Wei Wu - weiwu2@cs.cmu.edu
Computational Biology Department, School of Computer Science, Carnegie Mellon University
Office hours: Wednesdays, 1-2 pm EST via Zoom (see Canvas for link)
Ju-Chun Huang: juchunh@andrew.cmu.edu
Office hours: Mondays, 2-3pm, GHC 7404
Raehash Shah: raehashs@andrew.cmu.edu
Office hours: Tuesdays, 4-5pm, GHC 7404
Wenduo Cheng: wenduoc@andrew.cmu.edu
Office hours: Fridays, 2-3pm, GHC 7404
Tu,Th 11:00am-12:20 pm, WEH 5403
Canvas and Piazza and this website
---------------------------------------------------
CMU 02-718 (12 units) & CMU 02-518 (12 units)
This course examines computational methods that enhance our ability to diagnose, treat, and understand human diseases. Topics will include techniques for learning models from clinical data types, including: genomics, transcriptomics, proteomics, metabolomics, imaging, and electronic medical records. Most of the techniques will involve Machine Learning. The course is organized into modules. The first module will be an introduction to the field of Medicine, and how it differs from the basic sciences, as well as personalized and precision medicine. Subsequent modules will focus specific clinical tasks, including: disease phenotyping, biomarker discovery, predictive modeling, and the design and optimization of medical interventions. Students will be assessed based on homeworks, quizzes, and a course project. Students are allowed to work in small teams (2-3 students) for the project. Students are not allowed to work in teams for homeworks and quizzes.
Pre/Co-requisites
The course is designed for graduate and upper-level undergraduate students from a wide variety of backgrounds. Students should have some background in Machine Learning, but no prior background in Medicine is required. Students must also understand and agree to comply with Carnegie Mellon University's policies on academic integrity (see also here).
--------------
Introduction to medicine (2 weeks)
Branches of medicine
Disease categories
The hierarchy of medical evidence (meta reviews; randomized clinical trials, etc)
Medical Education
The Healthcare Industry
Clinical data types
Computational challenges (10 weeks)
Phenotyping
Biomarker discovery
Predictive Modeling
Medical imaging analysis
Optimizing and/or selecting interventions
Advanced topics (2 weeks)
There is no required textbook.
Students who complete the course successfully will be able to:
Explain the core computational challenges in medicine
Discuss and critique computational methods used to address these challenges
Explain common clinical data types, and what they are used for
Use software to analyze clinical data and build predictive models for diseases
Take-home quizzes (40% of your final grade)
Five quizzes will be given online via canvas. Students must complete the quizzes alone.
See Canvas for due dates.
Homeworks (20% of your final grade)
Five homeworks will be given online via canvas. Students must complete the homeworks alone.
See Canvas for due dates.
Course Project (40% of your final grade)
Students are allowed to work in small teams (2-3 students), or individually.
Students/Teams will submit a project proposal mid-semester (see Canvas for due date).
Students/Teams will submit a final report detailing their project and give an oral presentation.