Dataset Distillation for Histopathology Image Classification

TL;DR

This paper introduces Histo-DD, a dataset distillation algorithm for histopathology images that creates synthetic samples to improve training efficiency and reduce dataset size while maintaining classification performance.

Contribution

The paper presents a novel dataset distillation method tailored for histopathology images, integrating stain normalization and model augmentation to enhance synthetic sample quality.

Findings

Histo-DD generates more informative synthetic patches than previous methods.

Synthetic samples preserve discriminative information effectively.

The approach reduces training efforts and is architecture-agnostic.

Abstract

Deep neural networks (DNNs) have exhibited remarkable success in the field of histopathology image analysis. On the other hand, the contemporary trend of employing large models and extensive datasets has underscored the significance of dataset distillation, which involves compressing large-scale datasets into a condensed set of synthetic samples, offering distinct advantages in improving training efficiency and streamlining downstream applications. In this work, we introduce a novel dataset distillation algorithm tailored for histopathology image datasets (Histo-DD), which integrates stain normalisation and model augmentation into the distillation progress. Such integration can substantially enhance the compatibility with histopathology images that are often characterised by high colour heterogeneity. We conduct a comprehensive evaluation of the effectiveness of the proposed algorithm and the generated histopathology samples in both patch-level and slide-level classification tasks. The experimental results, carried out on three publicly available WSI datasets, including Camelyon16, TCGA-IDH, and UniToPath, demonstrate that the proposed Histo-DD can generate more informative synthetic patches than previous coreset selection and patch sampling methods. Moreover, the synthetic samples can preserve discriminative information, substantially reduce training efforts, and exhibit architecture-agnostic properties. These advantages indicate that synthetic samples can serve as an alternative to large-scale datasets.