An Accurate Multiple Sclerosis Detection Model Based on Exemplar Multiple Parameters Local Phase Quantization: ExMPLPQ

Multiple sclerosis (MS) is a chronic demyelinating condition characterized by plaques in the white matter of central nervous system that can be detected using magnetic resonance imaging (MRI). Many deep learning models for automated MS detection based on MRI have been presented literature. We developed computationally lightweight machine model diagnosis novel handcrafted feature engineering approach. The study dataset comprised axial and sagittal brain images were prospectively acquired from 72 59 healthy subjects who attended Ozal University Medical Faculty 2021. was divided into three subsets: only (n = 1652), 1775), combined 3427) both classes. All resized to 224 × 224. Subsequently, features generated with fixed-size patch-based (exemplar) extraction local phase quantization (LPQ) three-parameter settings. resulting exemplar multiple parameters LPQ (ExMPLPQ) concatenated form large final vector. top discriminative selected iterative neighborhood component analysis (INCA). Finally, k-nearest neighbor (kNN) algorithm, Fine kNN, deployed perform binary classification vs. ExMPLPQ-based attained 98.37%, 97.75%, 98.22% accuracy rates axial, sagittal, hybrid datasets, respectively, kNN 10-fold cross-validation. Furthermore, our outperformed 19 established pre-trained trained tested same data. Unlike models, yet highly accurate. It has potential implemented as an diagnostic tool screen MRIs lesions suspected patients.