NINEPINS: Nuclei Instance Segmentation with Point Annotations

TL;DR

This paper introduces a nuclei instance segmentation method using point annotations to generate pseudo-labels, reducing the annotation burden in digital pathology while maintaining robust segmentation performance.

Contribution

It presents a novel approach that leverages automatically generated pseudo-labels from point annotations to train a modified HoVer-Net for nuclei segmentation.

Findings

Robust to inaccuracies in point annotations

Comparable tissue classification performance despite segmentation degradation

Reduces annotation effort in digital pathology

Abstract

Deep learning-based methods are gaining traction in digital pathology, with an increasing number of publications and challenges that aim at easing the work of systematically and exhaustively analyzing tissue slides. These methods often achieve very high accuracies, at the cost of requiring large annotated datasets to train. This requirement is especially difficult to fulfill in the medical field, where expert knowledge is essential. In this paper we focus on nuclei segmentation, which generally requires experienced pathologists to annotate the nuclear areas in gigapixel histological images. We propose an algorithm for instance segmentation that uses pseudo-label segmentations generated automatically from point annotations, as a method to reduce the burden for pathologists. With the generated segmentation masks, the proposed method trains a modified version of HoVer-Net model to achieve instance segmentation. Experimental results show that the proposed method is robust to inaccuracies in point annotations and comparison with Hover-Net trained with fully annotated instance masks shows that a degradation in segmentation performance does not always imply a degradation in higher order tasks such as tissue classification.