crossMoDA challenge: Evolution of cross-modality domain adaptation techniques for vestibular schwannoma and cochlea segmentation from 2021 to 2023

The cross-Modality Domain Adaptation (crossMoDA) challenge series, initiated in 2021 in conjunction with the International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), focuses on unsupervised cross-modality segmentation, learning from contrast-enhanced T1 (ceT1) and transferring to T2 MRI....

Continue reading at PubMed →

No actionable change for current clinical practice; these are research-stage AI segmentation tools that have not yet been validated for routine clinical deployment.

Automated, accurate segmentation of vestibular schwannoma and cochlea across MRI modalities could eventually reduce radiologist workload and improve treatment planning consistency for skull-base tumors.

1. 01CrossMoDA is a multi-year international challenge benchmarking AI-driven cross-modality MRI segmentation.
2. 02Targets are vestibular schwannoma (a benign ear tumor) and cochlea across unpaired MRI sequences.
3. 03Performance of submitted algorithms improved measurably from 2021 to 2023.
4. 04Challenge data and benchmarks are publicly available, supporting reproducible research.
5. 05Results remain at the research stage and are not yet integrated into clinical workflows.

PMID

42673840

DOI

10.1016/j.media.2026.104282.

Journal

Medical Image Analysis

Publication type

review

Evidence level

2a

Population

Algorithm submissions from international research teams targeting MRI datasets containing vestibular schwannoma and cochlea

Intervention

Cross-modality domain adaptation algorithms for automated MRI segmentation of vestibular schwannoma and cochlea

Comparator

Successive challenge editions (2021, 2022, 2023) benchmarked against each other