Related Experiment Video
Updated: Dec 9, 2025
09:41
Comprehensive Autopsy Program for Individuals with Multiple Sclerosis
Published on: July 19, 2019
11.8K
Normal appearing brain white matter changes in relapsing multiple sclerosis: Texture image and classification
Christos P Loizou1, Marios Pantzaris2, Constandinos S Pattichis3
1Faculty of Engineering & Technology, Department of Electrical Engineering and Computer Engineering and Informatics, Cyprus University of Technology, 30 Arch. Kyprianos Str., Limassol CY-3036, Cyprus.
Magnetic Resonance Imaging
|September 5, 2020
Summary
Texture analysis of normal-appearing white matter (NAWM) in multiple sclerosis (MS) brain MRI scans can predict disease progression. This method accurately identifies early changes in NAWM, aiding in monitoring MS development.
Area of Science:
- Neuroimaging
- Medical Physics
- Biomedical Engineering
Background:
- Multiple Sclerosis (MS) is a central nervous system disease characterized by demyelination.
- Normal-appearing white matter (NAWM) in MS patients is susceptible to disease development and progression.
- Understanding changes in NAWM is crucial for monitoring MS.
Purpose of the Study:
- To investigate the utility of texture features in T2-weighted MRI scans for identifying and tracking changes in NAWM in subjects with a first demyelinating event.
- To assess the potential of texture analysis to differentiate between various white matter states (NAWM, developing lesions, established lesions) over time.
Main Methods:
- Utilized MRI scans from 38 untreated subjects experiencing a first demyelinating event (Clinically Isolated Syndrome).
- Manually delineated lesions at baseline (Time0) and follow-up (Time6-12 months).
- Extracted texture features from normal-appearing white matter areas (NAWM0, ROIS0) and corresponding control regions (ROIS_C0), and compared them with newly developed lesions (L0, L6-12).
- Employed Support Vector Machine (SVM) models to classify these regions based on texture features.
Main Results:
- SVM models achieved high classification accuracy, distinguishing between different white matter regions.
- Accuracies ranged from 65% (ROIS0 vs ROIS_C0) to 98% (NAWM0 vs L0).
- Texture features demonstrated significant potential in differentiating normal-appearing white matter from areas that later developed into lesions.
Conclusions:
- Texture features extracted from T2-weighted MRI scans offer valuable complementary information for monitoring MS disease development and progression.
- The proposed texture analysis method shows promise for early detection of pathological changes in NAWM.
