Neural mechanisms underlying the hierarchical construction of perceived aesthetic value.
Level 4 - case-series / case-control
Level by design analogy; human functional neuroimaging and computational modeling study.
PubMed 36693833 · doi:10.1038/s41467-022-35654-y
What was done
The authors used functional neuroimaging (fMRI) combined with computational modeling and deep convolutional neural networks to investigate how the human brain computes the aesthetic value of complex visual art stimuli.
What was found
The abstract reports no numerical data or effect sizes. Qualitatively, aesthetic value was found to be constructed through a hierarchical, weighted integration of low- and high-level visual features represented across early and late visual cortex, parietal cortex, and lateral prefrontal cortex, culminating in an overall aesthetic value representation in the medial prefrontal cortex.
Why it matters
The findings outline a potential general neurocomputational architecture explaining how the brain rapidly integrates multi-level sensory and cognitive features into subjective value judgments for complex stimuli.
Limits
The abstract does not disclose participant sample size, demographics, statistical test metrics, or effect sizes. As an fMRI correlational and modeling study, it cannot establish direct causality, and generalizability beyond visual art stimuli remains unverified.
Cited by
- supports Computational neural network models decomposing low-level visual features of abstract art, such as colors and visual composition, can decode or predict whether human observers find the art pleasant or liked.