Browse by author
Lookup NU author(s): Professor Alexander Thiele
This work is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0).
© The Author(s) 2024.Perceptual learning refers to an improvement in perceptual abilities with training. Neural signatures of visual perceptual learning have been demonstrated mostly in mid- and high-level cortical areas, while changes in early sensory cortex were often more limited. We recorded continuously from multiple neuronal clusters in area V1 while macaque monkeys learned a fine contrast categorization task. Monkeys performed the contrast discrimination task initially when a constant-contrast sample stimulus was followed by a test stimulus of variable contrast, whereby they had to indicate whether the test was of lower or higher contrast than the sample. This was followed by sessions where we employed stimulus roving; i.e. the contrast of the sample stimulus varied from trial to trial. Finally, we trained animals, under ‘stimulus roving-with-flanker’ conditions, where the test stimuli to be discriminated were flanked by ‘flanking stimuli’. Perceptual discrimination abilities improved under non-roving conditions and under roving-with-flanker conditions as training progressed. Neuronal discrimination abilities improved with training mostly under non-roving conditions, but the effect was modest and limited to the most difficult contrast. Choice probabilities, quantifying how well neural activity is correlated with choice, equally increased with training during non-roving, but not during either of the roving conditions (with and without flankers). Noise correlations changed with training in both monkeys, but the changes were not consistent between monkeys. In one monkey, noise correlations decreased with training for non-roving and both roving conditions. In the other monkey, noise correlations changed for some conditions, but lacked a systematic pattern. Thus, while perceptual learning occurred under non-roving and roving-with-flanker conditions, the changes in neural activity in V1 were overall modest and were essentially absent under the different roving conditions.
Author(s): Thiele A, Chen X, Sanayei M, Chicharro D, Distler C, Panzeri S
Publication type: Article
Publication status: Published
Journal: Journal of Cognitive Enhancement
Year: 2024
Pages: epub ahead of print
Online publication date: 23/05/2024
Acceptance date: 25/04/2024
Date deposited: 04/06/2024
ISSN (print): 2509-3290
ISSN (electronic): 2509-3304
Publisher: Springer Nature
URL: https://doi.org/10.1007/s41465-024-00298-x
DOI: 10.1007/s41465-024-00298-x
Data Access Statement: Original data for figure reproduction are stored on Newcastle University servers (https://doi.org/10.25405/data.ncl.25109123) and are publicly available. Code Availability Data were processed with Neuralynx and custom-written Matlab code. Software for figure reproduction is stored on Newcastle University servers (https://doi.org/10.25405/data.ncl.25109123) and are publicly available.
Altmetrics provided by Altmetric
