EMG-BIDS: an extension to the Brain Imaging Data Structure for electromyography

Area of Science:

• Neuroscience
• Biomedical Engineering
• Data Science

Background:

• Electromyography (EMG) is crucial for clinical assessment, rehabilitation, and research.
• Current lack of a standardized format hinders EMG data sharing, reuse, and aggregation.
• Existing data structures like BIDS are not tailored for EMG's complexities.

Purpose of the Study:

• To introduce EMG-BIDS, an extension of the Brain Imaging Data Structure (BIDS), for standardizing EMG data organization.
• To address specific challenges in EMG data, including diverse electrode types, placements, and montages.
• To facilitate reproducible documentation of sensor positioning.

Main Methods:

• Developed EMG-BIDS as an extension to the BIDS specification.
• Incorporated hierarchical coordinate systems for precise electrode placement documentation.
• Integrated EMG-BIDS into existing BIDS-compatible tools like MNE-BIDS and EEGLAB.

Main Results:

• EMG-BIDS is now an official part of the BIDS standard (version 1.11.0).
• The specification successfully handles diverse EMG recording setups, including high-density surface EMG.
• Demonstrated utility through public datasets, showcasing reproducible sensor positioning.

Conclusions:

• EMG-BIDS establishes a foundation for FAIR (Findable, Accessible, Interoperable, Reusable) EMG data.
• Enables large-scale meta-analyses, multi-site studies, and machine learning applications.
• Promotes standardized and well-documented EMG datasets for advanced research.