Welcome!

This is the GitHub repository for the course:

ECE 594N: Geometric Machine Learning for Biomedical Imaging and Shape Analysis at UC Santa Barbara.

• Instructor: Prof. Nina Miolane, BioShape Lab, UC Santa Barbara.
• Lectures: Mondays, Wednesdays 12:00 - 1:50 PM in PHELP 1437

• Join ECE 594n Slack workspace with your @ucsb.edu email address through the invitation you have received via email.

• Make it a habit to check ECE 594n Slack several times a week.

• Slack is our preferred way of communicating. Avoid emails and use Slack to ask questions about syllabus, lectures, project.

Advances in biomedical imaging techniques have enabled us to access the shapes of a variety of structures: organs, cells, proteins. Since biological shapes are related to physiological functions, shape data may hold the key to unlock outstanding mysteries in biomedicine — specifically when combined with genomics and transcriptomics data.

Machine learning is poised to play a major role in analyzing this new wealth of imaging information and testing novel biomedical hypotheses. In this class you will learn how to perform geometric machine learning on biomedical images and shapes. The course will cover basics of geometric machine learning and delve into specific methods for shape analysis of proteins, cells and organs. This course will feature guest lectures from invited speakers.

• Unit 1 (Geometry - Math!): Differential Geometry for Engineers

  • A) Manifolds
  • B) Lie groups
  • C) Riemannian metrics

• Unit 2 (Shapes): Computational Representations of Biomedical Shapes

  • A) Shapes of landmarks
  • B) Shapes of curves
  • C) Shapes of surfaces
  • D) Deformations

• Unit 3 (Machine Learning): Geometric Machine Learning for Shape Analysis

  • A) Statistics on manifolds
  • B) Classification
  • C) Regression
  • D) Clustering
  • E) Dimension reduction

The lecture notes are available in the lectures folder.

• Monday 01/16/2023 (11:59 PM PST): Intermediate Python - Datacamp. <4 hours.
• Monday 01/30/2023 (11:59 PM PST): Geomviz HW.
• Monday 02/20/2023 (11:59 PM PST): Shape HW.
• Monday 03/13/2023 and Wednesday 03/15/2023: Presentations of final projects.