TY - JOUR
T1 - Prospective slice-by-slice motion correction reduces false positive activations in fMRI with task-correlated motion
AU - Schulz, J.
AU - Siegert, T.
AU - Bazin, P. L.
AU - Maclaren, J.
AU - Herbst, M.
AU - Zaitsev, M.
AU - Turner, R.
N1 - Funding Information:
The authors acknowledge the financial support of the Max Planck Society and thank Andreas Schäfer and Robert Trampel for their scientific support, as well as Domenica Wilfling and Elisabeth Wladimirow for their assistance with the volunteer measurements.
PY - 2014/1/1
Y1 - 2014/1/1
N2 - Objective: We aimed to test the hypothesis that slice-by-slice prospective motion correction at 7. T using an optical tracking system reduces the rate of false positive activations in an fMRI group study with a paradigm that involves task-correlated motion. Materials and methods: Brain activation during right leg movement was measured using a block design on 15 volunteers, with and without prospective motion correction. Clearly erroneous activations were compared between both cases, at the individual level. Additionally, conventional group analysis was performed. Results: The number of falsely activated voxels with T-values higher than 5 was reduced by 48% using prospective motion correction alone, without additional retrospective realignment. In the group analysis, the statistical power was increased - the peak T-value was 26% greater, and the number of voxels in the cluster representing the right leg was increased by a factor of 9.3. Conclusion: Slice-by-slice prospective motion correction in fMRI studies with task-correlated motion can substantially reduce false positive activations and increase statistical power.
AB - Objective: We aimed to test the hypothesis that slice-by-slice prospective motion correction at 7. T using an optical tracking system reduces the rate of false positive activations in an fMRI group study with a paradigm that involves task-correlated motion. Materials and methods: Brain activation during right leg movement was measured using a block design on 15 volunteers, with and without prospective motion correction. Clearly erroneous activations were compared between both cases, at the individual level. Additionally, conventional group analysis was performed. Results: The number of falsely activated voxels with T-values higher than 5 was reduced by 48% using prospective motion correction alone, without additional retrospective realignment. In the group analysis, the statistical power was increased - the peak T-value was 26% greater, and the number of voxels in the cluster representing the right leg was increased by a factor of 9.3. Conclusion: Slice-by-slice prospective motion correction in fMRI studies with task-correlated motion can substantially reduce false positive activations and increase statistical power.
KW - FMRI
KW - False positive activations
KW - Magnetic resonance imaging
KW - Optical motion tracking
KW - Prospective motion correction
UR - http://www.scopus.com/inward/record.url?scp=84883666577&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84883666577&partnerID=8YFLogxK
U2 - 10.1016/j.neuroimage.2013.08.006
DO - 10.1016/j.neuroimage.2013.08.006
M3 - Article
C2 - 23954484
AN - SCOPUS:84883666577
SN - 1053-8119
VL - 84
SP - 124
EP - 132
JO - NeuroImage
JF - NeuroImage
ER -