Abstract

The surgical procedure or Deep Brain Stimulation (DBS) is an established technique aimed at neuromodulation and is most often used for treating neurological and neuropsychiatric disorders. In clinical studies, Magnetoencephalography (MEG) is a widely used technique for characterizing the effects of Deep Brain Stimulation (DBS). A significant limitation of DBS stimulation is the inability to distinguish stimulation artifacts from actual neuronal activity, especially in the lower frequencies, where it may obscure the biological response and be a confounding factor. This paper aims to understand how to circumvent the DBS artifactual component in the summary metrics of DBS-corrupted MEG data. To do the same, we used a watermelon as a phantom model of the head to deploy DBS electrodes. The spectral signature of the artifactual component of DBS and it's power in different clinically significant bands of interest were analyzed in the frequency domain. We present the preliminary results for the same in this paper.