Validation of Non-Invasive MEG Techniques for Detecting Hippocampal and Visual Cortex Responses to Entropy and Surprise

Understanding electrophysiological brain activity in response to cognitive tasks is a key focus in neuroscience. Despite advances in brain data acquisition technologies, invasive methods remain the gold-standard to accurately localise electrophysiological activity, particularly in deep brain structures. This study aims to provide preliminary results to validate superconducting quantum interference device (SQUID)-based magnetoencephalography (sMEG) and optically pumped magnetometer-based (OPM)-MEG against previously described gold-standard intracranial electroencephalograpy (iEEG) in detecting hippocampal and visual cortex responses to uncertain visual stimuli. Using two well-established source reconstruction techniques - dynamic Statistical Parametric Mapping (dSPM) and Linearly Constrained Minimum Variance (LCMV) beamformer - this study compared results from time-frequency analysis across different acquisition methods. While hippocampal time-resolved spectral activity detected via dSPM source reconstruction in sMEG data was consistent with previous iEEG findings, other approaches yielded inconsistent results compared to iEEG. These preliminary findings contribute to the understanding of non-invasive detection of cortical and subcortical activity and underscore the need for further methodological improvements in the analysis of non-invasive neurophysiological data to achieve reliable detection and interpretation of brain activity.