neurokit2

Computing HRV and EDA metrics from raw biosignal recordings requires a sampling rate–aware workflow that cleans the signals, detects peaks and events, and returns structured physiological indices and analysis tables. This process supports both event-related epochs and continuous interval-related recordings for autonomic feature extraction.

Extracting EEG microstates and EOG blink events involves cleaning the raw time-series data, segmenting EEG signals into stable topographic maps, and detecting ocular artifacts via peak identification. The mechanism returns structured indices and analysis tables for interpreting neural dynamics and blink counts.

Processing multimodal ECG and respiration recordings together enables integrated indices like RSA by combining cardiovascular peak detection with respiratory phase tracking. This cross-signal analysis requires a compatible workflow that synchronizes the modalities based on their sampling rates to derive autonomic metrics.

This biosignal processing approach supports both event-related and interval-related analysis workflows. It handles event-related epochs by segmenting data around specific triggers and processes continuous interval-related recordings by deriving global metrics like complexity, entropy, and nonlinear HRV features across the entire signal.

The best way to clean ECG and EMG signals for physiological feature extraction is applying sampling rate–aware computations that filter noise and detect activation events. This prepares the raw time-series for deriving cardiovascular metrics like HRV and muscle activation features in structured analysis tables.

Biosignal processing metrics vary between continuous recordings and event-related epochs because interval-related analysis computes global complexity and entropy across the entire signal, while event-related analysis isolates specific physiological responses around triggers. Both require distinct peak detection and epoching workflows to generate accurate structured indices.