COMP 775 - Fall 2017

COMP/BMME 775: Medical Image Analysis

Fall 2017, T TH 9:30am-10:45am, FBF 007

Instructor: Marc Niethammer
Email: (mn -at- cs.unc.edu)
Phone: 919-590-6149
Office: 219 Sitterson Hall

Overview

This course will provide an introduction to image processing and analysis motivated by medical image analysis. Starting with an overview of imaging modalities, we will explore simple methods for image preprocessing and feature extraction, methods for general image segmentation, for image analysis, as well as for image registration. We will also briefly touch on more modern approaches to image analysis making use of machine learning, in particular, deep neural networks.

Prerequisites: Some knowledge of linear algebra and calculus. MATH 547, STOR 435. Ability to program.

Resources: A textbook is not explicitly required. However, I will follow loosely material in the book 'Image Processing, Analysis, and Machine Vision' by Sonka, Hlavac, Boyle, Third Edition. A number of related books are are listed on the syllabus.

Lecture materials including a detailed syllabus will be posted on Sakai

Grading: Grades will be based on a set of homework problems and one final exam. The split-up is as follows:

  • Homework: 60%
  • Final exam: 40%

Tentative Schedule (subject to change)

Date Topic
Tue Aug. 22 Medical Imaging Modalities
Thu Aug. 24 Medical Imaging Modalities (continued)
Tue Aug. 29 Simple feature extraction and preprocessing
Thu Aug 31 Simple feature extraction and preprocessing (continued)
Tue Sep. 5 Binary images: morphology, skeletonization
Thu Sep. 7 Simple Segmentation: Thresholding and edge-based methods
Tue Sep. 12 Simple Segmentation: Merging, splitting, watershed, and matching
Thu Sep. 14 Bayesian estimation: ML, MAP ...
Tue Sep. 19 Bayesian estimation, ML, MAP, ... (continued)
Thu Sep. 21 Energy minimization for segmentation and registration
Tue Sep. 26 Data mismatch
Thu Sep. 28 Data mismatch (continued)
Tue Oct. 3 Geometric atypicality
Thu Oct. 5 Geometric atypicality (continued)
Tue Oct. 10 Shape Spaces
Thu Oct. 12 No class due to university day
Tue Oct. 17 Brief Overview of Deterministic Optimization Methods
Thu Oct. 19 No class due to Fall break
Tue Oct. 24 Calculus of Variations
Thu Oct. 26 Calculus of Variations
Tue Oct. 31 Deformable Model Segmentation Approaches
Thu Nov. 2 Deformable Model Segmentation Approaches (continued)
Tue Nov. 7 Image Registration
Thu Nov. 9 Image Registration (continued)
Tue Nov. 14 Deep Learning
Thu Nov 16 Deep Learning (continued)
Tue Nov. 21 Deep Learning (continued)
Thu Nov. 23 No class due to Thanksgiving
Tue Nov 28 Statistical shape analysis
Thu Nov 30 Statistical shape analysis (continued)
Thu Dec. 5 Hypothesis testing