Post-doctoral position in Medical Imaging

Post-doctoral position in Medical Imaging

Post-doctoral position in Medical Imaging

Ecole Centrale de Paris/INRIA Saclay, Ile-de-France

 

 

Topic: Multi-Modal Registration of Medical Images

 

Research Context:

Registration between pre-operative and Interventional data is a major  challenge and demand in a number of surgical operations In such a  context, first the operation procedure and the target are determined  through the use of high resolution appropriate imaging modalities. This  information is then used as guide for surgeons in the operation room. To  this end an alignment between the pre-operative data and the acquisition  on the fly during the surgery is required. The use of global parametric  transformations assumes that the tissue remains static that is not the  case when  one has to account for tissue shift because of the cut and an  alignment that also involves a local deformation component is to be  considered.

 

Post-doctoral Research work Description:

The aim of this post-doc is to address registration between  pre-operative 3D annotated data and  2D interventional images during  surgery. In such a context, the use of discrete optimization will be  considered, where one seeks for an out of plane projection of the 2D  images to the 3D volume and an in-plane non-rigid deformation towards  addressing tissue shift. Efficient techniques from discrete optimization  will be considered to address the problem. The estimation will  decomposed into a master (plane selection) and a slave problem (in plane  deformation). In order to account for different modalities, a  kernel-based approach will be considered. Once such a metric has been  determined, the registration will be casted using a continuous  interpolation model over non-uniform control points. This is a joint  work with one of the leading university hospitals in Paris. Successful  candidates are expected to contributer with other ongoing project in the  field of medical image analysis and computer aided diagnosis.

 

Required knowledge and background:

PhD in electrical or computer engineering, or computer science or  applied mathematics in the areas of medical imaging or computer vision  demonstrating ability to carry on research of excellent quality. Prior  work on registration, motion analysis or discrete optimization will be a  plus while good programming skills in Matlab and C++ and ability to work  in a team will be appreciated.

 

Duration : 12-24 months, Gross salary/month : 2500-3000€

Contact: For more information and to apply please contact Prof. Nikos

Paragios (nikos.paragios@ecp.fr)  with a resume and a list of two

potential references.

 

-- Nikos Paragios http://vision.mas.ecp.fr