IRMA is a meta-algorithm for image registration (image alignment), evaluating results under multiple metrics using the LONI Pipeline workflow infrastructure,
on the LONI/CCB grid computing facility. IRMA manages these results in a model base implemented with PostgreSQL. It permits
scientists to catalog the results such as provenance information, and permits subsequent mining — exploring the space of alternatives
in an organized fashion and building understanding about individual algorithms, and learn about strengths and weaknesses of
algorithms over time.
This work supported by NIH grant 1U54RR021813 and the Center for Computational Biology, an NIH National Center of Biomedical
Computing.