New subject: Variation in volume when recon-all is run 1416 times on the same subject

28 Jun 2023


      External Email - Use Caution
Dear FreeSurfer team,
We were interested in the reproducibility of multiple runs of recon-all and
repeated recon-all 1416 times on the same subject.
We then checked the left and right brain results for the volume obtained
from the Aseg process. The right brain showed exactly the same volume,
while the left brain showed a small variation.
We thought that the variation in volume between the left and right brains
would be the same. But why is this not the case in reality?
The volume_attemps1416.csv is attached. For example, the
Left-Lateral-Ventricle in this file has variation, but the
Right-Lateral-Ventricle has no variation, as shown below.
*Left-Lateral-Ventricle*12377.8
12375.9
12375.9
12380
12376.4
...
*Right-Lateral-Ventricle*7514.7
7514.7
7514.7
7514.7
7514.7
...
1) FreeSufer vesion: 7.2.0
2) Platform: Red Hat Enterprise Linux 8.6
3) uname -a: Linux fn01sv01 4.18.0-372.36.1.el8_6.x86_64 #1 SMP Thu Nov 17
16:37:07 EST 2022 x86_64 x86_64 x86_64 GNU/Linux
4) volume_attemps1416.csv: Summarized results of 1416 aseg.stats.
─────────────────────────────────
Takehiro Jimbo, Ph.D.
TEL: 050-5308-1575
Email: info@jras.co.jp
URL: https://secure-web.cisco.com/1v12MsuoCVOvbnqcvpETpiPtEYjXBUvx9IxdD0D4EvElKEm...
─────────────────────────────────