A simple linear readout of MT supports motion direction-discrimination performance

Abstract

Motion discrimination is a well-established model system for investigating how sensory signals are used to form perceptual decisions. Classic studies relating single-neuron activity in the middle temporal area (MT) to perceptual decisions have suggested that a simple linear readout could underlie motion discrimination behavior. A theoretically optimal readout, in contrast, would take into account the correlations between neurons and the sensitivity of individual neurons at each time point. However, it remains unknown how sophisticated the readout needs to be to support actual motion-discrimination behavior or to approach optimal performance. In this study, we evaluated the performance of various neurally plausible decoders, trained to discriminate motion direction from small ensembles of simultaneously recorded MT neurons. We found that decoding the stimulus without knowledge of the interneuronal correlations was sufficient to match an optimal (correlation aware) decoder. Additionally, a decoder could match the psychophysical performance of the animals with flat integration of up to half the stimulus and inherited temporal dynamics from the time-varying MT responses. These results demonstrate that simple, linear decoders operating on small ensembles of neurons can match both psychophysical performance and optimal sensitivity without taking correlations into account and that such simple read-out mechanisms can exhibit complex temporal properties inherited from the sensory dynamics themselves. NEW & NOTEWORTHY Motion perception depends on the ability to decode the activity of neurons in the middle temporal area. Theoretically optimal decoding requires knowledge of the sensitivity of neurons and interneuronal correlations. We report that a simple correlation-blind decoder performs as well as the optimal decoder for coarse motion discrimination. Additionally, the decoder could match the psychophysical performance with moderate temporal integration and dynamics inherited from sensory responses.

Behavioral read out of population value signals in primate orbitofrontal cortex

“…In concert with other findings [2][3][4] it evinces the likely causal role of MT neurons in motion perception, and, more broadly, provides the foundation for a robust computational framework for understanding how motion evidence is represented by neurons and transformed into a decision [5][6][7] .…”

“…The apparently linear read-out of value signals for value-based choice indicates potential similarities to mechanisms for perceptual motion discrimination, in which a linear read-out of motion-sensitive MT cells is sufficient to explain motion judgements 3,7 . The study of perceptual decision mechanisms may therefore provide a fruitful roadmap for future experiments.…”

Decoding of Attentional State Using High-Frequency Local Field Potential Is As Accurate As Using Spikes

“…The spiking activity of a simultaneously recorded population of neurons has also been used to predict the animals' focus of attention and behavior in a single trial Maunsell 2010, 2011). Since handling high-dimensional data is difficult, many studies have devised methods to obtain a onedimensional metric from a high-dimensional dataset to predict behavior (Cohen and Maunsell 2010;Cunningham and Yu 2014;Yates et al 2020). Using a simple method of projecting trials onto a one-dimensional axis constructed using the mean spiking activity of a neuronal population of two different attention conditions (Cohen and Maunsell 2010;Mayo et al 2015), Mayo and Maunsell (2016) showed that the projections of neutral cueing condition were also intermediate between cued and uncued conditions.…”

Stimulus-choice (mis)alignment in primate area MT

“…For simplicity, in this toy example, we assume there is no structure in the co-variability that can be exploited for decoding. In this case, increased variability along the stimulus axis (the so-called information-limiting noise) will change the amount of information about the stimulus, while, importantly, variability orthogonal to the stimulus axis will not [19,[22][23][24][25]. We call this variability direction the "non-stimulus axis" (Fig 1A).…”