Calibrating Histopathology Image Classifiers using Label Smoothing

TL;DR

This paper introduces label smoothing techniques that incorporate annotator agreement to improve the calibration of histopathology image classifiers, leading to more reliable confidence estimates without sacrificing accuracy.

Contribution

It proposes agreement-aware label smoothing methods that leverage per-image annotator agreement to enhance model calibration in histopathology classification tasks.

Findings

Agreement-aware label smoothing reduces calibration error by nearly 70%.

Methods improve calibration while maintaining or increasing accuracy.

Using model confidence as a proxy for agreement benefits datasets without multiple annotators.

Abstract

The classification of histopathology images fundamentally differs from traditional image classification tasks because histopathology images naturally exhibit a range of diagnostic features, resulting in a diverse range of annotator agreement levels. However, examples with high annotator disagreement are often either assigned the majority label or discarded entirely when training histopathology image classifiers. This widespread practice often yields classifiers that do not account for example difficulty and exhibit poor model calibration. In this paper, we ask: can we improve model calibration by endowing histopathology image classifiers with inductive biases about example difficulty? We propose several label smoothing methods that utilize per-image annotator agreement. Though our methods are simple, we find that they substantially improve model calibration, while maintaining (or even improving) accuracy. For colorectal polyp classification, a common yet challenging task in gastrointestinal pathology, we find that our proposed agreement-aware label smoothing methods reduce calibration error by almost 70%. Moreover, we find that using model confidence as a proxy for annotator agreement also improves calibration and accuracy, suggesting that datasets without multiple annotators can still benefit from our proposed label smoothing methods via our proposed confidence-aware label smoothing methods. Given the importance of calibration (especially in histopathology image analysis), the improvements from our proposed techniques merit further exploration and potential implementation in other histopathology image classification tasks.