainly used to generate target data. The typical generative network generally is usually the neural network based on deconvolution, such as FCN, U-Net, and SegNet, which recovers to the size of the input image through sampling on multiple deconvolution layers and finally obtains the generated image data. For image segmentation, the generative network is used as an image segmentation model, which receives the input of the original image and takes the predictive segmentation as the output. The adversarial network in GAN usually adopts convolutional neural networks. Its input includes real data and generated data from the generator. Through authenticity judgment and game learning with the generator, the ability of the generator to generate data is optimized.

The gray value of the input image to generative network is recorded as the random variable x, the distribution it obeys is set to P(x), and the noise data z obeys the uniform distribution. The generator maps the noise data z to G(z), and the authenticity discrimination probability of the image random variable x is D(x).

In learning, GAN is optimized according to the principle that the generative model maximizes log  D(x) and the discrimination model minimizes log(1 − D(G(z))). The objective function is defined as follows:

2.2. Nonlocal Attention Mechanism

The core idea of nonlocal attention mechanism is that the characteristic response value at a pixel is equal to the weighted average of the characteristic values at all points, that is, all points in the receptive field are connected, and the dependency relationship between pixels based on spatial distance is established to realize global information fusion. The calculation formula [38] is

where i represents the output position index, j represents any position of the input, x represents the input, y represents the output with the same size as x, f(·) is the similarity measurement function between pixels, g(·) represents the feature mapping of j, and C(x) is the normalization factor. (f) and g can be implemented in many ways. If g is a Gaussian function, its expression is

The operation in equation (3) represents the difference amplified exponentially after multiplying the two vector matrices, and equation (4) is the mathematical expression of the normalization function C(x).

Consider an extended form of equation (3). Firstly, the vector x is embedded into spatial mapping, that is, the two vectors are mapped to different feature spaces, and then, the Gaussian function is used to measure the similarity. The specific form is

The vector dot product of function f can be expressed as

Now, the normalization function C(x) is equal to N, and N is the number of positions of x.

Based on the paired function form in the relational network proposed by Santoro et al. [39], the function f can be expressed as a cascaded form:

where [,] represents cascade and w_f^T represents the weight vector that projects the cascade vector to the scalar.

The nonlocal attention mechanism is integrated into convolutional neural network to form a general nonlocal block. The module can be integrated into any neural network structure. The definition of nonlocal block is as follows:

where z is the output of the attention module, y is the output of equation (2), w_z is the weight matrix of the number of restored input channels, and x is the input.

The structure of nonlocal block is shown in Figure 2.

In Figure 2, the nonlocal block first performs feature mapping on the input matrix I(C × H × W) with a 1 × 1 convolution, that is, the mapping operation of the functions g, θ, and φ in (3) and (5), to get matrices w_r, w_k, and w_v. Then, the matrix w_k is deformed and transposed into a matrix F_k(HW × T), and w_r is transformed into a matrix F_r(T × HW). F_k is multiplied by F_r and then normalized by Softmax to obtain the similarity measure matrix F(HW × HW). Then, multiply the deformed matrix F_v(T × HW) from w_v and the matrix F to obtain the characteristic response matrix F_s(T × HW). Finally, F_s is deformed and multiplied by the convolution kernel w_z to restore the original number of channels. The operation result is added to the input matrix I(C × H × W) to obtain the output O(C × H × W), and the final output feature map is a feature map with enhanced global information dependence.

2.3. Residual Generation Network Based on Dual-Attention Mechanism

In the complex images' multitarget segmentation, if the number of image instance targets is large and the difference of individual gray value is small, the neural network needs to have strong ability of feature extraction, fusion, and recognition. In this paper, the residual network is used as the main body of the image segmentation model, and the shortcut-connection structure is added between different convolution layers, that is, the input of the upper network is directly superimposed with the output of the lower network at the element level, so as to reduce the loss of feature information and realize feature fusion of different levels. In this way, the network can still maintain good convergence properties in the deep case. The residual structure is shown in Figure 3.

At present, there are many classical residual network models, such as ResNet18, ResNet34, and ResNet50 [16]. In practice, the selection of the model is mainly based on the requirements of input image size, quality, and segmentation accuracy. Generally, the deeper the network, the stronger the feature extraction ability. In this paper, ResNet50 is selected as the basic model according to the problem of small-target region detection in complex image multitarget segmentation. The structure of segmentation model based on ResNet50 is shown in Figure 4.

