High-Density Mobile EEG Preprocessing Pipeline, Turn Detection, and Component-Level Spectral Analysis During Gait

This preregistration documents a standardized preprocessing and analysis pipeline for high-density mobile EEG recorded during overground walking and turning, synchronized with inertial measurement unit (IMU). The primary purpose of this registration is to ensure methodological transparency, reproducibility, and temporal precedence of all signal processing decisions prior to cohort-specific hypothesis testing and statistical analyses. The pipeline is designed for movement-intensive, ecologically valid recordings and addresses challenges inherent to mobile EEG, including motion artifacts, line noise instability, and non-stationary high-amplitude disturbances. Preprocessing includes high-pass and low-pass filtering, adaptive line-noise removal, automated bad channel detection, common average re-referencing, Artifact Subspace Reconstruction (ASR) with predefined thresholds, and Adaptive Mixture Independent Component Analysis (AMICA) with iterative sample rejection. Independent components are classified using combined dipole localization and probabilistic criteria to isolate brain-derived sources based on predefined residual variance and classification thresholds. A reproducible turn-detection algorithm based on IMU-derived yaw angle and angular velocity is also specified, including filtering, smoothing, derivative computation, thresholding procedures, and quality control steps. All parameters, thresholds, and processing decisions are explicitly defined to minimize analytic flexibility. This preregistered pipeline will serve as a reusable framework for future analyses, ensuring consistency across datasets while allowing hypothesis-driven statistical models to be specified separately for each study.