LAMINAR SPECIFICITY OF BSSFP AND VASO FMRI AT 9.4T IN DENSELY SAMPLED SUBJECTS

Submillimeter fMRI enables the investigation of laminar organization of cortical activity and functional connectivity. However, the commonly used Gradient-Echo BOLD (GE-BOLD) contrast suffers from reduced spatial specificity due to its sensitivity to larger veins, which are located on the cortical surface leading to superficial biases. Alternative contrasts like vascular space occupancy (VASO) and balanced steady-state free precession (bSSFP) have been proposed to mitigate vascular biases; however, their relative suitability for task-based and resting-state laminar fMRI remains incompletely characterised.

We present a preliminary comparison of VASO and bSSFP. To enable reliable laminar measurements, we adopted a dense sampling design. We acquired submillimeter fMRI data from a single subject during resting state and a visuomotor task. VASO data produced cerebral blood volume (CBV)–weighted signals with interleaved BOLD contrast. Regions of interest in motor and visual cortices were defined using HCP-MMP1.0 parcels with task-evoked activation. An anatomical scan (MP2RAGE) was used to segment gray matter and generate depth bins, which were applied to extract laminar task responses. We used resting state data to estimate intra-regional functional connectivity, summarizing each layer's connectivity by its correlation to the mean of all other layers.

The BOLD task responses showed a bias towards the surface that was less pronounced in bSSFP; CBV showed such a bias for V1 only. Resting state correlations exhibited more uniform patterns for all contrasts. These initial findings support the feasibility of VASO and bSSFP for laminar analyses and motivate their further evaluation in our ongoing study.