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

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kim, Jonghun, Kim, Mansu, Park, Hyunjin
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913525077639168
author Kim, Jonghun
Kim, Mansu
Park, Hyunjin
author_facet Kim, Jonghun
Kim, Mansu
Park, Hyunjin
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2410_00410
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Domain Aware Multi-Task Pretraining of 3D Swin Transformer for T1-weighted Brain MRI
Kim, Jonghun
Kim, Mansu
Park, Hyunjin
Image and Video Processing
Computer Vision and Pattern Recognition
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.
title Domain Aware Multi-Task Pretraining of 3D Swin Transformer for T1-weighted Brain MRI
topic Image and Video Processing
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2410.00410