Multi-echo fMRI of the cortical laminae in humans at 7T

• Published in: NeuroImage (Orlando, Fla.) Vol. 56; no. 3; pp. 1276 - 1285
• Main Authors: Koopmans, Peter J., Barth, Markus, Orzada, Stephan, Norris, David G.
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier Inc 01-06-2011; Elsevier Limited
• Subjects: 7 Tesla; Cortical layers; fMRI; High resolution; Multi-echo; NMR; Noise; Nuclear magnetic resonance; Physiology; Transverse relaxation; TE; fMRI; High resolution; Transverse relaxation; DL; TL; 7Tesla; Multi-echo; Cortical layers; PVE; ML
• ISSN: 1053-8119; 1095-9572
• DOI: 10.1016/j.neuroimage.2011.02.042

Recent developments in ultra high field MRI and receiver coil technology have opened up the possibility of laminar fMRI in humans. This could offer greater insight into human brain function by elucidating both the interaction between brain regions on the basis of laminar activation patterns associated with input and output, and the interactions between laminae in a specific region.
We used very high isotropic spatial resolution (0.75mm voxel size), multi-echo acquisition (gradient-echo) in a 7T fMRI study of human primary visual cortex (V1) and novel data analysis techniques to quantitatively investigate the echo time dependence of laminar profiles, laminar activation, and physiological noise distributions over an extended region of cortex. We found T2⁎ profiles to be explicable in terms of variations in myelin content. Laminar activation profiles vary with echo time (TE): at short TE the highest signal changes are measured at the pial surface; this maximum shifts into grey matter at longer TEs. The top layers peak latest as these have the longest transverse relaxation time. Theoretical simulations and experiment suggest that the intravascular contribution to functional signal changes is significant even at long TE. Based on a temporal noise analysis we argue that the (physiological) noise contributions will ameliorate differences in sensitivity between the layers in a statistical analysis, and correlates with laminar blood volume distribution. We also show that even at this high spatial resolution the physiological noise limit to sensitivity is reached within V1, implying that cortical sub-regions can be examined with this technique.
► High-resolution (0.75mm) fMRI of cortical laminae at 7T using multi-echo. ► Large, convoluted brain areas can be analysed using FreeSurfer's surface reconstruction. ► Laminar T2⁎ profiles correlate well with cortical myelin density. ► Laminar temporal noise profiles correlate well with cortical blood density. ► Optimal TE is layer-dependent and T2⁎ of ‘active’ venous blood is not negligible.