Colorectal Polyp Segmentation in the Deep Learning Era: A Comprehensive Survey

• DOI: 10.48550/arxiv.2401.11734
• Publication Date: 2024-01-01
• Authors: Zhenyu Wu; Fengmao Lv; Chenglizhao Chen; Aimin Hao; Shuo Li

Colorectal polyp segmentation (CPS), an essential problem in medical image analysis, has garnered growing research attention. Recently, the deep learning-based model completely overwhelmed traditional methods field of CPS, and more CPS have emerged, bringing into learning era. To help researchers quickly grasp main techniques, datasets, evaluation metrics, challenges, trending this paper presents a systematic comprehensive review deep-learning-based from 2014 to 2023, total 115 technical papers. In particular, we first provide current with novel taxonomy, including network architectures, level supervision, paradigm. More specifically, architectures include eight subcategories, supervision comprises six paradigm encompasses 12 totaling 26 subcategories. Then, provided analysis characteristics each dataset, number annotation types, resolution, size, contrast values, location. Following that, summarized CPS's commonly used metrics conducted detailed 40 SOTA models, out-of-distribution generalization attribute-based performance analysis. Finally, discussed methods' challenges opportunities.