Examining Hering’s theory for color responses in human V1 and V4

Poster Presentation 53.304: Tuesday, May 23, 2023, 8:30 am – 12:30 pm, Banyan Breezeway
Session: Color, Light, and Materials: Neural mechanisms, models

There is a Poster PDF for this presentation, but you must be a current member or registered to attend VSS 2023 to view it.
Please go to your Account Home page to register.

William Narhi-Martinez1 (), Zitong Lu1, Angela M. Brown1, Julie D. Golomb1, Delwin T. Lindsey1; 1The Ohio State University

We report an fMRI-based examination of Hering’s influential theory of color appearance based on color opponency. Our study asked whether representations of stimulus color in V1 and V4, cortical regions containing color-selective cells, vary according to Hering’s theory when participants perform a Hering-style hue-scaling task. While spectral opponency has been observed in primate visual cells, no one has yet found neurophysiological evidence for Hering-style color processing in the human brain. On each trial, participants viewed a 2-second duration colored stimulus (rotating spiral), with occasional null (gray) stimuli. On each trial, the spiral was colored with one of 8 hues: each participant’s personal unique red, yellow, green, and blue hues, plus their four binary-composite hues: cyan, lime, orange and purple. Over 2 scanning sessions, participants viewed each of these hues while performing 4 different tasks (16 trials per hue per task). Within a given task block, participants decided “What proportion of the spiral’s color has ___-ness”, with one color percept (“redness”, “greenness”, “blueness”, or “yellowness”) assigned to each block. Participants indicated by button press their responses on a 4-alternative scale: “none”, “some”, “mostly” or “all.” We localized each individual participant’s left and right V1 and V4 ROIs and used Representational Similarity Analysis to compare the neural similarity patterns based on ROI voxel activations to hypothetical similarity patterns derived from several plausible theoretical models of representational similarity, which were based on the colored stimuli and/or the task. Both V1 and V4 activation patterns were most similar to a model based on each individual’s responses to the hue scaling task. Our results suggest that the neural patterns in these regions reflect not simply the colorimetric properties of the stimuli participants were viewing, but were strongly influenced by top-down modulation based on task-specific attention and/or responses.

Acknowledgements: NSF BCS-1152841 (D.T.L.), NIH R01-EY025648 (J.G.), NSF 1848939 (J.G.)