Detecting and diagnosing epileptic seizures are still difficult problems, despite significant improvements in (i) MagnetoEncephaloGraphy (MEG), a neuro-imaging technique where scanners measure brain activity, and (ii) Magnetic Resonance Imaging (MRI), a medical imaging technique for visualizing the brain anatomy.
To obtain a brain map and detect what neurons were affected by an epileptic seizure, it is necessary to merge MEG and MRI data. This merger is an iterative operation that converges, after a number of iterations, to an approximate transfer function between the input (MEG data) and the output (MRI data). This operation consumes many CPU resources and processes a very large amount of data.
The objective of this project is to solve this problem using a large-scale distributed system and Grid computing techniques.