Scale dependant layer for self-supervised nuclei encoding

TL;DR

This paper introduces a scale-dependent convolutional layer for self-supervised nuclei encoding in histopathology images, improving unsupervised cellular feature extraction across varying nuclei sizes.

Contribution

It proposes a novel scale-dependent convolutional layer that enhances self-supervised nuclei encoding, outperforming existing methods in low sample settings.

Findings

The new layer boosts encoding performance across datasets.

Combining the layer with Barlows-Twins improves nuclei representation.

Outperforms supervised and other unsupervised methods in experiments.

Abstract

Recent developments in self-supervised learning give us the possibility to further reduce human intervention in multi-step pipelines where the focus evolves around particular objects of interest. In the present paper, the focus lays in the nuclei in histopathology images. In particular we aim at extracting cellular information in an unsupervised manner for a downstream task. As nuclei present themselves in a variety of sizes, we propose a new Scale-dependant convolutional layer to bypass scaling issues when resizing nuclei. On three nuclei datasets, we benchmark the following methods: handcrafted, pre-trained ResNet, supervised ResNet and self-supervised features. We show that the proposed convolution layer boosts performance and that this layer combined with Barlows-Twins allows for better nuclei encoding compared to the supervised paradigm in the low sample setting and outperforms all other proposed unsupervised methods. In addition, we extend the existing TNBC dataset to incorporate nuclei class annotation in order to enrich and publicly release a small sample setting dataset for nuclei segmentation and classification.