Home  | Publications | EBS+25

MultiMAE for Brain MRIs: Robustness to Missing Inputs Using Multi-Modal Masked Autoencoder

MCML Authors

Link to Profile Benedikt Wiestler PI Matchmaking

Benedikt Wiestler

Prof. Dr.

Principal Investigator

AI for Image-Guided Diagnosis and Therapy

Abstract

Missing input sequences are common in medical imaging data, posing a challenge for deep learning models reliant on complete input data. In this work, inspired by MultiMAE [2], we develop a masked autoencoder (MAE) paradigm for multi-modal, multi-task learning in 3D medical imaging with brain MRIs. Our method treats each MRI sequence as a separate input modality, leveraging a late-fusion-style transformer encoder to integrate multi-sequence information (“multi-modal”) and individual decoder streams for each modality for “multi-task” reconstruction. This pretraining strategy guides the model to learn rich representations per modality while also equipping it to handle missing inputs through cross-sequence reasoning. The result is a flexible and generalizable encoder for brain MRIs that infers missing sequences from available inputs and can be adapted to various downstream applications. We demonstrate the performance and robustness of our method against an MAE-ViT baseline in downstream segmentation and classification tasks, showing absolute improvement of 10.1 overall Dice score and 0.46 MCC over the baselines with missing input sequences. Our experiments demonstrate the strength of this pretraining strategy. The implementation is made available.

inproceedings EBS+25

MLMI @MICCAI 2025

16th International Workshop on Machine Learning in Medical Imaging at the 28th International Conference on Medical Image Computing and Computer Assisted Intervention. Daejeon, Republic of Korea, Sep 23-27, 2025.

Authors

A. C. Erdur • C. Beischl • D. Scholz • J. Pan • B. Wiestler • D. Rückert • J. C. Peeken

Links

DOI GitHub

Research Area

 C1 | Medicine

BibTeXKey: EBS+25

#p-rueckert #p-wiestler
Back to Top