PhD position in Computational Statistics for Medical Imaging at DIKU, University of Copenhagen

Most data analysis methods assume that data is represented by a fixed size vector or template. However, human anatomy does not follow a template. Organs differ both in their arrangement and in whether or not they exist, and the anatomy of patients suffering from disease can be very different from that of healthy individuals.
The PhD candidate will develop statistical and computational models for the analysis of medical imaging data with variable topological structure, and will apply these methods to the analysis of retinal vessel trees, airway trees from lungs and models of local brain connectivity estimated from diffusion MR images. The candidate will utilize the geometry of stratified spaces to model topological differences in the different applications. Based on these models, the candidate will develop principled methodological and numerical tools for statistical analysis of tree-structured data and local diffusion models.
The candidate must have:
A burning desire to submerse her/himself in mathematical, statistical and computational problems that appear when trying to solve medical imaging problems.
No fear of mathematics
A MSc degree in Computer Science, Mathematics/Statistics, Engineering or Physics, or alternatively a BSc degree plus one year of MSc studies in Computer Science, Mathematics/Statistics, Engineering or Physics.
A documented solid background in either geometry or statistics
The PhD student will join the Image Section at the Department of Computer Science, and will be part of the Centre for Stochastic Geometry and Advanced Bioimaging (CSGB) http://csgb.dk. The principal supervisor will be Aasa Feragen, Department of Computer Science aasa-AT-di.ku.dk.
For details, see:
http://www.diku.dk/ominstituttet/ledige_stillinger...