Abstract:In this paper, we propose an improved Unet model for brain tumor image segmentation, which combines coordinate attention mechanism and ASPP module to improve the segmentation effect. After the data set is divided, we do the necessary preprocessing to the image and use the improved model to experiment. First, we trained and validated the traditional Unet model. By analyzing the loss curve of the training set and the validation set, we can see that the loss value continues to decline at the first epoch and becomes stable at the eighth epoch. This process shows that the model constantly optimizes its parameters to improve performance. At the same time, the change in the miou (mean Intersection over Union) index shows that the miou value exceeded 0.6 at the 15th epoch, remained above 0.6 thereafter, and reached above 0.7 at the 46th epoch. These results indicate that the basic Unet model is effective in brain tumor image segmentation. Next, we introduce an improved Unet algorithm based on coordinate attention mechanism and ASPP module for experiments. By observing the loss change curves of the training set and the verification set, it is found that the loss value reaches the lowest point at the sixth epoch and then remains relatively stable. At the same time, the miou indicator has stabilized above 0.7 since the 20th epoch and has reached a maximum of 0.76. These results show that the new mechanism introduced significantly improves the segmentation ability of the model. Finally, we apply the trained traditional Unet model and the improved Unet model based on the coordinate attention mechanism and ASPP module to the test set for brain tumor image segmentation prediction. Compared to the traditional Unet, the enhanced model offers superior segmentation and edge accuracy, providing a more reliable method for medical image analysis with the coordinate attention mechanism and ASPP module.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the accuracy of brain tumor image segmentation. Specifically, the author aims to improve the segmentation effect of brain tumor images by improving the Unet model and combining the Coordinate Attention Mechanism and the ASPP module (Atrous Spatial Pyramid Pooling Module). ### Problem Background 1. **Limitations of Manual Segmentation**: Manual segmentation of brain tumor images is not only time - consuming and labor - intensive, but also easily affected by the subjective factors of the operator. 2. **Deficiencies of Traditional Algorithms**: Due to the significant differences among different patients, traditional algorithms have poor performance in dealing with tumors with complex shapes and blurred boundaries. 3. **Clinical Requirements**: Accurate brain tumor segmentation is crucial for tumor diagnosis, treatment planning and prognosis evaluation. Therefore, the development of automated image segmentation methods has important clinical significance. ### Solutions In order to improve the effect of brain tumor image segmentation, the author proposes the following improvement measures: - **Introducing the Coordinate Attention Mechanism**: By introducing spatial position information, the model's ability to focus on specific regions is enhanced, thereby improving feature extraction and representation. - **Introducing the ASPP Module**: Use dilated convolution to increase the receptive field while maintaining the resolution of the feature map, thereby effectively capturing multi - scale information. ### Experimental Verification The author verified the effectiveness of the improved model through the following steps: 1. **Dataset Partitioning and Pre - processing**: Divide the dataset into a training set and a validation set, and perform necessary pre - processing on the images (such as grayscale conversion, histogram equalization, etc.). 2. **Preliminary Experiment**: Use the traditional Unet model for training and validation, and observe the change curve of the loss function and the change of the mIoU (Mean Intersection over Union) index. 3. **Improved Model Experiment**: After introducing the Coordinate Attention Mechanism and the ASPP module, conduct the experiment again and compare the performance differences between the models before and after improvement. ### Experimental Results - **Traditional Unet Model**: After multiple iterations, the mIoU value exceeds 0.6 at the 15th epoch and reaches more than 0.7 at the 46th epoch, indicating that the traditional Unet model has a certain effect in the brain tumor image segmentation task, but there is still room for improvement. - **Improved Unet Model**: The loss value of the improved model reaches the lowest point and tends to be stable at the 6th epoch, and the mIoU value stabilizes above 0.7 at the 20th epoch and reaches the maximum value of 0.76, which is significantly better than the traditional Unet model. ### Conclusion By introducing the Coordinate Attention Mechanism and the ASPP module, the improved Unet model shows stronger feature extraction and edge recognition abilities in the brain tumor image segmentation task, significantly improving the segmentation effect. This result not only provides new technical means for medical image analysis, but also provides valuable references for subsequent related research. ### Formula Display Some key formulas involved in the paper include: - **Dice Coefficient**: \[ Dice=\frac{2|X\cap Y|}{|X| + |Y|} \] where \(X\) and \(Y\) represent the prediction result and the ground - truth label respectively. - **Cross - Entropy Loss Function**: \[ L_{CE}=-\sum_{i = 1}^{N}[y_i\log(p_i)+(1 - y_i)\log(1 - p_i)] \] where \(y_i\) is the ground - truth label and \(p_i\) is the predicted probability. These formulas are used to evaluate the segmentation effect of the model and optimize the model parameters.

Improved Unet model for brain tumor image segmentation based on ASPP-coordinate attention mechanism

Improved Unet brain tumor image segmentation based on GSConv module and ECA attention mechanism

Segmentation of brain tumor MRI image based on improved attention module Unet network

Improved MRI Brain Tumor Segmentation Method Based on UNet++

A Medical Image Segmentation Method Based on Improved UNet 3+ Network

Research on the magnetic resonance imaging brain tumor segmentation algorithm based on DO‐UNet

Brain tumor magnetic resonance image segmentation by a multiscale contextual attention module combined with a deep residual UNet (MCA-ResUNet)

Residual UNet with spatial and channel attention for automatic magnetic resonance image segmentation of rectal cancer

MA-Unet: An improved version of Unet based on multi-scale and attention mechanism for medical image segmentation

A Novel Brain Tumor Segmentation Method Based on Im-proved Spatial Attention Mechanism and Multi-path Deep Neural Network

Brain tumor image segmentation algorithm based on multimodal feature fusion of Bayesian weight distribution

Research on improved Unet in microscopic image segmentation of esophageal cancer

A4-Unet: Deformable Multi-Scale Attention Network for Brain Tumor Segmentation

Brain tumor image segmentation method using hybrid attention module and improved mask RCNN

Multiscale lightweight 3D segmentation algorithm with attention mechanism: Brain tumor image segmentation

Increase Channel Attention Based on Unet++ Architecture for Medical Images.

Brain tumor segmentation with corner attention and high-dimensional perceptual loss

Improving Whole-Heart CT Image Segmentation by Attention Mechanism

Application of U-Net Network Utilizing Multiattention Gate for MRI Segmentation of Brain Tumors

Multimodal brain tumor image segmentation based on DenseNet

MRS‐Net: Brain tumour segmentation network based on feature fusion and attention mechanism