Spatial accuracy evaluation of magnetic source imaging methods on OPM-based MEG

Nan An; Fuzhi Cao; Wen Li; Wenli Wang; Weinan Xu; Chunhui Wang; Min Xiang; Yang Gao; Binbin Sui; Dawei Wang; Dexin Yu; Xiaolin Ning

doi:10.1016/j.isci.2022.105177

Spatial accuracy evaluation of magnetic source imaging methods on OPM-based MEG

iScience. 2022 Sep 22;25(10):105177. doi: 10.1016/j.isci.2022.105177. eCollection 2022 Oct 21.

Authors

Nan An^{1

2}, Fuzhi Cao^{1

2}, Wen Li^{1

2}, Wenli Wang^{1

2}, Weinan Xu^{1

2}, Chunhui Wang^{1

2}, Min Xiang^{1

3}, Yang Gao^{1

4}, Binbin Sui⁵, Dawei Wang^{6

7

8}, Dexin Yu^{6

7

8}, Xiaolin Ning^{1

3}

Affiliations

¹ Zhejiang Provincial Key Laboratory of Ultra-Weak Magnetic-Field Space and Applied Technology, Hangzhou Innovation Institute, Beihang University, Hangzhou 310051, China.
² Key Laboratory of Ultra-Weak Magnetic Field Measurement Technology, Ministry of Education, School of Instrumentation and Optoelectronic Engineering, Beihang University, Beijing 100191, China.
³ Research Institute of Frontier Science, Beihang University, Beijing 100191, China.
⁴ School of Physics, Beihang University, Beijing 100191, China.
⁵ Beijing Tiantan Hospital, Capital Medical University, Beijing 100050, China.
⁶ Department of Radiology, Qilu Hospital of Shandong University, Jinan 250012, China.
⁷ Shandong Key Laboratory: Magnetic Field-free Medicine and Functional Imaging, Jinan 250012, China.
⁸ Research Institute of Shandong University: Magnetic Field-free Medicine and Functional Imaging, Jinan 250012, China.

Abstract

The emergence of the optically pumped magnetometer (OPM)-based magnetoencephalography (MEG) has led to new developments in MEG technology. The source imaging results of different magnetic source imaging (MSI) methods show considerable differences, which makes it difficult for researchers to choose an appropriate method. This study assessed time-domain MSI methods implemented in the Brainstorm, FieldTrip, and SPM12 toolboxes using simulations. We proposed using a metric, variational free energy under the Bayesian framework, as an indicator to evaluate source imaging results because it does not require the ground truth of sources but uses the fitness of the measurement data. Our simulations demonstrated the effectiveness of the variational free energy in indicating the quality of the source reconstruction results. We then applied each MSI method to the real OPM-MEG experimental data. We aimed to highlight the characteristics of each method and provide references for researchers choosing an appropriate MSI method.

Keywords: Biological sciences research methodologies; Biology experimental methods; Biotechnology; Techniques in neuroscience.