Domain generalization across tumor types, laboratories, and species -- insights from the 2022 edition of the Mitosis Domain Generalization Challenge

TL;DR

This paper reviews the 2022 Mitosis Domain Generalization Challenge, demonstrating that deep learning models can generalize across tumor types, laboratories, and species, but performance drops with unseen domain characteristics.

Contribution

It provides an overview of challenge tasks, participant strategies, and factors influencing success in domain generalization for mitosis detection in histologic images.

Findings

Top F1 score of 0.764 shows feasibility of domain generalization.

Performance decreases with unseen species, morphology, and scanner.

All methods show reduced recall against immunohistochemistry standards.

Abstract

Recognition of mitotic figures in histologic tumor specimens is highly relevant to patient outcome assessment. This task is challenging for algorithms and human experts alike, with deterioration of algorithmic performance under shifts in image representations. Considerable covariate shifts occur when assessment is performed on different tumor types, images are acquired using different digitization devices, or specimens are produced in different laboratories. This observation motivated the inception of the 2022 challenge on MItosis Domain Generalization (MIDOG 2022). The challenge provided annotated histologic tumor images from six different domains and evaluated the algorithmic approaches for mitotic figure detection provided by nine challenge participants on ten independent domains. Ground truth for mitotic figure detection was established in two ways: a three-expert consensus and an independent, immunohistochemistry-assisted set of labels. This work represents an overview of the challenge tasks, the algorithmic strategies employed by the participants, and potential factors contributing to their success. With an $F_1$ score of 0.764 for the top-performing team, we summarize that domain generalization across various tumor domains is possible with today's deep learning-based recognition pipelines. However, we also found that domain characteristics not present in the training set (feline as new species, spindle cell shape as new morphology and a new scanner) led to small but significant decreases in performance. When assessed against the immunohistochemistry-assisted reference standard, all methods resulted in reduced recall scores, but with only minor changes in the order of participants in the ranking.