Optimized sampling strategy for 3D magnetic field measurement

Abstract

Recently, along with non-pharmacological brain stimulation-based disease treatment research that stimulates neural cells in the brain without using drugs, there is a significant amount of research aimed at implementing sensation in actual physical extremities through stimulation of the brain's motor and sensory areas. A representative example of non-invasive brain stimulation devices is Transcranial Magnetic Stimulation (TMS) is a representative non-invasive device, but precise application to the desired brain location poses challenges. Accurate understanding of the magnetic field distribution is crucial for optimal use of these devices in terms of intensity and depth of penetration. However, data sets of magnetic fields generated by TMS are generally not available, and data sets of some products developed in previous studies are accessible to only a few. The simplest and most powerful method commonly known is to cover all areas by linearly scanning the entire area. However, these methods are time-consuming and place a time burden on the development and research of various stimulators. In this research, to address these issues, a new BPP algorithm is proposed by combining Breadth-first search, Path planning, and Pruning techniques, and naming it based on the initials of each technique. The purpose is to find an optimal algorithm that maintains the quality of the result while completing the measurement quickly compared to the previously proposed SSAS(Sequential Spatial Adaptive Sampling) algorithm and full-range linear scanning method. Experiments comparing BPP and SSAS algorithms with full-range linear scanning methods showed that both methods maintained good quality with Pearson correlation coefficients above 0.9. The SSAS algorithm took one-third of the time compared to the full-range scanning, while the BPP algorithm took one-eleven of the time, confirming the time efficiency of the BPP algorithm. As for the Pearson correlation coefficient for the entire area, the SSAS algorithm showed slightly better results, but the BPP algorithm showed slightly better results in the ROI area, which means the center of the TMS. The proposed BPP algorithm is expected to have applicability not only in verifying the magnetic field distribution of TMS but also in various fields such as detecting abnormal magnetic field generation in specific components due to overcurrent or confirming defects in TMS products.