In Figure 4, the segmentation model is divided into six parts. (1) 7 × 7 convolution with step size of 2 and the number of output image channels is 64. (2) The pooling operation with step size of 2 and the 3 × 3 sliding window cascades three residual blocks. Each residual block contains three convolution layers. The first and third are 1 × 1 convolution to adjust the number of channels of the feature maps. The second is 3 × 3 convolution with a step size of 1. Equations (3) to (5) are stacked residual blocks. The last one includes average pooling with a step size of 1 and 7 × 7 sliding window, as well as fully connection and Softmax classification. Extract the output from the second to the fifth part of the model, and then, upsample the four output feature maps for prediction segmentation.

In practice, only relying on the residual network to perform image multitarget segmentation will result in the blurring of the segmentation boundary and the loss of small-target information, that is, the residual network cannot make full use of the image feature information to segment the image. In this paper, based on the residual network, a nonlocal attention mechanism is introduced to construct a dual-attention mechanism model that can integrate spatial and channel attention. The structure is shown in Figure 5.

In Figure 5, the spatial-attention module is used on a nonlocal mechanism to strengthen the dependency between all pixels on the feature maps, which is represented by a similar weight matrix. In this module, the output feature map of the residual network first undergoes 1 × 1 convolution to obtain three feature maps F1, F2, and F3, which have the same width and height as the input image, but the number of channels is reduced to 1/4 of the original. On this basis, F1 is transposed and multiplied by F2 and then processed by Softmax normalization to obtain the interpixel similarity matrix F(HW × HW). F is multiplied by F3, and the original input is added to obtain the feature map after spatial feature optimization.

The channel-attention mechanism is used to capture the interdependence between any two channels and update the value in one channel by using the weighted average of all channels. Its implementation steps are similar to the spatial-attention module. The feature maps obtained by the spatial-attention module and the channel-attention module are added and fused to generate a segmented image strengthened by the attention mechanism.

In Figure 5, the mathematical expression of Softmax is

where z_i^l represents the i^th pixel value in the l^th column of the feature map F. This formula normalizes the similarity measure matrix F by column so that the weight value of each column is within the interval [0,1].

The nonlocal attention mechanism is combined with the residual network model to construct the generator module in the GAN. The structure is shown in Figure 6.

In Figure 6, the generator module receives image data input, and the input image is extracted features by the ResNet50 network based on the dual-attention mechanism; four output feature maps at different levels are obtained, denoted as F0, F1, F2, and F3. The F1, F2, and F3 feature maps are upsampled to the same size as F0, and then, the four output feature map channels are spliced and 3 × 3 convolution is performed to obtain the fused feature map F. The fusion feature map F is spliced with F0, F1, F2, and F3 at channel level, respectively, and then, input into the self-attention mechanism module after 3 × 3 convolution to obtain four prediction segmentation maps. They are added and fused and averaged, and finally, the image multitarget prediction segmentation map is obtained.

2.4. The Adversarial Network Model Based on Convolutional Network

In GAN, the adversarial network module penalizes the lost details and small-size target information of the network through game learning, which makes the multitarget images segmented by the generative network more accurate. In this paper, the adversarial network module in GAN is constructed based on CNN, and its structure is shown in Figure 7.

In Figure 7, the adversarial network model includes two inputs: one is dot-product image of segmentation label and the original image, and the other is dot-product image of generative network and the original image. Dot-product operation is a mask operation on the original image to obtain the area of label and prediction segmentation on the original image. The adversarial network performs feature extraction on the obtained detection area, distinguishes whether the input is a real or predicted segmentation area, and optimizes the segmentation result through the adversarial learning with the generative network. In this paper, the CNN in the adversarial network model contains a total of 5 convolutional layers, 5 pooling layers, and two fully connected layers. The size of the convolution kernel is 3 × 3, and the step size is 1 × 1, that is, the convolution operation does not change the size of the input. The pooling layer is the maximum pooling with a size of 2 × 2, and the step size is 2 × 2, which reduces the input resolution by half. The ReLu function is selected as the activation function, and the Sigmoid function is used for classification operations.

2.5. The Generative Adversarial Network-Fused Attention Mechanism

In this paper, a generative adversarial network segmentation model fused with attention mechanism (AM-GAN) is proposed and the overall structure is shown in Figure 8.

