Automatic multiple sclerosis lesion detection in brain MRI by FLAIR thresholding
Mariano Cabezas1, Arnau Oliver1, Eloy Roura1
1Department of Computer Architecture and Technology, University of Girona, Spain.
Abstract:
Magnetic resonance imaging (MRI) is frequently used to detect and segment multiple sclerosis lesions due to the detailed and rich information provided. We present a modified expectation-maximisation algorithm to segment brain tissues (white matter, grey matter, and cerebro-spinal fluid) as well as a partial volume class containing fluid and grey matter. This algorithm provides an initial segmentation in which lesions are not separated from tissue, thus a second step is needed to find them. This second step involves the thresholding of the FLAIR image, followed by a regionwise refinement to discard false detections. To evaluate the proposal, we used a database with 45 cases comprising 1.5T imaging data from three different hospitals with different scanner machines and with a variable lesion load per case. The results for our database point out to a higher accuracy when compared to two of the best state-of-the-art approaches.
Insights
A new modified expectation-maximisation algorithm improves multiple sclerosis (MS) lesion detection and segmentation in MRI scans. This method enhances accuracy compared to existing state-of-the-art approaches for brain tissue and lesion segmentation.
Area of Science:
- Medical Imaging
- Neuroscience
- Computer Vision
Background:
- Magnetic resonance imaging (MRI) is crucial for detecting and segmenting multiple sclerosis (MS) lesions.
- Accurate segmentation of brain tissues and MS lesions is essential for diagnosis and monitoring.
Purpose of the Study:
- To present a modified expectation-maximisation algorithm for improved segmentation of brain tissues and MS lesions.
- To enhance the accuracy of MS lesion detection and segmentation in MRI data.
Main Methods:
- A modified expectation-maximisation algorithm was developed for initial segmentation of brain tissues (white matter, grey matter, CSF) and partial volume classes.
- A two-step approach was employed, including thresholding of FLAIR images and regionwise refinement for lesion detection.
- The algorithm was evaluated on a diverse database of 45 cases with 1.5T MRI data from multiple hospitals and scanners.
Main Results:
- The proposed algorithm demonstrated higher accuracy in MS lesion segmentation compared to two leading state-of-the-art methods.
- The method effectively segmented brain tissues and identified lesions, even with variable lesion loads.
- Validation across different scanner machines and hospitals confirmed the robustness of the approach.
Conclusions:
- The modified expectation-maximisation algorithm offers a more accurate and robust solution for MS lesion detection and segmentation.
- This advancement in MRI analysis can improve the clinical assessment and management of multiple sclerosis.
- The proposed method shows significant potential for widespread adoption in MS research and clinical practice.
