neurodsp-tools/neurodsp

What it solves

NeuroDSP provides a specialized set of tools for analyzing and simulating neural time series data. It simplifies the process of applying digital signal processing (DSP) techniques to understand oscillatory activity and rhythmic patterns in brain signals.

How it works

The library is organized into several functional modules that handle different stages of signal analysis:

• Filtering: Uses bandpass, highpass, lowpass, and notch filters to clean data.
• Spectral Analysis: Computes power spectra and estimates instantaneous measures of oscillatory activity.
• Pattern Detection: Identifies bursting oscillations and rhythmic or recurrent patterns in neural signals.
• Aperiodic Analysis: Analyzes non-periodic features of the time series.
• Simulation: Generates plausible simulations of neural time series, including both periodic and aperiodic components.

Who it’s for

Researchers and data scientists working with neural time series and brain signal analysis who need a standardized way to process, analyze, and simulate these signals.

Highlights

• Comprehensive Toolset: Covers everything from basic filtering to advanced spectral and aperiodic analysis.
• Simulation Capabilities: Includes tools to create synthetic neural data for testing and simulation.
• Integrated Plotting: Built-in utilities for visualizing neural time series and their derived measures.
• Scientific Rigor: Published in the Journal of Open Source Software (JOSS) and supported by the NIH.