Authorea

        Does the method here proposed offer a more cost-efficient approach relative to other studies? The claim that this work proposes a more efficient multi-centric calibration approach should be reconsidered in the light of past work. It is true that the ADNI method involved considerable effort to install on different vendors particular sequences, which was possible by having agreements with the various vendors. However, simpler approaches have recently shown that high multi-centric reliability can be also achieved by taking the standard sequences available on various makes and models and limit the calibration to the setting of a few basic key acquisition parameters that optimize tissue contrast (Jovicich et al., 2013). This approach did not require software or hardware adjustments, did not use geometric calibration phantoms and did not use a group of volunteers travelling for scans at multiple sites distributed across countries. I highly doubt that sending 12 people across the USA and Europe to be scanned on 20 MRI sites can be considered more efficient and less costly than sending no person at all, or maybe one to check the local protocol after the setting of a few acquisition parameters. Or maybe the authors meant something else, please clarify.

    