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

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. 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.

Radial Prediction Domain Adaption Classifier for the MIDOG 2022 challenge

In this paper, we describe our contribution to the MIDOG 2022 challenge without using additional data. A challenge to handle the distribution shift between different tissues for detection of mitosis cells. The main characteristics parts can be distinguished into three parts: We modify the Radial Prediction Layer (RPL) to integrate the layer in a domain adaption classifier, the Prediction Domain Adaption Classifier (RP-DAC). This developed variant learns prototypes for each class and brings more related classes closer. We used this to learn the scanner, the tissue, and the case id. We used multiple trained YOLO models with different modified input variants of the image. We combine the outputs of the model with an ensembling strategy. We use the HED color space for data augmentation by calculating different magnitudes for each scanner/tissue type to create more variance in the training set.