Handcrafted Histological Transformer (H2T): Unsupervised Representation of Whole Slide Images

TL;DR

This paper introduces H2T, a handcrafted, transparent deep learning framework for representing whole-slide histological images, achieving competitive accuracy and significantly faster processing compared to existing Transformer-based methods.

Contribution

The paper presents a novel handcrafted approach that interprets Transformer processes for histological image analysis, enhancing transparency and speed while maintaining competitive performance.

Findings

H2T achieves comparable accuracy to state-of-the-art methods.

H2T is up to 14 times faster than traditional Transformer models.

The framework is effective across multiple datasets with over 5,300 WSIs.

Abstract

Diagnostic, prognostic and therapeutic decision-making of cancer in pathology clinics can now be carried out based on analysis of multi-gigapixel tissue images, also known as whole-slide images (WSIs). Recently, deep convolutional neural networks (CNNs) have been proposed to derive unsupervised WSI representations; these are attractive as they rely less on expert annotation which is cumbersome. However, a major trade-off is that higher predictive power generally comes at the cost of interpretability, posing a challenge to their clinical use where transparency in decision-making is generally expected. To address this challenge, we present a handcrafted framework based on deep CNN for constructing holistic WSI-level representations. Building on recent findings about the internal working of the Transformer in the domain of natural language processing, we break down its processes and handcraft them into a more transparent framework that we term as the Handcrafted Histological Transformer or H2T. Based on our experiments involving various datasets consisting of a total of 5,306 WSIs, the results demonstrate that H2T based holistic WSI-level representations offer competitive performance compared to recent state-of-the-art methods and can be readily utilized for various downstream analysis tasks. Finally, our results demonstrate that the H2T framework can be up to 14 times faster than the Transformer models.