Cookies on this website
We use cookies to ensure that we give you the best experience on our website. If you click 'Continue' we'll assume that you are happy to receive all cookies and you won't see this message again. Click 'Find out more' for information on how to change your cookie settings.


Moises Hernandez Fernandez

Accelerating computational diffusion MRI using GPUs

DPhil Candidate

Research Summary

Research summary

My research focuses on high performance computing applied to the analysis of diffusion MRI (dMRI) data. I am interested in how  parallel computer architectures, such as Graphics Processing Units (GPUs), can be used in scientific applications that require very high computational resources. 

I am using GPUs for resolving tissue microstructural patterns and for estimating long-range brain connectivity. The immense computational power provided by modern GPUs is exploited and accelerations of up to two orders of magnitude are obtained when comparing single GPU with single-threaded CPU implementations. Coronal, sagittal and axial views comparing FSL’s CPU tractography tool with GPU tractography framework reconstructing some major brain white matter tracts. These paths are binarised versions of the path distributions, after these have been thresholded at 0.5%.

My DPhil is supervised by Dr. Stamatios Sotiropoulos and Prof. Stephen Smith from FMRIB Analysis Group and by Dr. Istvan Reguly and Prof. Mike Giles from Oxford e-Research Centre.

Path probability map from a vertex in the Motor Cortex. The map is extracted from a dense connectome matrix (path probabilities between grayordinates, when these are used as seeds) generated with the FSL’s CPU tractography tool and our GPU tractography framework. Connectome Workbench tool ( was used for visualisation.

False True