Quantitative image analysis of damaged brains: study of acute stroke
Stroke is a major cause of death and disability in both the more developed and the less developed world. Recent advances in Magnetic Resonance Imaging (MRI) offer unique advantages for the evaluation of cerebral acute strokes. Higher strength of magnetic field (1.5-3.0 T field strength) yielding better resolution of images and newer sequences of images have lead to widespread use of this technology in diagnosis and management of acute stroke.
In this context, and in collaboration with Dr. Bogdan Draganski of the Laboratoire de Recherche en Neuroimagerie at the Lausanne University Hospital, we will investigate computer assisted image analysis methods that will allow an automatic segmentation of healthy brain tissues and damaged tissues due to stroke. Note that the precise segmentation of lesions is essential for the understanding of lesion-deficit mappings in human brain. MRI multimodal imaging data will be provided to take advantage of the different kinds of anatomical information provided by different imaging modalities. We will develop a robust framework for both tissue and lesion detection including tissue segmentation and a priori knowledge.
Requirements: This project will be developed in Matlab and ITK (C++ environment). Basic knowledge in signal/image processing is required.
Supervisors: Dr. Meritxell Bach Cuadra, Signal Processing Laboratory (LTS5) & Dr. Bogdan Draganski, Laboratoire de recherche en neuroimagerie (LREN), Lausanne University Hospital (CHUV).
Master Student: Rapaël Burgener, EPFL.