Brain MR image segmentation improves with combined pretraining tasks
Combining General and Domain-Specific Pretext Tasks for Brain MR Image Segmentation
Computer Vision and Pattern RecognitionArtificial Intelligence
Summary
Deep learning needs lots of labeled brain scans to learn how to identify different regions or lesions, but such labeled data is hard to get. The authors studied two ways for a computer to learn from unlabeled brain scans: predicting brain age voxel-by-voxel (a brain-specific task) and filling in missing parts of images (a general task). They found that training a model to do both tasks at the same time helps it learn better features for segmenting brain images. This combined approach improved performance when detecting multiple sclerosis lesions, stroke lesions, and brain structures.
What this means in practice
- •For medical imaging software developers: Create brain MRI segmentation tools that need fewer manual labels by using multitask self-supervised pretraining to improve lesion and structure detection accuracy.
- •For healthcare ai platform engineers: Enhance brain disease diagnosis systems with foundation models pretrained on combined domain-specific and general tasks to better transfer to varied clinical MR datasets.$Commercial implications: Enables developing advanced diagnostic AI products with improved segmentation precision for diverse brain conditions, attractive to hospitals and medical device companies.
Authors
Tasneem Nasser, Susanne Schmid, Roberto Souza, Naser El-Sheimy
Abstract
A key challenge in medical image analysis is the scarcity of large annotated datasets for specific populations and diseases. As deep learning models rely heavily on labeled data, effective transfer learning strategies are needed to reduce the dependence on manual annotations. Self-supervised learning has emerged as a promising approach for developing foundation models by enabling the learning of transferable feature representations from large-scale unlabeled medical imaging datasets. In this study, we investigate voxel-level brain age prediction as a domain-specific self-supervised pretext task and compare it with image inpainting, a widely used non-domain-specific alternative. We further propose a multitask self-supervised pretraining framework that jointly optimizes both objectives to learn complementary neuroimaging representations. The pretrained models are evaluated on three downstream magnetic resonance image segmentation tasks: multiple sclerosis lesion segmentation, ischemic stroke lesion segmentation, and cortical brain structure segmentation. Overall, the proposed multitask pretraining framework consistently outperformed the single-task pretrained models and training from scratch across most experimental settings, demonstrating the benefit of combining domain-specific and general self-supervised learning pretext tasks for the development of generalizable neuroimaging foundation models.\ Code Availability: The source code used in this study is publicly available at https://github.com/TasneemN/Combining-General-and-Domain-Specific-Pretext-Tasks-for-Brain-MR-Image-Segmentation/