Medical Grid: Using Grid Technology for Brain Studies

Abstract

ABSTRACT The Medical Grid (MedGrid) Project was initiated in 2004 to study and demonstrate the use of grid technology in medical applications, in particular, in the management and analysis of functional magnetic resonance imaging (fMRI) datasets. For this, the MedGrid testbed was formed to provide a platform for the study and to demonstrate the feasibility of the idea. In this paper, we summarize the progress we’ve made in the last four years in implementing a grid-based fMRI data management and analysis framework within the MedGrid testbed. Keywords Grid computing, functional MRI, fMRI, BAXGrid, BAXSQL 1. INTRODUCTION Advances in digital medical imaging systems such as magnetic resonance imaging (MRI), computed tomography, among others, have led to an explosion in the number of medical digital images produced in hospitals worldwide. As this growing number of datasets necessitates more storage capacity and more computational resources, imaging sites are facing several new challenges in terms of the management and processing of these datasets. In the study of human brain functions using functional magnetic resonance imaging (fMRI), the demand for more storage and processing resources is also rising. With the availability of higher field strength MRI scanners, acquiring veryhigh-resolution functional MR images becomes possible. This raises the storage and computational requirement for these datasets. This is further amplified by the increasing complexity in the type of analysis performed on acquired datasets and the growing trend of fusing several imaging modalities to obtain more reliable results. Because of this, we explored the potential use of grid technology[1] to meet the storage and computational demands for brain studies. Advances in grid technology enabled the sharing of geographically distributed resources, such as supercomputers, PC clusters, ultra-high capacity storage devices, and scientific instruments. With grid computing, on-demand, collaborative, and data-intensive computing is also now feasible. With on-demand computing, imaging sites can access high-end computational resources from service providers when only needed, minimizing acquisition and maintenance cost. Collaborative computing allows the federation of several resources from different organizations, forming a virtual organization with computing and/or storage capacity much higher than individual members. Computing requirements for data-intensive applications can be shared among participating organizations to minimize computational load. To study and demonstrate the use of grid technology, the Medical Grid (MedGrid) Project