Diagnosis and Severity Assessment of Ulcerative Colitis using Self Supervised Learning

TL;DR

This paper introduces a self-supervised learning approach for diagnosing and assessing the severity of Ulcerative Colitis from colonoscopy images, reducing the need for large annotated datasets and outperforming supervised models.

Contribution

It presents the first application of SSL frameworks like SwAV and SparK for UC diagnosis and severity assessment, demonstrating superior performance over supervised methods.

Findings

SSL models outperform supervised models on the LIMUC dataset

Self-supervised learning reduces dependency on annotated data

SSL frameworks achieve higher accuracy in UC severity classification

Abstract

Ulcerative Colitis (UC) is an incurable inflammatory bowel disease that leads to ulcers along the large intestine and rectum. The increase in the prevalence of UC coupled with gastrointestinal physician shortages stresses the healthcare system and limits the care UC patients receive. A colonoscopy is performed to diagnose UC and assess its severity based on the Mayo Endoscopic Score (MES). The MES ranges between zero and three, wherein zero indicates no inflammation and three indicates that the inflammation is markedly high. Artificial Intelligence (AI)-based neural network models, such as convolutional neural networks (CNNs) are capable of analyzing colonoscopies to diagnose and determine the severity of UC by modeling colonoscopy analysis as a multi-class classification problem. Prior research for AI-based UC diagnosis relies on supervised learning approaches that require large annotated datasets to train the CNNs. However, creating such datasets necessitates that domain experts invest a significant amount of time, rendering the process expensive and challenging. To address the challenge, this research employs self-supervised learning (SSL) frameworks that can efficiently train on unannotated datasets to analyze colonoscopies and, aid in diagnosing UC and its severity. A comparative analysis with supervised learning models shows that SSL frameworks, such as SwAV and SparK outperform supervised learning models on the LIMUC dataset, the largest publicly available annotated dataset of colonoscopy images for UC.