Retinal metric: a stimulus distance measure derived from population neural responses

TL;DR

This paper introduces a neural metric derived from retinal responses that measures stimulus distinguishability, revealing a non-Euclidean structure that impacts neural decoding strategies.

Contribution

It presents a novel stimulus distance measure from retinal neural responses that captures the true distinguishability, differing from traditional static metrics.

Findings

Retinal distance deviates from Euclidean metrics.

The neural metric reveals stimulus features the retina is sensitive to.

Non-Euclidean structure affects neural decoding approaches.

Abstract

The ability of the organism to distinguish between various stimuli is limited by the structure and noise in the population code of its sensory neurons. Here we infer a distance measure on the stimulus space directly from the recorded activity of 100 neurons in the salamander retina. In contrast to previously used measures of stimulus similarity, this "neural metric" tells us how distinguishable a pair of stimulus clips is to the retina, given the noise in the neural population response. We show that the retinal distance strongly deviates from Euclidean, or any static metric, yet has a simple structure: we identify the stimulus features that the neural population is jointly sensitive to, and show the SVM-like kernel function relating the stimulus and neural response spaces. We show that the non-Euclidean nature of the retinal distance has important consequences for neural decoding.