Related Experiment Video
Updated: May 31, 2026
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
Published on: July 5, 2024
A multi-sequence MRI integration framework using SwinUNETR-v2 for multiple sclerosis lesion segmentation
Rezq Muhammed Thabet1, Howida A Shedeed2, Maryam Al-Berry2
1Department of Scientific Computing, Faculty of Computer and Information Sciences, Ain Shams University, Cairo, Egypt. rizk.mohamed@cis.asu.edu.eg.
Abstract:
Multiple Sclerosis (MS) is a chronic brain disease that affects the brain and spinal cord, where Magnetic Resonance Imaging (MRI) plays a key role in diagnosis. While manual analysis of brain MRIs is important, it is time-consuming and prone to human error. Artificial Intelligence (AI)-driven Computer Aided Diagnostic (CAD) systems have therefore gained traction due to their ability to provide more consistent and reliable assessments. This study presents a multi-sequence framework that integrates four MRI modalities with the SwinUNETR-v2 backbone for MS lesion segmentation. The main contribution is a task-oriented integration of multi-sequence input design (implemented as a four-channel volume representation), refined preprocessing (including balanced foreground/background patch extraction), and a weighted loss formulation under a controlled five-fold evaluation protocol. This approach achieved a peak DSC of 90.7% and a mean DSC of 88.3%. Moreover, when compared against other state-of-the-art segmentation methods-including AttentionUNet, DenseResidualUNet, SegResNet, FCNN, and nnUNet-v2-the multi-sequence SwinUNETR-v2 setup consistently outperformed these models across all key metrics, demonstrating strong effectiveness in identifying MS lesions.
Insights
This study introduces an AI system using multi-sequence MRI scans to accurately detect Multiple Sclerosis (MS) lesions. The advanced SwinUNETR-v2 model significantly improves lesion segmentation, aiding in faster and more reliable MS diagnosis.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Neurology
Background:
- Multiple Sclerosis (MS) is a chronic neurological disease impacting the brain and spinal cord.
- Magnetic Resonance Imaging (MRI) is crucial for MS diagnosis, but manual analysis is labor-intensive and error-prone.
- AI-driven Computer Aided Diagnostic (CAD) systems offer consistent and reliable assessments for MS detection.
Purpose of the Study:
- To develop and evaluate an AI framework for accurate Multiple Sclerosis lesion segmentation using multi-sequence MRI.
- To enhance the diagnostic process for MS by leveraging advanced AI techniques.
Main Methods:
- A multi-sequence framework integrating four MRI modalities with the SwinUNETR-v2 backbone was developed.
- Task-oriented integration included a four-channel volume representation, refined preprocessing with balanced patch extraction, and a weighted loss function.
- A five-fold cross-validation protocol was employed for rigorous evaluation.
Main Results:
- The proposed framework achieved a peak Dice Similarity Coefficient (DSC) of 90.7% and a mean DSC of 88.3% for MS lesion segmentation.
- The multi-sequence SwinUNETR-v2 model consistently outperformed other state-of-the-art segmentation methods (AttentionUNet, DenseResidualUNet, SegResNet, FCNN, nnUNet-v2).
Conclusions:
- The AI-driven multi-sequence MRI framework demonstrates strong effectiveness in identifying Multiple Sclerosis lesions.
- This approach offers a more consistent, reliable, and efficient alternative to manual analysis for MS diagnosis.
