Purchase this article with an account.
Chou Hung, Chang Mao Chao, Li-feng Yeh, Yueh-peng Chen, Chia-pei Lin, Yu-chun Hsu, Ding Cui; Monkey neuronal assemblies predict (across objects) human fMRI and behavior. Journal of Vision 2013;13(9):1004. doi: 10.1167/13.9.1004.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Complex systems (e.g. for object recognition) are classically explored via hierarchical computational models that are derived from principles of primary sensory cortex. Biological guidance for higher stages of these models remains poor, and existing approaches cannot bridge this gap. Single-electrode recordings do not reveal how neurons act as part of local assemblies, e.g. populations of neurons spanning ~500 μm grouped by tuning covariation and coincident spiking. Functional MRI is an indirect measure that blurs across multiple assemblies within each mm[sup]3[/sup] voxel. This problem of spatial scale is compounded by investigator-driven biases in stimulus manipulation, making it virtually impossible to find specific neuronal assemblies and circuits underlying the same complex feature computations across animals and species, or even to confirm feature-specific results across studies.
We measured spiking assemblies in inferior temporal cortex, the last stage of the macaque visual object pathway, by combining precise mapping via multi-electrode arrays with precise measurement of assembly tuning covariation and coincident spiking. The assemblies were tuned to complex features in stimulus space, and these ‘key features’ enabled identification of the same feature assembly in another monkey via optical imaging. These monkey complex features also accurately predicted, across objects and in humans, fMRI patterns in lateral occipital complex and perception of a complex visual illusion. Altogether, this preliminary evidence suggests that a common library of complex visual object features, with common organizing principles and computations, is shared across individuals and across species.
Meeting abstract presented at VSS 2013
This PDF is available to Subscribers Only