GCLDNet: Gastric cancer lesion detection network combining level feature aggregation and attention feature fusion

Xu Shi; Long Wang; Yu Li; Jian Wu; Hong Huang

doi:10.3389/fonc.2022.901475

GCLDNet: Gastric cancer lesion detection network combining level feature aggregation and attention feature fusion

Front Oncol. 2022 Aug 29:12:901475. doi: 10.3389/fonc.2022.901475. eCollection 2022.

Authors

Xu Shi¹, Long Wang¹, Yu Li², Jian Wu³, Hong Huang¹

Affiliations

¹ Key Laboratory of Optoelectronic Technology and Systems of the Education Ministry of China, Chongqing University, Chongqing, China.
² Department of Pathology, Chongqing University Cancer Hospital and Chongqing Cancer Institute and Chongqing Cancer Hospital, Chongqing, China.
³ Head and Neck Cancer Center, Chongqing University Cancer Hospital and Chongqing Cancer Institute and Chongqing Cancer Hospital, Chongqing, China.

Abstract

Background: Analysis of histopathological slices of gastric cancer is the gold standard for diagnosing gastric cancer, while manual identification is time-consuming and highly relies on the experience of pathologists. Artificial intelligence methods, particularly deep learning, can assist pathologists in finding cancerous tissues and realizing automated detection. However, due to the variety of shapes and sizes of gastric cancer lesions, as well as many interfering factors, GCHIs have a high level of complexity and difficulty in accurately finding the lesion region. Traditional deep learning methods cannot effectively extract discriminative features because of their simple decoding method so they cannot detect lesions accurately, and there is less research dedicated to detecting gastric cancer lesions.

Methods: We propose a gastric cancer lesion detection network (GCLDNet). At first, GCLDNet designs a level feature aggregation structure in decoder, which can effectively fuse deep and shallow features of GCHIs. Second, an attention feature fusion module is introduced to accurately locate the lesion area, which merges attention features of different scales and obtains rich discriminative information focusing on lesion. Finally, focal Tversky loss (FTL) is employed as a loss function to depress false-negative predictions and mine difficult samples.

Results: Experimental results on two GCHI datasets of SEED and BOT show that DSCs of the GCLDNet are 0.8265 and 0.8991, ACCs are 0.8827 and 0.8949, JIs are 0.7092 and 0.8182, and PREs are 0.7820 and 0.8763, respectively.

Conclusions: Experimental results demonstrate the effectiveness of GCLDNet in the detection of gastric cancer lesions. Compared with other state-of-the-art (SOTA) detection methods, the GCLDNet obtains a more satisfactory performance. This research can provide good auxiliary support for pathologists in clinical diagnosis.

Keywords: artificial intelligence; attention feature fusion; convolutional neural network; deep learning; gastric cancer lesion detection; image segmentation; level feature aggregation.