In Figure 8, the gold standard is an accurate result of manual segmentation by experts. In AM-GAN, the generative network is composed of the ResNet50 model and the nonlocal dual-attention mechanism module. It takes the original image as input and the predicted segmentation as output. The adversarial network is composed of CNNs. Its inputs include the mask of the predicted segmentation and the original image and the mask of the gold standard and the original image. The mask of the predicted segmentation is judged as false (marked as 0), and the mask of the gold standard is judged to be true (marked as 1).

Combining the generative network based on the residual network and the nonlocal dual-attention mechanism with the adversarial network based on the CNNs to build a generative adversarial network model for the multitarget image segmentation. After AM-GAN is trained to reach the optimum, the model used for image segmentation is a generator network. The structure of the AM-GAN segmentation model is shown in Figure 9.

As shown in Figure 9, in the generator network, the conv2x, conv3x, conv4x, and conv5x parts of the ResNet50 based on the dual-attention mechanism all have a prediction segmentation output, and the final prediction segmentation of the generator network is the average value of the prediction segmentation of these four parts. The input of the generator is subjected to feature extraction through the extended ResNet50 network, and the rough feature map is obtained by upsampling. The calculation formula of the upsampling output F₁ of the conv3x part in the extended ResNet50 network is

where I represents the input image, conv3x(I) represents the convolution output of the conv3x part, w₁ represents the 1 × 1 convolution, whose purpose is to reduce the number of channels of the convolution output, b₁ represents the bias, BN(·) represents the batch normalization, Relu(·) is the activation function, and Umsample(·) is the interpolation upsampling. In this paper, the bilinear interpolation algorithm is used to upsample the feature map, and the output is F₁. Using the same algorithm, the upsampled output F₂ and F₃ of the conv4x and conv5x parts can be obtained. The output F₀ of the conv2x part does not need to be upsampled, and the sizes of F₁, F₂, and F₃ are the same as F₀. After feature extraction and upsampling of the input image, the output information of each part is fused. The calculation formula of the fusion feature map F is

where cat(·) represents the splicing operation of feature map channel, w₂ is a 3 × 3 convolution, which is used to fuse the information between the four feature maps, w₃ is a 1 × 1 convolution, which is used to reduce the number of channels of the fusion feature map, b₂ and b₃ are biases, and Relu(·) is the activation function. After the information of each part is fused, the feature map is input into the dual-attention mechanism module for feature enhancement. The calculation formula of the predicted output P₁ of the conv3x module is

where pAttention(·) represents the spatial-attention mechanism module, cAttention(·) represents the channel-attention mechanism module, w₄ represents 1 × 1 convolution, which is used to convert the number of channels into the number of classification categories, Upsample(·) represents a bilinear interpolation up-sampling operation, which is used to restore the size of output feature map to the size of the input image to obtain a predicted segmented image. Using the same method, the predicted output P₀, P₁, and P₃ can be obtained. Finally, P₀, P₁, P₂, and P₃ are added and averaged to obtain the predicted segmented image. The calculation formula is

Equation (13) is the calculation expression for predicting segmentation P, where avg(·) represents the average operation.

In the GAN segmentation model established in this paper, the data processing capability of the ResNet50 extended based on the nonlocal dual-attention mechanism can mechanically ensure that the image multitarget features can be accurately extracted, while nonlocal attention mechanism also strengthens the output feature map of the ResNet50, which can further improve the accuracy of segmentation. The authenticity judgment of the adversarial network can guide the segmentation network to avoid the loss of detailed information as much as possible and optimize boundary segmentation and small-target segmentation.

3.1. The Loss Function

In the complex images' multitarget segmentation, it is necessary to calculate classification errors of multiclass pixel. In this paper, the multiclassification cross-entropy loss [40] is used as the loss function of the generative network, which is defined as follows:

Equation (14) represents the multiclassification cross-entropy loss of the label image x and the predicted segmentation y, where (H, W, C) represents the length, width, and number of channels of the image.

In the AM-GAN segmentation model, the adversarial network uses the authenticity loss to perform game learning with the generative network, which is essentially a binary classification problem. Therefore, the two-classification cross-entropy loss [41] is used as the loss function of the adversarial network, which is defined as follows:

Equation (15) represents the two-classification cross-entropy loss, where x represents the classification label and y represents the classification probability output of the adversarial network.

