Skip to contents

PhysioIO PhysioIO logo

I/O Functions for PhysioExperiment Objects

PhysioIO provides comprehensive file I/O capabilities for the PhysioExperiment ecosystem. With 50 exported functions, it supports reading and writing physiological signal data across all major formats used in neuroscience, rehabilitation, and clinical research – including EDF/EDF+, HDF5, BIDS, CSV, MATLAB .mat, and RDS. It also integrates with DuckDB for efficient querying and management of large-scale physiological datasets.

Installation

You can install PhysioIO from r-universe:

install.packages("PhysioIO",
  repos = c("https://x-biosignal.r-universe.dev", "https://cloud.r-project.org"))

Or install the development version from GitHub:

# install.packages("remotes")
remotes::install_github("x-biosignal/PhysioIO")

Quick Start 

library(PhysioIO)

# Read an EDF file into a PhysioExperiment object
pe <- readEDF("recording.edf")
pe
#> PhysioExperiment with 256000 timepoints x 32 channels
#> Sampling rate: 512 Hz | Duration: 500.0 s
#> Assays: raw

# Write to HDF5 for out-of-memory analysis of large datasets
writePhysioHDF5(pe, "recording.h5")
pe_h5 <- readPhysioHDF5("recording.h5")
isHDF5Backed(pe_h5)  # TRUE

# Export to BIDS format
writeBIDS(pe, path = "bids_dataset", subject = "01", session = "pre",
          task = "rest", modality = "eeg")

# Query from a DuckDB database
con <- physioDBConnect("experiments.duckdb")
physioDBRegister(con, pe, dataset_id = "rest_eeg_01")
pe_loaded <- physioDBLoad(con, "rest_eeg_01")
physioDBDisconnect(con)

Features

EDF/EDF+ Format

Full support for the European Data Format, the most widely used standard for polysomnography, EEG, and clinical recordings:

  • readEDF() – read EDF and EDF+ files with automatic header parsing
  • writeEDF() – write PhysioExperiment objects to EDF format

HDF5 Format

High-performance hierarchical data format with out-of-memory support for large datasets:

BIDS Format

Brain Imaging Data Structure support for standardized neuroimaging datasets:

CSV Format

Flexible CSV import and export for tabular physiological data and clinical metadata:

  • readCSV() – read CSV files with automatic channel detection
  • writeCSV() – export signal data and metadata to CSV
  • Clinical metadata CSV functions for participant and session information

MATLAB .mat Format

Interoperability with MATLAB-based analysis pipelines:

  • readMAT() – import .mat files (v5 and v7.3/HDF5-based)
  • writeMAT() – export PhysioExperiment objects to .mat format

RDS Format

Native R serialization for fast save/load workflows:

DuckDB Database Integration

Efficient database-backed storage and querying for large-scale studies:

  • physioDBConnect() – connect to a DuckDB database
  • physioDBRegister() – register a PhysioExperiment in the database
  • physioDBQuery() – query experiments by metadata criteria
  • physioDBLoad() – load experiments from the database
  • physioDBDisconnect() – close the database connection

Dependencies

  • R (>= 4.2)
  • PhysioCore
  • HDF5Array
  • rhdf5
  • jsonlite
  • DBI

PhysioExperiment Ecosystem

PhysioIO is the I/O layer of the PhysioExperiment ecosystem, a suite of R packages for multi-modal physiological signal analysis:

Package Description
PhysioCore Core data structures and accessors
PhysioIO File I/O (EDF, HDF5, BIDS, CSV, MAT)
PhysioPreprocess Preprocessing (filters, ICA, resampling)
PhysioAnalysis Analysis and visualization

Visit the r-universe page to browse all available packages.

License

MIT License. See LICENSE for details.

Author

Yusuke Matsui

Governance & support

Part of the Physio ecosystem. Community and policy documents live in the umbrella repository: