Nuclear Segmentation and Classification: On Color & Compression Generalization

TL;DR

This paper evaluates the robustness of nuclear segmentation and classification models in computational pathology, revealing that models are sensitive to color shifts but resilient to compression artifacts, and that neural style transfer can improve performance under color variation.

Contribution

It provides a comprehensive evaluation of model robustness to color and compression variations and highlights the effectiveness of neural style transfer for handling color domain shifts.

Findings

Models are robust to compression artifacts.

Models' performance drops with color domain shifts.

Neural style transfer improves robustness to color variations.

Abstract

Since the introduction of digital and computational pathology as a field, one of the major problems in the clinical application of algorithms has been the struggle to generalize well to examples outside the distribution of the training data. Existing work to address this in both pathology and natural images has focused almost exclusively on classification tasks. We explore and evaluate the robustness of the 7 best performing nuclear segmentation and classification models from the largest computational pathology challenge for this problem to date, the CoNIC challenge. We demonstrate that existing state-of-the-art (SoTA) models are robust towards compression artifacts but suffer substantial performance reduction when subjected to shifts in the color domain. We find that using stain normalization to address the domain shift problem can be detrimental to the model performance. On the other hand, neural style transfer is more consistent in improving test performance when presented with large color variations in the wild.