The loss function of the generative network is defined as follows:

where x_n is the input image, y_n is the label segmentation, and θ_g is the training parameter set of the generative network. According to equation (16), the loss of generative network is composed of two parts, that is, the multiclassification cross-entropy loss of prediction segmentation and label segmentation and the loss of adversarial network which predict segmentation. The hyperparameter λ is used to balance these two losses. When optimizing the generative network, we minimize the objective function.

The loss function of the adversarial network is defined as

where θ  _d represents the training parameter set of the adversarial network. From (17), the loss function of the adversarial network consists of two parts. The first is the two-classification cross-entropy loss of the mask map of input image x_n and the label segmentation y_n, which is judged to be true (that is, the value is 1). The second is the two-classification cross-entropy loss of the mask map of input image x_n and the label segmentation g(x_n), which is judged to be false (that is, the value is 0). When optimizing the adversarial network, we minimize the loss function.

The loss function of GAN is composed of the loss of the generative network and the loss of the adversarial network. The calculation expression is

3.2. The Training Process

AM-GAN adopts the method of alternate training of generative network and adversarial network. In order to ensure the stability of training, the training times of the discriminating module are generally more than that of the generative module. In this paper, the minibatch gradient descent (MBGD) algorithm is used for AM-GAN training. The specific process is as follows:

• (1)

  Randomly sample n samples z_n from noise samples, and n samples x_n from real samples.

• (2)
  Gradient ascent algorithm is used to update the adversarial network:

  $$\begin{array}{c}{\nabla }_{{\theta }_d}\frac{1}{n}{\displaystyle \sum _{i=1}^n\left(\log \left(D\left(x^i\right)\right)+\log \left(1-D\left(G\left(z^i\right)\right)\right)\right)}.\end{array}$$

  (19)

• (3)

  Repeat steps (1) and (2) k times to update the adversarial network k times.

• (4)
  Sampling n generated samples from the noise samples, and we update the generative network once using gradient descent algorithm:

  $$\begin{array}{c}{\nabla }_{{\theta }_s}\frac{1}{n}{\displaystyle \sum _{i=1}^n\log }\left(1-D\left(G\left(z^i\right.\right)\right).\end{array}$$

  (20)

• (5)

  Sequentially, we repeat the above steps until the model training is stable and optimal.

The specific algorithm implementation is as follows:

• Step 1: determine and initialize the GAN training parameter set. The training parameter set of the generated network is θ_g=(w₀, w₁,…, w_m, b₀, b₁,…, b_m, g, θ, φ, w_z, w_c), where (w₀, w₁,…, w_m) is the set of convolution kernel weight of extended ResNet50 network, (b₀, b₁,…, b_m) is the set of biases, g, θ, φ, and w_z are 1 × 1 convolution kernel weights of spatial-attention mechanism, and w_c is 1 × 1 convolution kernel weights of channel-attention mechanism. The training parameter set of the adversarial network is θ  _d=(w₀, w₁,…, w_n, b₀, b₁,…, b_n), where (w₀, w₁,…, w_n) is the set of convolution kernel weights of CNN, and (b₀, b₁,…, b_n) is the set of biases.

• Step 2: randomly select N image samples x_n from the sample set, and select corresponding N label segmentation samples y_n.

• Step 3: calculate the loss of the adversarial network: ∇L(θ  _d)=1/N∑_n=1^N[L_bec(d(x_n, y_n, θ  _d), 1)+L_bec(d((x_n, g(x_n), θ  _d)), 0)].

• Step 4: calculate the training parameter gradient according to the chain rule, a_i=σ(z_i)=σ(w_ia_i−1+b_i). σ is the activation function, a_i is the input feature map of the i^th layer, z_i is the feature map of the i^th layer after convolution, w_i is the weight of the i^th layer, and b_i is the bias of the i^th layer.

• The formula for calculating the weight gradient using the chain rule is

  $$\begin{array}{c}\frac{\partial L\left({\theta }_d\right)}{\partial w_i}=\frac{\partial L\left({\theta }_d\right)}{\partial z_i}·a_{i-1}={\sigma }^{\prime }\left(z_i\right)·{\left(w_{i+1}\right)}^T·\left(\frac{\partial L\left({\theta }_d\right)}{\partial z_{i+1}}\right)·a_{i-1},\end{array}$$

  (21)

• where σ′(·) is the derivative of the activation function and (·)^T represents the matrix transpose.

