Identification of Functional Subunits of the Human Cortex using Resting State fMRI

r i rA i i D y i D r R μ μ μ μ − − ∈ , where ) ( . Method: Imaging was performed on a 3D Siemens Trio scanner at the Yale MRRC. A T1-weighted 3-plane localizer was used to localize the slices to be obtained and T1 anatomic scans were collected in the axial-oblique orientation parallel to the ac-pc line. Resting state connectivity data was obtained using a gradient echo T2*-weighted echo planar imaging sequence, 64×64 matrix, alpha/TE/TR = 80/30ms/1550ms, with 25 slices 6mm thick, slip 0mm, 22×22 cm FOV, providing whole-brain coverage with voxel size of 3.4mm×3.4mm×6mm. Eight 6-min runs of resting state data were collected. Data was motion corrected using SPM5 and slice time corrected. Physiological noise from both respiration and cardiac pulsations were removed using the RETROICOR approach. Time courses were detrended and lowpass filtered before the segmentation algorithm was applied. Results: The segmentation results were averaged over a group of 22 subjects. I. Resting-State Vs. Histology: resting state time courses from Brodmann Areas BA 17/18 were extracted from each subject. The weighted Kmeans clustering was applied with R=2. The partitioning was highly consistent across subjects. The composite map of the 2-way segmentation is shown in Fig. 1(B). Comparing to the BA map shown in Fig 1(A), we see that the two maps are in general similar to each other, while the red region in Fig 1(B) extends more medially. II. Resting-State Vs. Task-based: an area in the intraparietal sulcus was discovered to be involved in working memory update/refresh tasks. Resting state data from the same region were recruited for a 2-way segmentation. The results from the resting state data show similar delineation of the functional subunits responding to different tasks (Fig 2). Conclusion: The segmentation is driven by the resting state fMRI with no structure information. The approach generates stable functional subdivision in both sensory and cognitive regions, which show large overlap with known histology and task-based findings. The results support the idea of developing a functional brain atlas based on resting state fMRI.