EEG-fMRI: Ballistocardiogram Artifact Reduction by Surrogate Method for Improved Source Localization

• DOI: 10.3389/fnins.2022.842420
• Publication Date: 2022-03-10T12:56:09Z
• Authors: Mateusz Rusiniak; Harald Bornfleth; Jae-Hyun Cho; Tomasz Wolak; Nicole Ille; Patrick Berg; Michael Scherg
• Keywords: Artifact (error); Distortion (music); SIGNAL (programming language); Surrogate data; Source Separation

For the analysis of simultaneous EEG-fMRI recordings, it is vital to use effective artifact removal tools. This applies in particular ballistocardiogram (BCG) which difficult remove without distorting signals interest related brain activity. Here, we documented surrogate source models separate artifact-related from with minimal distortion activity interest. The topographies used for separation were created automatically using principal components (PCA-S) or by manual selection utilizing independent (ICA-S). Using real resting-state data 55 subjects superimposed simulated auditory evoked potentials (AEP), both approaches compared three established BCG methods: Blind Source Separation (BSS), Optimal Basis Set (OBS), and a mixture (OBS-ICA). Each method was evaluated its applicability ERP following criteria: number events surviving threshold scans, signal-to-noise ratio (SNR), error localization, signal variance explained dipolar model. these criteria, PCA-S ICA-S fared best overall, highly significant differences methods, especially localization. approach also applied single subject Berger experiment performed MRI scanner. Overall, artifacts methods provides substantial improvement methods.