Retinotopic visual mapping of brain oxygenation and neuronal activity using simultaneous fast and slow near-infrared optical brain imaging in humans.
63.438, Wednesday, 21-May, 8:30 am - 12:30 pm, Banyan Breezeway
Kyle E. Mathewson1, Kathy A. Low1, Nils Schneider-Garces1,2, Antonio Chiarelli1, Chin Hong Tan1,2, Tania Kong1,2, Courtney R. Burton1, Mark A. Fletcher1,3, Benjamin Zimmerman1,3, Brad P. Sutton1,4, Edward L. Maclin1, Monica Fabiani1,2, Gabriele Gratton1,2; 1Beckman Institute for Advanced Science and Technology, University of Illinois, 2Department of Psychology, University of Illinois, 3Neuroscience Program, University of Illinois, 4Department of Bioengineering, University of Illinois
High-precision retinotopic mapping of the visual field normally uses BOLD signal. Given the divergence between neuronal and hemodynamic responses with aging, however, a more direct measure of neuronal activity is needed to make accurate maps of retinotopic organization. We employed tri-modal neuroimaging of 18-78 year-old subjects using BOLD-fMRI, along with a separate session of simultaneously acquired near-infrared spectroscopy (NIRS) and fast-optical imaging (FOI), to test the correspondence among retinotopic maps derived from these techniques. Identical stimuli presented in the MRI and optical sessions consisted of checkerboard gratings expanding or revolving every 48 seconds (1/48 Hz), flickering at 5 Hz. We passed modulated near-infrared light through visual cortex using 8 pairs of 690- and 830-nm laser diodes and 16 photomultiplier tube detectors. The source and detector locations were co-registered with the scalp from each individual’s MPRAGE image, and the light’s path through the tissue was modeled in the subject’s original brain space. The activity in each voxel was computed as a weighted sum of the amplitude (NIRS-HbO2) and phase (FOI) of light passing though each channel. We segmented and parcellated each subject’s MPRAGE using FreeSurfer, and flattened the occipital surface cut along the calcarine fissure, mapping both the fMRI and optical retinotopy results in this space. For both BOLD and HbO2, we extracted the phase and amplitude of the 1/48-Hz signal. For FOI we used short moving-window wavelets to compute the time course of 5-Hz flicker power in each voxel, from which the phase and amplitude of the 1/48-Hz grating were computed. Results show a strong correspondence between optical and BOLD retinotopic maps, onto which we can map event-related optical signals (EROS) evoked by stimuli of varying eccentricities. These results reveal the potential for retinotopic mapping in both space and time afforded by fast optical imaging of the human brain.