A recurrent positional encoding circular attention mechanism network for biomedical image segmentation

作者全名："Yu, Xiaoxia; Qin, Yong; Zhang, Fanghong; Zhang, Zhigang"

作者地址："[Yu, Xiaoxia] Chongqing Univ Technol, Coll Mech Engn, Chongqing 400054, Peoples R China; [Qin, Yong] Chongqing Med Univ, Childrens Hosp, Chongqing 400016, Peoples R China; [Zhang, Fanghong] Univ Ghent, Ghent, Belgium; [Zhang, Fanghong] Chongqing Normal Univ, Chongqing 400054, Peoples R China; [Zhang, Zhigang] Chongqing Univ Technol, Coll Mech Engn, Chongqing 400045, Peoples R China"

通信作者："Yu, XX (通讯作者)，Chongqing Univ Technol, Coll Mech Engn, Chongqing 400054, Peoples R China."

来源：COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE

ESI学科分类：COMPUTER SCIENCE

WOS号：WOS:001183638700001

JCR分区：Q1

影响因子：6.1

年份：2024

卷号：246

期号：　

开始页：　

结束页：　

文献类型：Article

关键词：Biomedical image; Relative positional encodings; RPECAMNet; Segmentation

摘要："Deep-learning-based medical image segmentation techniques can assist doctors in disease diagnosis and rapid treatment. However, existing medical image segmentation models do not fully consider the dependence between feature segments in the feature extraction process, and the correlated features can be further extracted. Therefore, a recurrent positional encoding circular attention mechanism network (RPECAMNet) is proposed based on relative positional encoding for medical image segmentation. Multiple residual modules are used to extract the primary features of the medical images, which are thereafter converted into one-dimensional data for relative positional encoding. The recursive former is used to further extract features from medical images, and decoding is performed using deconvolution. An adaptive loss function is designed to train the model and achieve accurate medical-image segmentation. Finally, the proposed model is used to conduct comparative experiments on the synapse and self-constructed kidney datasets to verify the accuracy of the proposed model for medical image segmentation."

基金机构：Scientific Research Foundation of Chongqing University of Technology [0119230961]; Program for Innovation Team at Institution of Higher Education in Chongqing [CXQT21027]; Natural Science Foundation of Chongqing [cstc2020jcyj-msxmX0235]

基金资助正文："This research is supported by the Scientific Research Foundation of Chongqing University of Technology (No. 0119230961), the Program for Innovation Team at Institution of Higher Education in Chongqing (No.CXQT21027), the Natural Science Foundation of Chongqing (No. cstc2020jcyj-msxmX0235). This work was conducted at the Children's Hospital, Chongqing Medical University, Chongqing 400,016, China."