Deep brain stimulation device-specific artefacts in MEG recordings

Background Deep brain stimulation (DBS) has strong beneficial effects for treating movement disorders. The related cortical mechanisms can be studied with magnetoencephalography (MEG) during active DBS. However, MEG is prone to artefacts induced by the electrical stimulation and the movement of ferromagnetic DBS components. Although artefacts might vary between DBS devices from different manufacturers, no such comparison has been performed. To date, no combined MEG-DBS studies have been conducted within Yokogawa MEG systems. Objective The aim of the present study was to compare DBS artefacts in MEG phantom recordings acquired with two MEG systems (Neuromag, Yokogawa) using DBS devices from three different manufacturers (Abbott, Boston Scientifc, Medtronic) and to test whether established cleaning methods can sufficiently reduce artefacts. Methods DBS devices, electrodes, and extension cables were attached to a gelatine MEG phantom, and data was acquired in a Neuromag and a Yokogawa MEG system. Results There are device-specific differences in movement-related artefacts with weaker artefacts for Boston Scientific (BSC) devices. Stimulation-, movement- and IPG-related artefacts are best cleaned when combining the ICA-MI and Hampel filter across recordings and DBS devices. However, the cleaning of movement-related artefacts can result in artefactual spectral peaks in physiologically relevant frequencies below 20 Hz. Conclusions Device-specific IPG-related artefacts have to be considered for MEG-DBS studies and can be cleaned with combinations of published cleaning methods for MEG and EEG data. Critically, cleaning movement-related artefacts can potentially result in spectral peaks, which resemble physiological activity. Finally, combined MEG-DBS recordings are feasible in Yokogawa-MEG systems.