Multi-target segmentation of pancreas and pancreatic tumor based on fusion of attention mechanism. (January 2023)

• Record Type: Journal Article
• Authors: Cao, Luyang; Li, Jianwei; Chen, Shu
• Is Part Of: Biomedical signal processing and control. Volume 79(2023)Part 2
• Publication Date: 2023-01
• Subjects: Pancreas -- Pancreatic tumor -- Multi-object segmentation -- Attention mechanism -- Small target detection -- Cross entropy loss; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2022.104170
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: Multi-objective segmentation algorithms for pancreatic and pancreatic tumors are relatively scarce. The existing pancreas segmentation algorithms are weak in small target feature extraction. The existing pancreas segmentation algorithms require too much computation and high equipment requirements. Therefore, clinical application is more difficult. We design a novel loss function (WML), which effectively enhances the network's attention to target samples and the ability to capture edge detail pixels of pancreas and pancreatic tumor, so that the weights of the convolution kernel change in the direction favorable to pancreatic and pancreatic tumor feature learning. We propose a novel attention module (MDAG) for exploring multidimensional hybrid features of pancreas and pancreatic tumour. MDAG is able to efficiently learn contextual information within the U-shaped network, accomplish small target feature localisation in multiple dimensions of space and channels, and filter redundant information in shallow feature maps, thus enhancing the feature representation of pancreas and pancreatic tumour. Our proposed MDAG performs well on both the Task07_Pancreas dataset as well as the NIH_Pancreas dataset, and the MDAG is able to achieve higher segmentation performance with a smaller number of parameters. Abstract: Existing neural network segmentation schemes perform well in the task of segmenting images of organs with large areas and clear morphology, such as the liver and lungs. However, it is difficult to segment organs with variable morphology and small target area, such as pancreas and tumors. In order to achieve accurate segmentation of pancreas and its cysts, MDAG-Net (Multi-dimensional Attention Gate Network) is proposed in this paper. Combining three attention mechanisms: spatial, channel and multi-dimensional feature map input, MDAG(Multi-dimensional Attention Gate) obtains the global distribution of semantic information in spatial and channel dimensions, filters redundant information in shallow feature maps, realizes feature response, and recalibrates convolution kernel parameters. In addition, the WML(Weighted cross entropy and MIoU loss function) loss can adaptively assign the weight of category loss and count the classification error of global pixels, which can increase the error attention of the target area and improving the segmentation accuracy of the network. The algorithm is experimented on the Task07_Pancreas dataset, compared with U-Net under the same conditions, the Dice coefficient, Precision, Recall rate and MIoU (Mean Intersection over Union) of MDAG-Net are improved by 5.3%, 1.5%, 12.7% and 7.6% respectively. The results show that MDAG-Net can accurately segment the region of pancreas and its cyst in CT(Computed Tomography) images, which proves that MDAG has better segmentation efficiency for such small target regions.