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

Zhenyu Wu,Fengmao Lv,Chenglizhao Chen,Aimin Hao,Shuo Li

2024-01-22

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

Computer Vision and Pattern Recognition

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the key challenges in colorectal polyp segmentation (CPS) in medical image analysis. Specifically, the paper focuses on how to use deep - learning techniques to improve the accuracy and automation level of CPS, thereby assisting doctors in diagnosing and treating polyps that may develop into colorectal cancer earlier. The paper points out that traditional polyp segmentation methods mainly rely on manually - extracted low - level features, such as texture, shape or color distribution, and these methods are not effective in handling complex scenarios. With the development of deep - learning techniques, more and more methods based on convolutional neural networks (CNN) and transformers have been proposed for polyp segmentation. These methods far outperform traditional methods in performance, but lack a comprehensive review to systematically summarize the characteristics, datasets, evaluation metrics and challenges of these methods.

To this end, the paper provides a systematic and comprehensive review, covering 115 technical papers from 2014 to 2023, aiming to help researchers quickly master the main techniques, datasets, evaluation metrics, challenges and development trends of deep learning in the CPS field. In particular, the paper proposes a new taxonomy, including three main categories: network architecture, supervision level and learning paradigm, and each category is further subdivided into multiple sub - categories, with a total of 26 sub - categories. In addition, the paper also conducts a comprehensive analysis of the characteristics of each dataset and a detailed performance analysis of 40 of the latest deep - learning models, including generalization ability and attribute - based performance analysis. Finally, the paper discusses the main challenges and opportunities of deep - learning CPS methods, providing directions for future research.