Identifying functional subdivisions of the human MT complex

The motion-sensitive MT complex (MT+) in macaque monkey contains several functional subdivisions, including areas MT, MST and FST. Previous neuroimaging work has established functional localizers for the human analogue of areas MT and MST (Huk, Dougherty, & Heeger, 2002), but not FST. We employed a multifaceted approach, incorporating functional localizers, retinotopic organization, and myeloarchitecture to separate FST from MT and MST. We first used a traditional localizer to identify MT+ (radial and tangential 2D moving versus static dots). Considering FST’s sensitivity to 3D motion and its robustness to opponent motion in monkey, we then explored disparity-defined stereomotion (coherent versus temporally scrambled) and opponent motion (paired versus unpaired counter-phase oscillating dots) localizers. We also compared our results to myelin maps and retinotopic organization within the same participants. We consistently localized putative FST (pFST) within the ventral section of MT+. We found weaker responses to 2D motion, but stronger responses to stereomotion in pFST relative to MT/MST. Our results contrast with a previous report that identified an area outside and anterior to MT+ (Likova & Tyler, 2007). 2D motion and opponent motion localizers generally identified the same areas, suggesting a more suppressed response to locally opponent motion in MT/MST than in pFST. Retinotopic mapping aids in distinguishing MT and MST but is less effective for pFST. The MT+ foveal confluence was identifiable in all subjects, but the central visual field was overrepresented in pFST. Myelin mapping showed significant individual and hemispheric variability, though on average, converged at the border of MT and MST. Finally, in all participants, we observed mismatches between functionally and atlas-defined MT, MST, and pFST. Our results demonstrate the importance of functional localization for delineating subdivisions of human MT+.