Method for stroke lesion segmentation using the primary-auxiliary path attention compensation networkChinese Full TextEnglish Full Text (MT)
HUI Haisheng;ZHANG Xueying;WU Zelin;LI Fenglian;College of Information and Computer,Taiyuan University of Technology;
Abstract: When the feature of stroke lesions is non-distinct,and the boundary between the lesions and the healthy brain tissue is difficult to distinguish,the segmentation model based on the self-attention mechanism is prone to generate a wrong attention coefficient map of the focus area,which affects the segmentation performance.To solve this problem,based on the global-attention-upsample attention U-Net(GAU-AUNet),we propose a primary-auxiliary path attention compensation network(PAPAC-Net).The primary path network is responsible for accurate lesion segmentation and outputting the segmentation results while the auxiliary path network generates a tolerant auxiliary attention compensation coefficient to compensate for the primary path network’s potential attention coefficient map errors.Two compound loss functions are also proposed to realize the different functions for the primary and auxiliary path networks.Experimental results show that our GAU-A-UNet and PAPAC-Net both have a significant improvement in segmentation performance,which proves the effectiveness of our method.
- DOI:
10.19665/j.issn1001-2400.2021.04.026
- Series:
(I) Electronic Technology & Information Science; (E) Medicine & Public Health
- Subject:
Neurology; Computer Software and Application of Computer
- Classification Code:
TP391.41;R743.3
- Mobile Reading
Read on your phone instantly
Step 1Scan QR Codes
"Mobile CNKI-CNKI Express" App
Step 2Open“CNKI Express”
and click the scan icon in the upper left corner of the homepage.
Step 3Scan QR Codes
Read this article on your phone.
- HTML
- BETABETAEnglish HTML (MT)
- CAJ Download
- PDF Download
Download the mobile appuse the app to scan this coderead the article.
Tips: Please download CAJViewer to view CAJ format full text.
Download: 318 Page: 200-208 Pagecount: 9 Size: 2365K
Citation Network
Related Literature
- Similar Article
- Reader Recommendation
- Associated Author