Md. Johir Raihan

His research interests span temporal analysis, machine learning, computer vision, and brain-computer interfaces. His published work develops interpretable deep learning models for automated rehabilitation exercise assessment, including CNN, graph neural network, and squeeze-and-excitation architectures tuned with metaheuristic optimization, alongside explainable AI for medical diagnosis and machine learning for Bengali sign language interpretation. As a Graduate Research Assistant at the Neuro-Physiological Rehab Lab, Khulna University, he developed NeuroCue, an open-source framework for synchronized EEG and multimodal data acquisition, and built machine learning models to support improved autism spectrum disorder (ASD) diagnosis. His master’s thesis proposed a spatio-temporal AI framework for Parkinson’s disease detection using physiological signals.