Medical research in neurodegenerative diseases, both academic and industrial (pharmaceutical), greatly benefits from the results of MRI and PET data processing. Doctor-Researchers are not interested in computing – they just want someone to turn the raw image datasets into feature-rich analysis datasets as quickly as possible. There are very few “Technician-Researchers” to do this job and there are many different ways to process datasets, so a friendly interface is needed. A key factor in accelerating research is turnaround time. Clusters, Grids and Clouds can help here, each with their pluses and minuses.

The PIC port d’informació científica (PIC), has developed a web-based platform running over a cluster that allows researchers to easily store, organize and process data with a vast improvement of turnaround time (from several months to a few days), the so called PICNIC (PIC Neuroimaging Centre).

The PICNICC flagship aims to implement a commercial cloud back-end for PICNIC.

What we want to achieve.

• Improve the speed and quality of research for finding surrogate biomarkers based on brain images.
• Early detection of neurodegenerative diseases
• Study of pharmaceutical effects

Expected impact & benefits

Fast turnaround time in processing large amount of MRIs or PETs without relying on the “charity” of an LHC Tier-1 cluster
Execution of algorithms over several OS or processing units (CPU or GPU) and storage of large amounts of data without buying the hardware or hiring many system administrators
Maintain a user-friendly interface to submit jobs to the Cloud without specific knowledge of scripting, OS specificities, etc.
PICNIC on Cloud would make it easy to move to a full-cost, commercial charging model, which would benefit the expansion of the use of PICNIC by private industry (mainly pharmaceutical).