HIGH-PERFORMANCE PROCESSING AND STREAMING ARCHITECTURE FOR MULTIMODAL NEURO-MOTOR SIGNAL ANALYSIS

This paper presents performance-focused addition to multimodal neuro-motor analysis software that integrates electroencephalography (EEG) and motor deviation signals (ΔR). While traditional diagnostic approaches rely on static session-based evaluation, this work introduces computational complexity evaluation of core analytical modules, a scalable serverless cloud architecture for streaming biomedical data signals, and a transition from static indicators to time-resolved dynamic diagnostic parameters. Execution complexity of Fast Fourier Transform (FFT), cross-correlation between ΔR and 16 EEG channels, and a derived Sinusoidality Index (SI) is formally evaluated, demonstrating computational feasibility under near real-time constraints. A streaming-ready AWS-based architecture (includes data flow from WebSocket, API Gateway, Kinesis, Lambda to S3) is proposed to support batch diagnostics and continuous monitoring scenarios. The proposed framework transforms neuro-motor indicators from static session to dynamic time-dependent functions, enabling scalable medical analytics aligned with software engineering.