From neural decoding to design intervention: A systematic review of human emotional responses to architectural spaces through neuroimaging techniques

• EEG+VR is the primary experimental method. Highlights overreliance on VR simulations (66.7%) and EEG-only approaches (83.3%), with severe deficits in real-world multimodal validation (8.3%) and longitudinal neuroplasticity studies • The most commonly focused independent variables are spatial proportions and colors, followed by geometric shapes, scale, and window size. • Identifies neuro-design thresholds for emotion regulation. Reveals evidence-based spatial parameters (window-wall ratio 0.4–0.6, depth-width ratio 1.6:1, ceiling height >3.2m) that significantly modulate neural biomarkers—α/β oscillations (relaxation-arousal balance) and ventral striatum activation (aesthetic valuation)—providing quantifiable benchmarks for emotion-driven design. Architectural spaces profoundly shape human emotional states, yet the neural mechanisms underlying spatial-emotional interactions remain underexplored in design frameworks. This systematic review synthesizes evidence from 17 neuroimaging studies (2009-2023) to establish methodological paradigms and neural correlates of emotion-driven spatial design. Following PRISMA guidelines, we identified three critical findings: (1) Methodological convergence: electroencephalogram dominates neuro-architectural research (83.3% studies), predominantly paired with virtual reality (66.7%) for dynamic emotion tracking, while real-world assessments constitute only 8.3%; (2) Neural mechanisms: Spatial proportions (width-depth ratios), curvilinear forms, and openness significantly modulate relaxation-arousal balance through α/β oscillations and anterior cingulate cortex activation, while aesthetic valuation engages visual-spatial networks (precuneus, middle temporal gyrus) and reward circuits (ventral striatum); (3) Design thresholds: Empirical parameters such as window-to-wall ratios (0.4-0.6) and spatial proportions (1.6:1) emerge as neurophysiological benchmarks for emotion-oriented design. The review highlights an urgent need for multimodal frameworks integrating EEG, fMRI, and behavioral metrics to decode complex spatial interactions. Future research should prioritize longitudinal neural plasticity studies, computational models of design parameter optimization, and neuroevidence-based design guidelines. This work advances a paradigm shift from intuition-driven to neuro-informed spatial design by bridging the gap between neuroscientific insights and architectural practice.