biostudies-literatureBezsudnova YWellcome TrustBiotechnology and Biological Sciences Research CouncilEngineering and Physical Sciences Research Council119747https://www.ebi.ac.uk/biostudies/studies/S-EPMC7615061biostudies-literatureUnknown264Magnetoencephalography (MEG) based on optically pumped magnetometers (OPMs) has been hailed as the future of electrophysiological recordings from the human brain. In this work, we investigate how the dimensions of the sensing volume (the vapour cell) affect the performance of both a single OPM-MEG sensor and a multi-sensor OPM-MEG system. We consider a realistic noise model that accounts for background brain activity and residual noise. By using source reconstruction metrics such as localization accuracy and time-course reconstruction accuracy, we demonstrate that the best overall sensitivity and reconstruction accuracy are achieved with cells that are significantly longer and wider that those of the majority of current commercial OPM sensors. Our work provides useful tools to optimise theNeuroImageOptimising the sensing volume of OPM sensors for MEG source reconstruction.PMC7615061EP/T001046/1BB/R018723/1207550207550/Z/17/ZEP/W028050/1Koponen LMBezsudnova YBarontini GKowalczyk AUJensen OfalseOptimising the sensing volume of OPM sensors for MEG source reconstruction.Magnetoencephalography (MEG) based on optically pumped magnetometers (OPMs) has been hailed as the future of electrophysiological recordings from the human brain. In this work, we investigate how the dimensions of the sensing volume (the vapour cell) affect the performance of both a single OPM-MEG sensor and a multi-sensor OPM-MEG system. We consider a realistic noise model that accounts for background brain activity and residual noise. By using source reconstruction metrics such as localization accuracy and time-course reconstruction accuracy, we demonstrate that the best overall sensitivity and reconstruction accuracy are achieved with cells that are significantly longer and wider that those of the majority of current commercial OPM sensors. Our work provides useful tools to optimise the2022-01-01T00:00:00Z2022 Dec2025-04-04T02:25:43.268Z2025-04-04T02:25:43.268ZS-EPMC76150613640373310.1016/j.neuroimage.2022.119747