Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

AG-CUResNeSt: A Novel Method for Colon Polyp Segmentation

About

Colorectal cancer is among the most common malignancies and can develop from high-risk colon polyps. Colonoscopy is an effective screening tool to detect and remove polyps, especially in the case of precancerous lesions. However, the missing rate in clinical practice is relatively high due to many factors. The procedure could benefit greatly from using AI models for automatic polyp segmentation, which provide valuable insights for improving colon polyp detection. However, precise segmentation is still challenging due to variations of polyps in size, shape, texture, and color. This paper proposes a novel neural network architecture called AG-CUResNeSt, which enhances Coupled UNets using the robust ResNeSt backbone and attention gates. The network is capable of effectively combining multi-level features to yield accurate polyp segmentation. Experimental results on five popular benchmark datasets show that our proposed method achieves state-of-the-art accuracy compared to existing methods.

Dinh Viet Sang, Tran Quang Chung, Phan Ngoc Lan, Dao Viet Hang, Dao Van Long, Nguyen Thi Thuy• 2021

Related benchmarks

TaskDatasetResultRank
Polyp SegmentationCVC-ClinicDB (test)
DSC91.7
196
Polyp SegmentationKvasir-SEG (test)
mIoU84.5
87
Polyp SegmentationETIS (test)
Mean Dice70.1
86
Polyp SegmentationCVC-Clinic Scenario 5 (5-fold cross-validation)
mIoU90.2
9
Polyp SegmentationCVC-Clinic (test)
mDice77.1
9
Polyp SegmentationKvasir-SEG Scenario 6 (5-fold cross-validation)
mDice0.912
8
Polyp SegmentationCVC-Clinic Scenario 1 (test)
mDice83.3
7
Polyp SegmentationPolyp Segmentation
GFlops273.4
7
Showing 8 of 8 rows

Other info

Follow for update