• The formula for calculating the bias gradient using the chain rule is

  $$\begin{array}{c}\frac{\partial L\left({\theta }_d\right)}{\partial b_i}=\frac{\partial L\left({\theta }_d\right)}{\partial z_i}=\sigma \left(z_i\right)·{\left(w_{i+1}\right)}^T·\left(\frac{\partial L\left({\theta }_d\right)}{\partial z_{i+1}}\right).\end{array}$$

  (22)

• Step 5: use the MBGD optimizer to update the convolution weights and biases of the adversarial network:

  $$\begin{array}{c}{w}_i^t=w_i^{t-1}-\eta \frac{1}{N}{\displaystyle \sum _{n=1}^N\frac{\partial L\left({\theta }_d^{t-1}\right)}{\partial w_i^{t-1}}},\end{array}$$

  (23)

  $$\begin{array}{c}{b}_i^t=b_i^{t-1}-\eta \frac{1}{N}{\displaystyle \sum _{n=1}^N\frac{\partial L\left({\theta }_d^{t-1}\right)}{\partial b_i^{t-1}}}.\end{array}$$

  (24)

• Step 6: repeat Step 2 Step 5 k times.

• Step 7: randomly select N image samples x_n, and select corresponding N label segmentation samples y_n.

• Step 8: calculate the loss ∇L(θ_g) of the generative network:

  $$\begin{array}{c}\nabla L\left({\theta }_g\right)=\frac{1}{N}{\displaystyle \sum _{n=1}^N\left[L_{\text{mec}}\left(g\left(x_n,{\theta }_g\right),y_n\right)\right.}\\ \left.+\lambda L_{\text{bec}}\left(d\left(g\left(x_n,{\theta }_g\right),x_n\right),1\right)\right].\end{array}$$

  (25)

• Step 9: use the MBGD optimizer to update the parameters θ_g of the generation network.

• Step 10: if the network converges to the optimal, then the training ends; otherwise, it returns to Step 2.

The epoch of training is set to 200 and the batch size is 10. According to the above algorithm process, the pseudocode of the training algorithm is shown in Algorithm 1.

4.1. The Experiment Dataset

The dataset comes from the Combined Healthy Abdominal Organ Segmentation (CHAOS) competition dataset [42]. The CHAOS dataset contains the abdomen MRI images. In the experiment, the abdominal MRI abdominal images containing the label information of the liver, left/right kidney, and spleen were selected as the sample set. The typical abdomen image and label segmentation image are shown in Figure 10.

The experimental dataset contains 38 groups of abdominal MRI images, and each group has 26 slices with a size of 256 × 256, totaling 988 slices. The dataset is divided into a training set, a validation set, and a test set. The training set includes 30 groups of slices, the validation set includes 2 groups of slices, and the test set includes 6 groups of slices. Due to the small number of samples in the dataset, random rotation, mirroring, and other operations are used to enhance the data in the experiment. The final number of slices in the training set is expanded to 3120, the verification set is expanded to 104, and the test set is expanded to 312. The sample distribution of dataset is shown in Table 1.

4.2. The Model Structure and Parameter Settings

The training strategy of the AM-GAN is to alternate training between the generative network and the adversarial network. In order to ensure the stability of training, the ratio of the training times of the generative network and the adversarial network is set to 1 : 6, that is, the adversarial network training is performed 6 times first, and then, the generative network is trained once.

The size of the convolution kernel of the extended ResNet50 in the generative network is all 3 × 3, the step size of the downsampling is 2 × 2, and the other step size is 1 × 1. The activation function adopts the ReLu function, and the balance coefficient λ of the loss function of generative network is set to 0.2. The adversarial network includes 5 convolutional layers, 5 pooling layers, and 2 fully connected layers. The convolution kernel size of the convolutional layer is set to 3 × 3, and the step size is 1 × 1. The pooling window size of the pooling layer is set to 2 × 2, and the step size is 2 × 2. The Sigmoid function is used as the classification function. The learning rate of the MBGD optimizer is set to 0.01, and the weight parameters are initialized with truncated normal distribution. The number of training iterations is 200, and the batch of one training is 10. The experimental environment is Linux system with NVIDIA GeForce RTX 2080Ti GPU.

