[Development of a computer-aided diagnostic system for detecting multiple sclerosis using magnetic resonance images]

Area of Science:

• Neuroimaging
• Medical image analysis
• Machine learning for healthcare

Context:

• Multiple Sclerosis (MS) lesion evaluation is crucial for tracking disease progression.
• Accurate assessment of lesion location, shape, and area is vital.
• Current methods may lack efficiency in temporal change evaluation.

Purpose:

• To develop an automated method for detecting multiple sclerosis (MS) lesions.
• Utilize T1-weighted, T2-weighted, and FLAIR MRI sequences.
• Employ tri-linear interpolation, k-mean clustering, and Support Vector Machine (SVM) for classification.

Summary:

• An automated method was developed using T1, T2, and FLAIR MRI.
• Brain regions were segmented via tri-linear interpolation and k-mean clustering.
• A Support Vector Machine (SVM) classified true and false positive MS regions, achieving 81.8% sensitivity and 14.1% false positives.

Impact:

• The automated method shows promise for improving the diagnosis of multiple sclerosis.
• Facilitates more objective and efficient monitoring of MS lesion changes.
• Potential to aid clinicians in estimating disease progression more accurately.