Loading…
CNS*2020 Online has ended
Welcome to the Sched instance for CNS*2020 Online! Please read the instruction document on detailed information on CNS*2020.
Back To Schedule
Sunday, July 19 • 8:00pm - 9:00pm
P82: Contrast invariant tuning in primary visual cortex

Log in to save this to your schedule, view media, leave feedback and see who's attending!

Feedback form is now closed.
Zoom:    https://unimelb.zoom.us/j/97678012117?pwd=T2dBNE5BOVBxM3hSYXRUSnF0OXBQdz09 
  Password: 677923

Hamish Meffin
, Ali Almasi, Michael R Ibbotson

Previous studies show that neurons in primary visual cortex (V1) exhibit contrast invariant tuning to the orientation of spatial grating stimuli [1]. Mathematically this is equivalent to saying that their response is a multiplicatively separable function of contrast and orientation.

Here we investigated the contrast dependence of V1 tuning to visual features in a more general framework. We used a data-driven modelling approach [2] to identify the spectrum of spatial features to which individual V1 neurons were sensitive, from our recordings of single unit responses in V1 to white (Gaussian) noise and natural scenes. For each cell we identified between 1 and 5 spatial feature dimensions to which the cell was sensitive (e.g. Fig. 1A, with 2 feature dimensions; feature 1 & 2 as labelled, with red showing bright and blue showing dark regions of the feature). The response of a neuron to its set of features was estimated from the data as the spike rate equal to a function of the individual feature-contrasts :

r = F(c1,…,cK) (Eq. 1)

where c1,…,cK are the contrast levels of a cell’s spatial features, 1,..K, embedded in any stimulus (e.g. Fig. 1B). These features spanned a subspace, giving a spectrum of interpolated features to which the cell was sensitive (Fig. 1A, examples labelled). The identity of these features varied along the angular polar coordinate in this subspace, which we term the feature-phase, φ (Fig. 1A, labelled). In this angular dimension, characteristics of the features, such as their spatial phase, orientation or spatial frequency, were found to vary continuously. In the radial coordinate, the contrast of these features varied, c=||( c1,…,cK) ||(Fig. 1A, labelled).

We found that the neural response above the spontaneous rate, r0, was well approximated by a multiplicatively separable function of the feature-contrast and feature-phase (Fig. 1C):

r = fc(c) fφ(φ) + r0  (Eq.2) To quantify the accuracy of this approximation, we calculated a relative error between the original and separable forms of the feature-contrast response function (i.e. Eq. (1) & (2)). This relative error varied between 2% and 18% across the cell population, with a mean of 6%. This indicates that for most cells, the separable form of the feature-contrast response function was a good approximation.

This result may be interpreted as demonstrating a form of contrast invariant tuning to feature-phase in V1. This tuning to feature-phase is given by the function fφ(φ) (Fig. 1E), and the contrast response function is given by fc(c) (Fig. 1D). As several feature characteristics such as spatial phase, orientation or spatial frequency covary with feature-phase, this also leads to contrast invariant tuning under covariation in these characteristics as feature-phase varies.

**Acknowledgements ** The authors acknowledge the support the Australian Research Council Centre of Excellence for Integrative Brain function (CE140100007), the National Health and Medical Research Council (GNT1106390), and Lions Club of Victoria.

References

1. Alitto, H. J., & Usrey, W. M. (2004). Journal of neurophysiology , 91 (6), 2797-2808.
2. Almasi, A., Meffin, H., Cloherty, S. L., Wong, Y., Yunzab, M., & Ibbotson, M. R. (2020). Mechanisms of feature selectivity and invariance in primary visual cortex. bioRxiv.

Speakers
avatar for Hamish Meffin

Hamish Meffin

Senior Researcher, Department of Biomedical Engineering, The University of Melbourne
Primary interests: Visual neuroscience, data driven modelling, modelling electrical stimulation of neurons



Sunday July 19, 2020 8:00pm - 9:00pm CEST
Slot 14