LoGex: Improved tail detection of extremely rare histopathology classes via guided diffusion

TL;DR

This paper introduces LoGex, a method that uses guided diffusion and low-rank adaptation to improve detection of extremely rare classes in histopathology, addressing class imbalance without sacrificing overall accuracy.

Contribution

The paper proposes a novel approach combining LoRA and diffusion guidance to generate synthetic data for rare classes, enhancing out-of-distribution detection in long-tailed medical datasets.

Findings

Significant improvement in OOD detection performance on histopathology data.

Effective detection of tail classes with only ten samples per class.

Maintains classification accuracy on common classes.

Abstract

In realistic medical settings, the data are often inherently long-tailed, with most samples concentrated in a few classes and a long tail of rare classes, usually containing just a few samples. This distribution presents a significant challenge because rare conditions are critical to detect and difficult to classify due to limited data. In this paper, rather than attempting to classify rare classes, we aim to detect these as out-of-distribution data reliably. We leverage low-rank adaption (LoRA) and diffusion guidance to generate targeted synthetic data for the detection problem. We significantly improve the OOD detection performance on a challenging histopathological task with only ten samples per tail class without losing classification accuracy on the head classes.