Image Analysis Applications in Perinatal Diagnostics: What has Worked and What we've Learned

Carolyn Salafia
Placental Analytics

Since the Random Shapes meeting of Spring 2007, we have embarked on a series of collaborations with IPAM associated investigators involving image analysis. We will briefly discuss four specific applications:


1. Automated neutrophil extraction from tissues, a reliable means of diagnosing an important risk factor for cerebral palsy (work shared with Dr Matthew Sottile and Kristy Thomas, Master's Candidate in Computation Sciences, University of Oregon).


2. Solution of diffusion equations of single villi: method for quantitation of placental function in a digitized slide (work shared with Drs Dimitri Vvedensky and Joshua Gill, PhD candidate in Physics at Imperial College UK and Dr Denis Grebenkov, Ecole Polytechnique)


3. 3D reconstructions of placental volume from surface pictures and slices (work performed as part of RIPS 2009, mentored by Drs Xiaoqun Zhang and Luminita Vese, Department of Mathematics, UCLA)


4. Analysis of chorionic surface vasculature: manual and automated extraction methods (work shared with Drs. Michael Yampolsky and Alex Shlakhter, Department of Mathematics University of Toronto, and the REU 2009 program, mentored by Drs Todd Wittmann and Luminita Vese and Pascal Getruer, PhD candidate in Applied Mathematics, UCLA).


The discussion will show successes, and limitations of differenta approaches to these image analysis problems that have important clinical diagnostic implications.


Back to Mathematical Problems, Models and Methods in Biomedical Imaging