Brain Imaging Behav. 2026 Feb 9;20(1):7. doi: 10.1007/s11682-026-01092-3.
ABSTRACT
Brain analysis software uses different atlases producing different results. These variations arise mainly from differences in algorithms and region-of-interest definitions used for segmentation and parcellation. We compared the results of FreeSurfer (FS) and BrainSuite (BS), in brain segmentation and parcellation for group comparisons. Magnetic resonance images from 45 adults (23 females, 22 males) were analyzed using FS and BS. Our analysis included the volume, surface area and thickness of the total brain, right motor cortex, and right thalamus. Statistical analyses were done between software outputs and sex-based variations in brain structure. The mean volumes (cm³) of the thalamus, motor cortex, and total brain were 7.20, 28.54, and 951.9 for FS, and 6.25, 25.47, and 772.6 for BS (P = 0.001 for all). Volumes were higher with FS than BS. The mean surface areas (cm²) of the motor cortex and total brain were 59.30 and 1966.71 for FS, and 57.80 (P = 0.001 for both) and 1797.97 for BS (P = 0.001 for both). Motor cortex surface area was significantly higher in FS compared to BS, while total brain surface area was also larger in FS. The mean thickness (mm) of the motor cortex and total brain was 2.32 and 2.37 for FS, and 4.05 and 3.95 for BS (P = 0.001 for all). Thickness values were lower in FS than in BS. Sex comparison results differed for total brain surface area, but other comparisons revealed similar statistical results across sexes for both software. FS provides higher volume and surface area measurements than BS, while BS yields greater thickness measurements. Despite differences in thickness, surface area, and volume, group comparisons from FS and BS data yielded similar results.
PMID:41661408 | DOI:10.1007/s11682-026-01092-3
