A general method for remontaging based on a singular value decomposition algorithm.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Electroencephalography (EEG) data acquisition often involves various electrode montages.
• Converting between different EEG montages (e.g., referential, bipolar, Laplacian) is crucial for data analysis and comparison.
• Existing methods for montage conversion can be complex or limited in scope.

Purpose of the Study:

• To develop a general mathematical algorithm for converting any EEG montage to any other using linear transformation.
• To provide a robust method for recalculating EEG signals from different electrode configurations.
• To identify limitations in montage conversion, specifically which output channels cannot be derived from the input.

Main Methods:

• The algorithm represents input and output montages as matrices.
• Singular value decomposition (SVD) is employed to determine the linear transformation between montages.
• An error signal is incorporated to assess the validity of the montage conversion process.
• The algorithm identifies output channels that are not obtainable from the specified input montage.

Main Results:

• The developed algorithm successfully converts between different EEG montages via linear transformation.
• Testing with a system retrieving digital EEG data from videotape showed good agreement between converted and directly calculated montages.
• Referential and Laplacian data derived from bipolar output matched montages calculated directly from referential output.
• The algorithm effectively identified output channels that could not be derived from the input.

Conclusions:

• The proposed algorithm offers a general and accurate method for EEG montage conversion.
• This tool enhances the flexibility of EEG data analysis by enabling transformation between various montages.
• The error signal and channel identification features improve the reliability and interpretability of converted EEG data.