Home Publications Research Grants Inventions & Patents Awards Additional Research Activities Faculties & Institutions Research Areas

Self Pre-Training with Adaptive Mask Autoencoders for Variable-Contrast 3D Medical Imaging

Das BK, Zhao G, Liu H, Re TJ, Comaniciu D, Gibson E, Maier A (2025)

---

Publication Language: English

Publication Type: Conference contribution

Publication year: 2025

Journal

Publisher: IEEE Computer Society

Conference Proceedings Title: 2025 IEEE 22nd International Symposium on Biomedical Imaging (ISBI)

Event location: Houston, TX

ISBN: 979-8-3315-2053-3

DOI: 10.1109/ISBI60581.2025.10981097

Abstract

The Masked Autoencoder (MAE) has recently demonstrated effectiveness in pre-training Vision Transformers (ViT) for analyzing natural images. By reconstructing complete images from partially masked inputs, the ViT encoder gathers contextual information to predict the missing regions. This capability to aggregate context is especially important in medical imaging, where anatomical structures are functionally and mechanically linked to surrounding regions. However, current methods do not consider variations in the number of input images, which is typically the case in realworld Magnetic Resonance (MR) studies. To address this limitation, we propose a 3D Adaptive Masked Autoencoders (AMAE) architecture that accommodates a variable number of 3D input contrasts per subject. A magnetic resonance imaging (MRI) dataset of 45,364 subjects was used for pretraining and a subset of 1648 training, 193 validation and 215 test subjects were used for finetuning. The performance demonstrates that self pre-training of this adaptive masked autoencoders can enhance the infarct segmentation performance by 2.8%-3.7% for ViT-based segmentation models.

Authors with CRIS profile

Involved external institutions

How to cite

APA:

Das, B.K., Zhao, G., Liu, H., Re, T.J., Comaniciu, D., Gibson, E., & Maier, A. (2025). Self Pre-Training with Adaptive Mask Autoencoders for Variable-Contrast 3D Medical Imaging. In 2025 IEEE 22nd International Symposium on Biomedical Imaging (ISBI). Houston, TX, US: IEEE Computer Society.

MLA:

Das, Badhan Kumar, et al. "Self Pre-Training with Adaptive Mask Autoencoders for Variable-Contrast 3D Medical Imaging." Proceedings of the 22nd IEEE International Symposium on Biomedical Imaging, ISBI 2025, Houston, TX IEEE Computer Society, 2025.

BibTeX: Download