1. Motivations
- Today there is no commercial solution for archiving clinical research images
- Phase 1: Develop a reliable and standard-compliant MRI image research archive system for all GE MRI
scanners at 221 Longwood Advanced MRI Center
- Phase 2: Develop a platform to support image re-analysis, population and time series studies,
computer-aided diagnosis, and imaging-based clinical trial management
2. Hardware
- A Linux server with high-speed RAID 5 disk array (3TB) storage
- Periodic disk-to-disk backup to another RAID unit (3TB) on the same
server (move to a physically-independent server in Phase 2)
- Windows XP application server hosting supporting applications
- Replica servers set up for redundancy and reliability
- Gigabyte network connectivity
3. Software
- Receiving application - PHP, Apache, mySQL
- fMRI/raw files storage - scp/ftp
- Custom CBI search application (prototype) - Apache,
JBoss and Oracle or other - SQL-compliance
database system.
4. Current Status
- Phase 1 completed and functioning
- The process from data acquisition to data preprocessing and archiving into DICOM archive is fully automated
- MIRA is designed to be scalable for more storage space and interactive query capability
- Phase 2 implementation will depend on the users' requirements and funding availability
For more information contact:
John Chow
jchow@bwh.harvard.edu
Send Feedback to: 
