Exploratory EEG Feature Profiles During Viewing of Six Combined Indoor Spatial-Color Conditions

Abstract

Indoor visual environments combine room function, spatial composition, materials, and color, but these factors are often examined with subjective ratings alone. This study developed an exploratory electroencephalography (EEG) workflow for describing neural feature profiles during viewing of six non-factorial, combined indoor spatial–color conditions. Twenty adults completed three repeated runs of a counterbalanced virtual-walkthrough paradigm while 32-channel EEG was recorded at 500 Hz. Participant-specific Latin-square sequences were used to reconstruct the actual condition represented by each F1–F12 event marker. The processing archive contained 720 planned event positions, 711 unique F markers after duplicate removal, 706 epochable events, and 691 clean event-locked epochs (97.9% of epochable events). Features comprised signed mean amplitudes in literature-supported event-related-potential (ERP) windows, sustained theta/alpha/beta relative power, early/middle/late relative-power summaries, and band-limited Hilbert power-change and phase-consistency summaries. Runs were processed separately, and features were summarized within-participant before group-level description. The primary article reports no inferential results; the condition-level values, participant distributions, and waveform summaries are presented only as exploratory descriptions. The observed profiles varied across event stage, scalp region, component window, and frequency band but they do not identify a pure color or room effect, establish condition superiority, support a causal neurophysiological mechanism, or provide a current indoor-design prescription. The principal contribution is a traceable workflow for sequence-aware event mapping, quality-control accounting, and multi-domain descriptive EEG feature extraction that can inform a future prospectively powered, factorial validation study.

The authors' abstract, as published at the source. Sensors, 2026 · DOI ↗