Seeded iterative clustering for histology region identification

TL;DR

This paper introduces seeded iterative clustering, a method that efficiently generates dense histology annotations at the whole slide level using limited annotations, reducing the need for extensive manual labeling and computational resources.

Contribution

It presents a novel seeded iterative clustering algorithm that produces dense histology segmentations and enables comparison of neural network representations for transfer learning.

Findings

Fast dense annotation generation for whole slide images

Effective classification of image patches with limited annotations

Framework for comparing neural network latent representations

Abstract

Annotations are necessary to develop computer vision algorithms for histopathology, but dense annotations at a high resolution are often time-consuming to make. Deep learning models for segmentation are a way to alleviate the process, but require large amounts of training data, training times and computing power. To address these issues, we present seeded iterative clustering to produce a coarse segmentation densely and at the whole slide level. The algorithm uses precomputed representations as the clustering space and a limited amount of sparse interactive annotations as seeds to iteratively classify image patches. We obtain a fast and effective way of generating dense annotations for whole slide images and a framework that allows the comparison of neural network latent representations in the context of transfer learning.