EEG Study Links AI-Generated Biodigital Architecture to Emotional Responses

A recent study featured on arXiv (2607.24808) investigated emotional reactions to images of AI-generated biodigital architecture using electroencephalographic (EEG) data. Initially, a pre-experiment involving 336 participants narrowed down 60 images from a selection of 600 that provoked strong feelings categorized as awe, disgust, or contentment. EEG recordings were then conducted with 52 volunteers, where channel selection and sample size were informed by existing dataset analyses. The findings revealed that gamma and delta bands provided the best classification accuracy, with the gamma band reaching 77.07 percent +/- 13.8 percent for awe. Notably, elements like greenery and varied textures were associated with positive emotions, while dampness was linked to negative feelings, highlighting the importance of these aspects in biodigital architecture design.