Quantitative performance assessments for neuromagnetic imaging systems

We have developed a Monte-Carlo simulation method to assess the performance of neuromagnetic imaging systems using two kinds of performance metrics: A-prime metric and spatial resolution. We compute these performance metrics for virtual sensor systems having 80, 160, 320, and 640 sensors, and discuss how the system performance is improved, depending on the number of sensors. We also compute these metrics for existing whole-head MEG systems, MEGvision™ (Yokogawa Electric Corporation, Tokyo, Japan) that uses axial-gradiometer sensors, and TRIUX™ (Elekta Corporate, Stockholm, Sweden) that uses planar-gradiometer and magnetometer sensors. We discuss performance comparisons between these significantly different systems.