Purpose

Comparability study of automated quantitative brain volumetry at 1.5T, 3T and 7T

Materials & Methods

In this study, brain scans of 7 volunteers (255y) were acquired on three different Siemens MR Scanners (1.5T-AERA, 3T-PRISMA and 7T-MAGNETOM) using 3D-T1w images (resolution 0.5×0.5×0.59mm³, no interpolation). Volumetric measurements were performed with the AI-powered commercial software mdbrain from mediaire as it showed high stability for repeated measurements in a previous study and its short calculation time. The evaluation pipeline includes a bias correction accounting for intensity non-uniformities. For statistical evaluation, repeated measures ANOVA was calculated followed by a post-hoc paired t-tests between the three conditions focusing on a selection of seven brain regions (grey- and white matter (GM, WM), hippocampus, putamen, amygdala, caudate and thalamus)

Results

As rated by a radiologist, images were of good quality for all three scanners. Except for the hippocampus, the repeated measures ANOVA revealed significant differences between the conditions 1.5T, 3T and 7T (p<0.05). Calculated values for the post-hoc paired t-tests showed significant differences for the 1.5T/3T data versus the 7T data: Volumes of WM and amygdala were decreased bv 10% and 5% while caudate was increased by ~4%. Except for the thalamus (~6% decrease in volume), no significant differences between 1.5T and 3T were present.

Discussion

To our knowledge, this is the first study where quantitative brain volumetry at 3 different field strengths was performed. Results showed that quantitative brain volumetry at different field strengths is possible. However, comparability cannot be guaranteed which has to be taken into account when longitudinal volumetry measurements shall be performed. This is especially true for measurements at higher fields.