The A.A. Martinos Center for Biomedical Imaging (formerly known as the MGH-NMR Center), has an acute and growing need for an archival system for storage and retrieval of imaging data. The Center's four (4) active human research MRI scanners and three (3) animal systems are used by over 100 different users, each of whom is responsible for archiving their own data. Most of these users have backgrounds that are not in the radiological sciences, and are therefore not accustomed to the detailed requirements for data archiving. Because of this, despite extensive training and efforts, data loss is all too common, perhaps 1% to as much as 3% among some investigators. A secure, redundant archiving system coupled with a radiology information system (RIS) could prevent such loss. In addition to this need, a number of currently funded projects would scientifically benefit from a cataloged archive that allows, under appropriate human studies guidelines, retrospective analyses. For example, a neuroscientist using fMRI to do brain mapping may learn of a new analysis technique and wish to re-analyze two year's worth of data using the new tool. A central archive makes the task of retrieving data to test such a hypothesis practical. Furthermore, technical analyses needed to undertake multicenter clinical studies would be facilitated by such an archive/RIS combination. Such analyses include intra-study comparison across time points to ensure scanner stability, measurements of technical scanner parameters such as the effect of time of day on scanner noise, effect of new software or hardware upgrades on image quality. The RIS system we propose could also store scanned-in consent forms and other data at the time of scanning for central archiving and retrieval, all with sophisticated access controls. Such a tool would also be highly beneficially in assisting investigators in maintaining full HIPAA compliance, especially as interpretations of such privacy requirements are modified over time. For example, certain volumetric MRI scans are of high enough quality that a subject's face is recognizable, a potential violation of privacy rules. A central archive would provide a common platform for us to automate needed privacy tools/protocols, thereby allowing our Center to offer fully HIPAA-compliant services to a user community that wishes to focus on science, not on the details of image storage. Overall, this needed resource will greatly facilitate many NIH-funded investigators and investigations for years to come.
Showing the most recent 10 out of 12 publications
