UniToPatho, a labeled histopathological dataset for colorectal polyps classification and adenoma dysplasia grading

TL;DR

UniToPatho is a comprehensive annotated dataset of over 9,500 histopathological images designed to improve deep learning models for colorectal polyps classification and adenoma grading, aiding in better diagnosis and patient management.

Contribution

The paper introduces UniToPatho, a large labeled dataset specifically created for training deep neural networks in colorectal polyps classification and dysplasia grading, addressing data scarcity in this domain.

Findings

Dataset contains 9536 annotated patches from 292 slides.

Provides insights into automatic colorectal polyps characterization.

Facilitates development of deep learning models for pathology analysis.

Abstract

Histopathological characterization of colorectal polyps allows to tailor patients' management and follow up with the ultimate aim of avoiding or promptly detecting an invasive carcinoma. Colorectal polyps characterization relies on the histological analysis of tissue samples to determine the polyps malignancy and dysplasia grade. Deep neural networks achieve outstanding accuracy in medical patterns recognition, however they require large sets of annotated training images. We introduce UniToPatho, an annotated dataset of 9536 hematoxylin and eosin (H&E) stained patches extracted from 292 whole-slide images, meant for training deep neural networks for colorectal polyps classification and adenomas grading. We present our dataset and provide insights on how to tackle the problem of automatic colorectal polyps characterization.