figshare
Browse
Adaptation poster 2019.pdf (779.94 kB)

Probing contrast sensitivity and adaptation in the head-fixed mouse

Download (779.94 kB)
poster
posted on 2019-01-07, 16:18 authored by Miles WellsMiles Wells, Peter Zatka-Haas, Kenneth HarrisKenneth Harris, Matteo Carandini
The ability to detect a visual pattern depends on multiple attributes of the pattern and on the recent history of visual stimulation, which determines the observer’s state of adaptation. The dependence of sensitivity on these attributes has been measured in multiple species and related to physiological measurements of visual sensitivity. The perceptual effects of visual adaptation, however, have been studied mostly in humans. Thus it is unknown if perceptual phenomena of contrast adaptation are due to phenomena of contrast adaptation measured in the cortex.
To relate perceptual and physiological adaptation, we developed a method to probe contrast sensitivity and adaptation in the head-fixed mouse, where methods of population recordings and causal circuit identification are well developed. We modified a two-alternative unforced choice contrast detection task (Burgess et al. bioRxiv 2016) whereby mice receive a reward for turning a wheel to indicate which hemifield contains a grating stimulus, or for refraining to do so (a “no-go” choice) if no stimulus is present. Grating contrast and left-right position were varied randomly. These "test" stimuli were preceded by an adaptor stimulus (to be ignored): a grating varying randomly in spatial phase, whose contrast could be low (5%) or high (100%) in different blocks. To study the effect of adaptation we fitted a simple logistic choice model (Zatka-Haas et al, SfN 2016). The model fits the probabilities of choosing left, right, or no-go using parameters reflecting choice bias and stimulus sensitivity and a nonlinear function of contrast with one key parameter, the half-maximal contrast (C50).
Preliminary results in two mice show marked effects of contrast adaptation on contrast sensitivity. Both mice yielded high-quality psychometric curves with barely any errors at high contrast. Preceding the test stimuli with the high contrast adaptor had two effects: it markedly increased the tendency of the mice to respond, therefore giving more false alarms at zero contrast, and it increased the contrast needed by the mice to reach a threshold performance, as if it had physically reduced the contrast of the test stimulus. These effects were captured by increases in both the bias and C50 parameters.
We are currently testing this effect in more mice and measuring whether it depends on the relative orientation between adaptor and test. If the effect is confirmed, the simplicity and flexibility of this task will allow us to probe multiple aspects of contrast adaptation not only perceptually but also neurally, through simultaneous optogenetic inactivation or population recordings.

Funding

Neuronal integration in sensory cortex: New paths to discovery

Wellcome Trust

Find out more...

History

Usage metrics

    Licence

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC