CSE 554: Geometric Computing for Biomedicine

CSE554 Introduction Slide 1

CSE 554: Geometric Computing for Biomedicine

CSE 554: Geometric Computing for Biomedicine

Fall 2015

http://www.seas.wustl.edu/


OutlineOutline

• Introduction to course

• Mechanics



OutlineOutline


• Mechanics



GeometryGeometry

• Greek word: Earth-measuring

• One of the oldest sciences

Chinese Chou Pei Suan Ching (500-200 BC) Euclid’s Element (300 BC)



GeometryGeometry

• Greek word: Earth-measuring

• One of the oldest sciences

Newton’s Principia Mathematica (1687) Einstein’s General Relativity (1915)



Geometric FormsGeometric Forms

• Continuous geometry

– Defined by mathematical functions

– E.g.: parabolas, splines, subdivision surfaces

• Discrete geometry

– Disjoint elements with connectivity relations

– E.g.: polylines, triangle surfaces, pixels and voxels

2xy ][][ ySinxSinz

Pixels

Triangle surfaces (meshes)

Polyline

Voxels

Curves Surfaces



Geometric ComputingGeometric Computing

• Algorithms and data structures for (discrete) geometry

– Creation

• From 2D/3D images, from point clouds, by hand, etc.

– Processing

• De-noise, simplify, repair, transform, animate, etc.

– Analysis

• Geometric, topological, shape and physical properties



ApplicationsApplications

Industrial design

Movie CGUrban design and evacuation planning

Engineering simulation3D printing



Application: BiomedicineApplication: Biomedicine

• Modeling biological structures as geometric forms

– A spectrum of scales: organs, tissues, cells, molecules, etc.

• With geometric representation, we can do

– Visualization

– Quantitative analysis

– Simulation and interaction

Human Virus

Surgical simulation

Treatment planning



This CourseThis Course

• Classical algorithms for geometric computing

– Widely used for biomedical image analysis

– Easy to understand, simple to implement




• Working with biomedical imaging data

– 2D: Light microscopy, slices of 3D images

– 3D: Magnetic resonance imaging (MRI), Computed tomography (CT), Cryo-Electron Microscopy (Cryo-EM)

CT Cryo-EMMicroscopy




• Creating, processing, deforming, and analyzing geometry

Segment Extract boundary Fair & Simplify

Align & DeformShape analysis

(Before) (After)



Beyond This CourseBeyond This Course

• On-going research projects on biomedical modeling

– Gorgon: shape analysis of proteins (Gorgon.wustl.edu)

– Geneatlas: image-based queries in mouse brains (Geneatlas.org)

– VolumeViewer: interactive 3D segmentation (Volumeviewer.cse.wustl.edu)

• Research opportunities in the M&M lab

– Biomedical modeling (Tao)

– Image analysis (Robert, Tao)

– Computer vision (Robert, Yasu)

– Human computer interaction (Caitlin)

http://gorgon.wustl.edu/

http://geneatlas.org/

http://volumeviewer.cse.wustl.edu/



OutlineOutline


• Mechanics



StaffStaff

• Instructor: Tao Ju

– Jolley 406 ([email protected])

• TA:

– Yajie Yan ([email protected])

– Zhiyang Huang ([email protected])

mailto:[email protected]





PrerequisitesPrerequisites

• Programming

– Experienced in at least one of the major programming languages

• C/C++, Java, Matlab, Python, etc.

– CSE332 is strongly recommended (required if you are a CS major)

• CS background

– Basic data structures (e.g., queues, trees, hash tables) and algorithms

– CSE241 is strongly recommended (required if you are a CS major)

• Math

– Linear algebra



OverviewOverview

• 2 meetings per week

– Lectures on Mondays (Cupples II 200)

– Labs on Wednesdays (Whitaker 130)

• 6 lab modules

– 2 weeks for each module

– Due and graded in Wednesdays labs

• 1 course project

– Proposal due in November

– Final presentation in December

• Check out the calendar on course webpage

No exams!

http://www.cse.wustl.edu/~taoju/cse554/



LecturesLectures

• Theory and algorithms

– Algorithms are explained in depth, pseudo-code given when possible

1. …

2. Repeat until Q is empty:

1. Pop a pixel x from Q.

2. For each unvisited object pixel y connected to x, add y to S, set its flag to be visited, and push y to Q.

3. Output S

Example:



Lab ModulesLab Modules

• Algorithm prototyping (in Mathematica)

– Step-by-step, easy to hard, 2D to 3D

– Unit tests

– Work individually

Example:



Course ProjectCourse Project

• A working tool that solves an existing problem in biomedical research

– Topics provided by the instructor or identified by students

• Use your favorite programming language

• Work in team or individually




• Example projects (Fall 2014)

– Measuring length of sperm cells of fruit flyies

(Luis Velazquez-Irizarry)





– Plotting concavity of bone surface

(Zhaonan Liu and Zhenyi Zhao)





– Segmenting skull from MRI scan

(Hang Yan)





– Measuring size of holes on skulls in CT scans

(Zhiyang Huang)





– Matching and superimposing ancient prints

(Tom Wilkinson)



GradingGrading

• Lab modules: 75% (graded during Wednesday labs)

• Course project: 25%

• Late policy

– Late modules are accepted till the Monday following the due date

– The late part will earn at most 50% credit

– Other extensions will be given only under exceptional conditions.



Action Items – This WeekAction Items – This Week

• Make sure you have a SEAS account

– Check with the help desk at EIT in Lopata 4nd floor.

• Get access to Mathematica

– Available on all SEAS machines; installed freely on campus computers

– Purchase for personal use for $45 / semester

• Module 0 is already out

– Due and graded two weeks from this Wednesday in lab (Sept. 9)

– I will give a quick tutorial this Wednesday

• See you all on Wednesday (Whitaker 130)!


Documents

CSE 554: Geometric Computing for Biomedicine