Abstract #1953
Using the ADNI1 data set to assess the back-to-back reproducibility of brain volumetrics
Keith S Cover 1 , Ronald A van Schijndel 1 , Adriaan Versteeg 1 , Alberto Redolfi 2 , Jrme Revillard 3 , Baptiste Grenier 3 , David Manset 3 , Hugo Vrenken 1 , Bob W van Dijk 1 , Giovanni B Frisoni 2 , and Frederik Barkhof 1
1
VU University Medical Center, Amsterdam,
Netherlands,
2
IRCCS
San Giovanni di Dio Fatebenefratelli, Italy,
3
gnbila
France, France
To compare the reproducibility of the hippocampal
atrophy rates over one year generated by different
versions of the FreeSurfer/ReconAll and FSL/FIRST
software packages using a back-to-back (BTB)
reproducibility test. The test is based on the first
Alzheimer's Disease Neuroimaging Initiative (ADNI1) data
set - which includes two MPRAGE scans for each patient
visit. The FSL/FIRST and FreeSurfer/ReconAll
reproducibilities of hippocampal atrophy rates were
similar. The poor reproducibility of the well documented
bug in longitudinal mode of FreeSurfer 5.0.0 is clearly
evident. The ADNI1-based BTB benchmark presented is a
valuable way to compare performance of volumetric
algorithms.
This abstract and the presentation materials are available to members only;
a login is required.
Join Here
