ME.210 Biomedical Engineering

This course introduces modern artificial intelligence (AI) methods in the context of neuroscience and neuroengineering. Topics include machine learning algorithms, neural network models, probabilistic models, evolutionary computation, and symbolic AI. Examples span from basic research to practical applications, with an emphasis on understanding the fundamental principles underlying these approaches.

Recent advances in machine learning, and deep convolutional neural networks in particular, together with modern GPU computing and greater data availability have accelerated the adoption of deep learning in medical imaging. This course introduces the foundations of deep learning, with applications ranging from image formation to analysis, through hands-on assignments and projects in: image denoising and artifact correction; image processing, segmentation, object detection, and classification; single- and multi-modality registration; generative modeling; and sequential learning across major medical imaging modalities. Recommended course background: Python and Linear Algebra

This course provides a rigorous introduction to the foundations of biomedical data science, emphasizing principled statistical modeling and machine learning with applications to biomedical and clinical data. Lectures focus on the theoretical and mathematical foundations underlying modern data science methods, while students develop practical proficiency through applied analysis of real biomedical datasets in Python. Topics include data wrangling, exploratory data analysis, and visualization; linear and regularized regression (feature weighting and selection, including LASSO); generative and discriminative classification (LDA/QDA, logistic regression); supervised learning (perceptron, support vector machines); non-linear methods (k-nearest neighbors, decision trees, random forests); and evaluation methodology (train/test splits, cross-validation, precision/recall and related metrics). The course also introduces dimensionality reduction and representation learning (PCA, t-SNE, UMAP), EM-based methods, and accessible deep learning fundamentals A dedicated topic on ethics, fairness, and responsible use of biomedical data is covered. Background in statistics and probability, linear algebra, and Python programming is useful.

Special Studies in Biomedical Engineering

N/A