4.3. The Experiment Result and Analysis

The AM-GAN training is carried out according to the algorithm strategy in Section 3 and the model structure and parameter setting in Section 4.2. Multitarget image segmentation is performed by generative network. Recall rate, precision, F1-score, accuracy, and Dice similarity coefficient are used to evaluate the properties of the segmentation model. The image segmentation experiment results of the test set are shown in Figure 11, and the index evaluation results are shown in Table 2.

In Figure 11, from left to right, there is the original abdominal image, the gold standard, that is, the image manually segmented by the expert, and the predicted segmented image by network. From Figure 11, the segmentation results of the GAN are clear and smooth. Whether it is a large-sized organ such as the liver or a small-sized organ such as the kidney and pancreas, the GAN can distinguish them more accurately, which shows that the proposed method has good applicability.

For analyzing the properties of the learning algorithm, the curves of the Dice coefficient indicators of four organs and tissues, including liver, left kidney, right kidney, and spleen changing with iteration, and the confusion matrix of the validation set are recorded. The results are shown in Figure 12 and Table 3.

In Figure 12, the Dice similarity coefficient training curves of the liver, left kidney, right kidney, and spleen are shown. The Dice curve of the liver converges quickly and is relatively stable in the later iterations. This is because its size is relatively large compared to other organs, which causes the segmentation network to be more inclined to learn the classification of the liver area pixels in the early stage. The left kidney, right kidney, and spleen organs are smaller in size than the liver, and the segmentation network has a deeper level, which leads to easy loss of information and large fluctuations in the training. In the later stage of training, the average Dice coefficient of liver, left kidney, right kidney, and spleen on the training set is 0.92, which shows that the overall effect of model segmentation is good, and it has good applicability to small-target segmentation.

From Table 3, the segmentation accuracy of liver organs is the highest, 97.55%. This is because the size of liver organs is relatively large, and the network tends to learn their pixel weights of liver organs. The spatial location of the left kidney and the liver area is relatively close, and the boundary tissues overlap, leading to the segmentation error of the left kidney mainly from the liver and background. The right kidney and spleen organs are close in space, leading to mutual influence. The right kidney and spleen are small-target organs, and there is no boundary overlap, so the segmentation is less affected than the left kidney. From the confusion matrix, the accuracy of left kidney segmentation was 82.39%, while that of right kidney segmentation was 92.68% and that of spleen segmentation was 94.12%. It shows that the method in this paper has a good ability to identify and distinguish the features of complex image pixel categories.

4.4. Comparative Experiment and Analysis

In order to verify the performance improvement of AM-GAN brought by the nonlocal attention mechanism, a ResNet-GAN model was constructed in the comparative experiment. This model removes the nonlocal attention mechanism in AM-GAN, and other structures are the same as the AM-GAN. In addition, current mainstream, AM-FCN [43] embedded based on attention, Attention U-Net [44] fused on attention in both encoder and decoder, DANet [45] embedded based on dual attention, and SEVNet [46] fused on SE module are selected as comparison models. The four index values of accuracy, recall, precision, and F1-score are used for comparative evaluation. In order to avoid misleading the performance evaluation by high background indicators, all evaluation indicators do not include the value of background indicators. The specific results are shown in Table 4.

From Table 4, the four indexes of AM-FCN and Attention U-Net are lower than the proposed model. This is because it organically integrates GAN's inherent ability to process and repair low-resolution images, and the global to local content association and feature enhancement mechanism of nonlocal attention improves the accuracy of multitarget segmentation results. Except for the recall rate, other indexes of DANet are lower than that of the proposed model, which shows that the nonlocal attention mechanism used in this model can gather the receptive field from global to local target area to realize information context correlation. In addition, GAN can optimize the segmentation results in mechanism and improve the segmentation properties of the model, and the model and algorithm are relatively robust. Compared with ResNet-GAN, the accuracy of proposed model is 1.1% higher, which shows that the nonlocal attention can effectively realize the information association from global to local. Combined with GAN's probability exploration mechanism, the segmentation accuracy of the model is comprehensively improved. Compared with the SEVNet, due to the inherent high-definition processing and repairing capabilities of the GAN for noise images and the use of a game strategy, the optimization of the segmentation results is realized, and the accuracy, continuity, and smoothness of the segmentation results are comprehensively improved. Based on the above analysis, it shows that the method in this paper has comprehensive advantages when performing image target segmentation with insignificant structural features and achieves a better segmentation effect.