External Email - Use Caution
Hello,
I had the problem already with FS 7.2.0, but since it also occurs in FS 8.1.0, which I currently have installed on UBUNTU 20.24, an answer would be helpful. I am fully aware that FS and Synthseg do not provide exactly the same results when calculating the volumes for the same T1 scan with FS and Synthseg. E.g when I calculate a T1 3D 1mm ISO scan and a FLAIR 3D 2mm ISO scan of the same person both with Synthseg, I also fully understand that the results will not be exactly the same here either. However, what I do not understand is that the results for sub-volumes sometimes deviate by a factor of 1000? I have attached a small table below: row 1 Synthseg T1 scan, row 2 Synthseg FLAIR scan, row 3 FS T1 scan same subject. E.g. for "left lateral ventricle," the Synthseg FLAIR and FS results are about the same, but the Synthseg T1 does not match by a factor of 1000. For "left inferior lateral ventricle," the Synthseg FLAIR result is too large. Where do these deviations come from? Is it a unit problem??? I do not understand what is behind it. These deviations occur in all individuals in this manner.
Thank you for your help, Jürgen
subject
total intracranial
left cerebral white matter
left cerebral cortex
left lateral ventricle
left inferior lateral ventricle
left cerebellum white matter
left cerebellum cortex
left thalamus
left caudate
left putamen
left pallidum
3rd ventricle
4th ventricle
brain-stem
left hippocampus
left amygdala
csf
left accumbens area
Syn= 02_t1_5_13
1581149.5
234530.16
279514.53
28584312
515.05
15703237
62220.02
5994828
3766241
4605884
1485176
1864659
2365228
23305277
3743831
1692543
258776.8
679.45
Syn= 02_flair_5_13
1540067.8
237580.17
278995.16
26533.9
462551
14112386
64456.68
6460912
3999.81
4585682
1717103
1535.82
1356496
26397467
3787048
1596163
211639.08
720396
fs= 02_t1.5.3
1558777.000000
212518.500000
248086.368363
30058.0
1061.2
16326.6
57684.7
5877.0
3879.4
4342.9
1751.2
2113.8
2982.3
20526.8
4015.2
1763.8
374287.9
539.0
ok
ok
ok
?
?
?
ok
?
?
?
?
?
?
?
?
?
ok
?