|Date & Location||Speakers|
Friday, April 18
2:00pm - 3:00pm
|Rachel Denison, Postdoctoral Fellow at NYU
Functional Mapping of the Magnocellular and Parvocellular Subdivisions of Human LGN
The magnocellular (M) and parvocellular (P) subdivisions of primate LGN are known to process complementary types of visual stimulus information, but a method for noninvasively defining these subdivisions in humans has proven elusive. To functionally map the M and P subdivisions of human LGN, we used high-resolution fMRI at high field (7T and 3T) together with a combination of spatial, temporal, luminance, and chromatic stimulus manipulations. We found that stimulus factors that differentially drive magnocellular and parvocellular neurons in primate LGN also elicit differential BOLD fMRI responses in human LGN and that these responses exhibit a spatial organization consistent with the known anatomical organization of the M and P subdivisions. Mapping LGN subdivisions opens possibilities for investigating their functions in human visual perception, in patient populations with suspected abnormalities in one of these subdivisions, and in visual cortical processing streams arising from parallel thalamocortical pathways.
|Friday, April 25
3:00 pm - 4:00 pm
|Noah Benson, Postdoctoral Fellow at University of Pennsylvania
Distortion of the Flattened Cortical Surface Enables Prediction of V1, V2, and V3 Retinotopic Maps From Individual Subject Anatomy
Idealized, map-like models of how location on the cortical surface is linked with sensory representation are common in sensory neuroscience. Within vision, algebraic transformations have been proposed to relate position in the visual field to the retinotopic representation of polar angle and eccentricity within cortical visual areas V1, V2, and V3. One limitation of these idealized models is that they are typically described with reference to a flat cortical sheet. Because the cortical surface is curved, however, flattening it to a plane unavoidably introduces local geometric distortions which are not accounted for in idealized models.
Here, we demonstrate how this limitation is overcome using a mass-spring-damper simulation to register functional MRI retinotopic mapping data to an idealized model of visual areas V1, V2, and V3. The simulation maximizes the alignment of the model and the functional data while respecting anatomical constraints. The resulting, registered cortex may be used to accurately predict the location and retinotopic organization of these early visual areas. Moreover, we show that the prediction accuracy extends beyond the range of data used to inform the model, indicating the the registration is accurately reflecting fundamental features of the retinotopic organization of visual cortex.