Hard Exudate Segmentation Supplemented by Super-Resolution with Multi-scale Attention Fusion Module

Jiayi Zhang1; Xiaoshan Chen1; Zhongxi Qiu2; Mingming Yang3; Yan Hu1; Jiang Liu2

10.1109/BIBM55620.2022.9995545

2022

2022 IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

978-1-6654-6820-6

2022 IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

1375-1380

6-8 Dec. 2022

Las Vegas, NV, USA

Hard exudates (HE) is the most specific biomarker for retina edema. Precise HE segmentation is vital for disease diagnosis and treatment, but automatic segmentation is challenged by its large variation of characteristics including size, shape and position, which makes it difficult to detect tiny lesions and lesion boundaries. Considering the complementary features between segmentation and super-resolution tasks, this paper proposes a novel hard exudates segmentation method named SSMAF with an auxiliary super-resolution task, which brings in helpful detailed features for tiny lesion and boundaries detection. Specifically, we propose a fusion module named Multi-scale Attention Fusion (MAF) module for our dual-stream framework to effectively integrate features of the two tasks. MAF first adopts split spatial convolutional (SSC) layer for multi-scale features extraction and then utilize attention mechanism for features fusion of the two tasks. Considering pixel dependency, we introduce region mutual information (RMI) loss to optimize MAF module for tiny lesions and boundary detection. We evaluate our method on two public lesion datasets, IDRiD and E-Ophtha. Our method shows competitive performance with low-resolution inputs, both quantitatively and qualitatively. On E-Ophtha dataset, the method can achieve $\ge 3$% higher dice and recall compared with the state-of-the-art methods.

Dual-Stream Learning Hard Exudates Semantic Segmentation Attention

第一 ; 通讯

IEEE

1.Department of Computer Science and Engineering, Southern University of Science and Technology, Shenzhen, Guangdong, China
2.Research Institute of Trustworthy Autonomous Systems Southern University of Science and Technology, Shenzhen, Guangdong, China
3.Department of Ophthalmology, Shenzhen People’s Hospital, Guangdong, China

Jiayi Zhang,Xiaoshan Chen,Zhongxi Qiu,et al. Hard Exudate Segmentation Supplemented by Super-Resolution with Multi-scale Attention Fusion Module[C],2022:1375-1380.