Domain Aware Multi-Task Pretraining of 3D Swin Transformer for T1-weighted Brain MRI

TL;DR

This paper introduces a domain-aware multi-task pretraining approach for a 3D Swin Transformer tailored to brain MRI data, leveraging large-scale datasets and domain knowledge to improve performance on medical diagnosis tasks.

Contribution

It proposes a novel multi-task pretraining method incorporating domain knowledge and contrastive learning for 3D brain MRI analysis, outperforming existing methods.

Findings

Outperforms existing supervised and self-supervised methods in downstream tasks.

Effective use of large-scale brain MRI data (13,687 samples).

Ablation study confirms the importance of proposed pretext tasks.

Abstract

The scarcity of annotated medical images is a major bottleneck in developing learning models for medical image analysis. Hence, recent studies have focused on pretrained models with fewer annotation requirements that can be fine-tuned for various downstream tasks. However, existing approaches are mainly 3D adaptions of 2D approaches ill-suited for 3D medical imaging data. Motivated by this gap, we propose novel domain-aware multi-task learning tasks to pretrain a 3D Swin Transformer for brain magnetic resonance imaging (MRI). Our method considers the domain knowledge in brain MRI by incorporating brain anatomy and morphology as well as standard pretext tasks adapted for 3D imaging in a contrastive learning setting. We pretrain our model using large-scale brain MRI data of 13,687 samples spanning several large-scale databases. Our method outperforms existing supervised and self-supervised methods in three downstream tasks of Alzheimer's disease classification, Parkinson's disease classification, and age prediction tasks. The ablation study of the proposed pretext tasks shows the effectiveness of our pretext tasks.