A neural timing model of visual threshold (Q1916542)

This theoretical paper introduces a model of incremental flash detection. The model accounts for psychophysical threshold data on the basis of assumptions that are part physiological and part psychophysical. The physiological part of the theory consists of a model of spike generation by idealized visual neurons that possess linear receptive field weighting functions. The psychophysical part of the theory consists of two assumptions: (1) that the observer monitors the neural mechanism that is ideally located and scaled for detecting the flash; and (2) that the observer employs a neural timing strategy in order to make target/blank discriminations based on a random sample of the neural spike train. The complete theory gives an account of the shapes of psychophysical threshold curves, including their dependence on test area and duration. Many of the results presented in this paper have also been previously reported in brief abstract form. The spike generation model that is proposed here is not a detailed biophysical model. It is, however, intended as a first-order approximation to the spike generation mechanism of real visual neurons. The spike generation model can be described quite succinctly; but many of its implications -- as will be demonstrated -- are nonintuitive.