Influence of Brain MRI Segmentation Techniques on EEG Source Analysis

This study investigates influence of brain MR segmentation techniques and implements a framework in Python for a uniform interface to the three different automated segmentation software tools (SPM,FSL and FreeSurfer) in order to be able to compare them consistently. This is done by using the Nipype (Neuroimaging in Python, Pipelines and Interfaces) which provides a uniform interface to these neuroimaging software tools. In addition, the tissue probability maps, masks and surfaces that are required for the head model construction are generated with this framework. The source reconstruction results with the head models constructed with different segmentation tools are evaluated based on the distance error from the source location to the lesion in the post-operative MR images. The SPM head models produced lowest distances error compared to the FSL and the FreeSurfer head models for two patients.

Autorentext

Azeez Adebimpe graduated with first class double honour in Electrical and Computer Engineering from FUT Minna, Nigeria. He worked briefly with Nigeria Atomic Energy Commission. He is presently now in Europe for postgraduate study and research in Biomedical Engineering with interest in system neuroscience and computational modeling.