Automatic Identification of Multiple Sclerosis

This study aims to develop an AI-driven method for rapid and accurate diagnosis of multiple sclerosis (MS) using brain MRI images. It employs feature extraction techniques and an optimized selection process to identify key markers, achieving a detection accuracy of 97.97% using the k-nearest neighbors classifier on a small dataset. Further validation on larger datasets demonstrated 92.94% accuracy with FLAIR images and 91.25% with T2-weighted images, surpassing existing methods and showing potential as a clinical tool for early MS detection and management.