Neuroimaging: From Classical Image Processing to Deep Learning

3 Credits.

This course covers practical neuroimaging image processing—emphasizing MRI—with a bridge to introductory AI. Students learn to convert raw scans into analyzable data; preprocess images (normalization, contrast, denoising), segment and label structures, correct bias fields, register images (rigid/affine), and carry out visualization and quality control. A brief AI module introduces convolutional models for segmentation/classification. Examples may include classical filters, thresholding and probabilistic methods, clustering/mixture models, and modern learning-based approaches. Labs in Python/MATLAB use real datasets and stress reproducibility and method selection. Suggested course background: Mathematical Methods For Engineers or equivalent course, Signals and Systems, and Probability.