Dysplasia grading of colorectal polyps through CNN analysis of WSI

TL;DR

This paper presents a deep learning pipeline using CNNs to classify dysplasia grades in colorectal polyps from whole slide images, achieving accuracy comparable to pathologists.

Contribution

It introduces a CNN-based method specifically designed for high-resolution, imbalanced WSI data to classify dysplasia grades in colorectal polyps.

Findings

Achieved 70% accuracy in dysplasia classification

Method handles high-resolution WSI and dataset imbalance effectively

Results align with pathologists' concordance levels

Abstract

Colorectal cancer is a leading cause of cancer death for both men and women. For this reason, histopathological characterization of colorectal polyps is the major instrument for the pathologist in order to infer the actual risk for cancer and to guide further follow-up. Colorectal polyps diagnosis includes the evaluation of the polyp type, and more importantly, the grade of dysplasia. This latter evaluation represents a critical step for the clinical follow-up. The proposed deep learning-based classification pipeline is based on state-of-the-art convolutional neural network, trained using proper countermeasures to tackle WSI high resolution and very imbalanced dataset. The experimental results show that one can successfully classify adenomas dysplasia grade with 70% accuracy, which is in line with the pathologists' concordance.