Purchase this article with an account.
Gerald P. Keith, Joseph F.X. DeSouza, Xiaogang Yan, Hongying Wang, J. Douglas Crawford; A new method for determining neuron receptive field reference-frames. Journal of Vision 2008;8(6):806. doi: 10.1167/8.6.806.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Mapping neuron receptive fields (RFs) only leads to an understanding of the role these neurons play in sensorimotor behavior if the proper reference-frame of each RF is determined as well. Our goal was to develop a method by which this reference-frame could be determined from the RF mapping itself. We developed this method using a data set that included single-unit recordings from 80 neurons in the intermediate superior colliculus of two head-unrestrained monkeys during performance of a visual saccade task. Using head-unrestrained animals represents an improvement over head-restrained animals in that head movements allow a more natural behavior and open up an additional reference frame in which neuron RFs might be represented. Because the torsional component of eye and head are significant in this condition, we recorded both positions in 3-D using dual search coils. We plotted RFs using both the final gaze and visual target directions relative to initial gaze fixation, in eye, head and space reference-frames. While the requirements of target fixation constrained both initial and final gaze directions, freedom of movement of both eye and head produced a natural variation in these orientations across successive trials. This meant that the RFs plotted in different reference frames would themselves be different. We modeled this using hypothetical RFs, and found that these variations produced a ‘smearing’ of the RF in all reference frames other than the proper reference-frame associated with the neuron's activity. A non-parametric fitting of each RF in the different reference-frames, and a Levine test comparing the variances of the residuals of these fits, produced a significant difference when eye and head orientation variability were of a size similar to that shown by monkeys performing saccade tasks, so that the proper reference-frame could be distinguished. We confirmed this using our actual neurophysiological data. Supported by: CIHR (Canada). JDC holds a Canada Research Chair.
This PDF is available to Subscribers Only