Comparing physiological emotional responses to traditional hutongs and modern streets in Beijing through EEG and an interpretable machine learning approach

Abstract

As public spaces that urban populations interact with in their daily lives, streets have been found to affect people’s emotional well-being. This study utilized electroencephalography (EEG) to measure and compare the emotional effects of traditional hutongs and modern streets. By combining Random Forest (RF) and Shapley Additive Explanations (SHAP) analysis, the research explored key built environment factors and their positive and negative impacts. The study found: (1) The questionnaire data revealed that traditional hutongs enhance calmness and deep relaxation more effectively than modern streets, as mirrored by the physiological responses recorded in the EEG analysis. (2) In traditional hutongs, $\alpha$ brainwaves are elicited by a green view index above 5.57 and enclosed spaces (width-to-height ratio <0.64), while an architectural form richness exceeding 3.57 enhances $\beta$ brainwaves. (3) In modern streets, a positive $\alpha$ brainwave response requires significantly higher environmental thresholds, with the green view index needing to exceed 15.46. Notably, a high floor area ratio demonstrates a dual effect: it boosts $\beta$ brainwaves when exceeding 3.21 but inhibits $\theta$ brainwaves once it exceeds 2.28. Based on these findings, we propose recommendations for urban planners and designers to optimize design strategies to enhance emotional well-being.