At the recent ISMRM meeting in Seattle, Dr. Elias Zerhouni predicted growth for the use of quantitation in medical imaging. It may seem obvious that quantitative information should be extracted from MRI brain images, but this currently isn’t being done in clinical cases because there are difficult issues to overcome, not just technical, but also political and regulatory. Computer aided detection (CAD) of structural brain changes promises to alleviate some of the barriers to the clinical application of quantitative measurements and this will allow new ways to diagnose, measure treatment response and guide interventions. We believe that for CAD to be successfully used on magnetic resonance brain scans, it not only needs to rely on automation for efficiency, but it also requires manual interaction and oversight. We present the results of a study involving 40 T1-weighted human MRI brain scans that demonstrates an increase in the efficiency of extracting quantitative measurements. The method begins with automated pre-processing to remove intensity inhomogeneities from the raw scans and then proceeds with interactive automation to generate outlines around specific neuroanatomical regions of interest. The outlines are finally converted into morphometric results (volumetric and/or shape metrics) but only after manual over-reads assure correctness and accuracy. Open source software, “NVM” provides the manual interface to automated analyses and “SegMentor” scripts glue all aspects of the data pipeline together while also assuring accuracy and accountability. We compare the results against purely manual and purely automated outputs and conclude with a discussion of the benefits of interactive automation.