Learning Melanocytic Cell Masks from Adjacent Stained Tissue

TL;DR

This paper introduces a deep learning approach to segment melanocytic cells in skin tissue images using paired stained sections, aiming to improve melanoma diagnosis consistency and automate pixel-level annotation.

Contribution

The study presents a novel method for training neural networks for melanocytic cell segmentation using paired H&E and IHC tissue sections, overcoming the need for extensive manual labeling.

Findings

Achieved a mean IOU of 0.64 in segmentation accuracy.

Demonstrated the feasibility of using paired tissue sections for training.

Reduced reliance on labor-intensive pathologist annotations.

Abstract

Melanoma is one of the most aggressive forms of skin cancer, causing a large proportion of skin cancer deaths. However, melanoma diagnoses by pathologists shows low interrater reliability. As melanoma is a cancer of the melanocyte, there is a clear need to develop a melanocytic cell segmentation tool that is agnostic to pathologist variability and automates pixel-level annotation. Gigapixel-level pathologist labeling, however, is impractical. Herein, we propose a means to train deep neural networks for melanocytic cell segmentation from hematoxylin and eosin (H&E) stained sections and paired immunohistochemistry (IHC) of adjacent tissue sections, achieving a mean IOU of 0.64 despite imperfect ground-truth labels.