Researchers developed a framework using population-specific artificial neural network models to predict image-level judgments for autistic and neurotypical participants, aiming to address variability in facial emotion perception studies.

  • The models were used to select novel faces predicted to maximize group separation, which produced larger behavioral differences than matched random images in an independent cohort.
  • A generative adversarial network was employed to transform diagnostic images toward greater predicted group agreement, reducing behavioral separation in phenotype-matched validation.
  • This approach establishes a model-guided method for discovering and transforming stimuli that reveal population-specific perceptual differences.

The results demonstrate how behavioral phenotyping can move beyond averaging across fixed stimulus sets toward optimized assays that identify conditions under which neurodivergent perception diverges or converges.