Joseph DeGutis^1,2 (), Asha Jotwani^1,2, Mia Haidamus^1,2, Regan Fry^1,2, Leah Kirsch^1,2, Maruti Mishra^1,2,3, Alison Campbell^1,4; ¹Boston Attention and Learning Laboratory, VA Boston Healthcare System, Boston, MA, ²Department of Psychiatry, Harvard Medical School, Boston, MA, ³Department of Psychology, California State University Bakersfield, CA, ⁴Department of Psychiatry, Boston University Chobanian & Avedisian School of Medicine, Boston, MA

Numerous studies have demonstrated that efficient processing of the information-rich eye region is important for face recognition. However, the extent to which eye processing is impaired in developmental prosopagnosia (DP) and how these potential impairments contribute to DPs’ face perception deficits remains unclear. To address this, we had 110 DPs and 133 controls perform validated face matching assessments (Cambridge Face Perception Test, Computerized Benton Facial Recognition Test, Same/Different Face Matching Test) and the Georges and Part-Whole tasks to measure feature (eye, nose, mouth, chin, forehead sensitivity) and holistic processing (part-whole holistic advantage) abilities. Regarding feature processing, DPs were most impaired on the eye conditions (Georges eye spacing and size trials, Part-Whole part eye trials) but showed relatively normal mouth (Georges mouth size trials and Part-Whole part mouth trials, though deficient Georges mouth spacing trials) and chin performance. DPs also showed a significantly reduced Part-Whole holistic advantage. We next ran factor analyses in DPs and controls using all the facial feature conditions and holistic advantage. In both groups, we found a three-factor solution: a clearly distinct eye processing factor as well as less distinct lower facial feature (mouth and chin trials) and holistic processing factors. Notably, in regression models using factor scores to predict face perception ability (face matching performance), both DPs and controls showed that eye processing ability predicted the most unique variance, while holistic processing also predicted unique variance. Finally, in a follow-up experiment to determine the generality of DP eye impairments, we examined eye gaze perception and reading emotion from the eyes and found subtle impairments in DPs. In sum, these results demonstrate that, in both DPs and controls, eye processing ability is distinct from more general facial feature processing and holistic processing and is a crucial contributor to DP face perception deficits.

Acknowledgements: Funded by National Eye Institute R01EY032510 awarded to JD