Talks and presentations
This page collects talks, posters and presentations given by the NeuroWaves MEG Laboratory and the Bio-Medical Imaging Core at NYU Abu Dhabi.
SAMBA × MEG 2026 - Conference Poster
Benchmarking SQUID and OPM Magnetoencephalography with a Dry Phantom: Source Localization Accuracy and Its Dependence on Sensor Count
- Authors:
Hadi Zaatiti 1, Daisuke Oyama 2, Yoshiaki Adachi 2
- Affiliations:
1 Bio-Medical Imaging Core & NeuroWaves MEG Laboratory, New York University Abu Dhabi, UAE · 2 Applied Electronics Laboratory, Kanazawa Institute of Technology, Japan
- Presenting author:
Conference details
Abstract
Magnetoencephalography (MEG) noninvasively records the magnetic fields of neuronal activity using either established cryogenic SQUID sensors or emerging optically pumped magnetometers (OPMs), which need no cooling and can sit closer to the scalp. Human comparisons are confounded by physiology, so we benchmarked both technologies on a dry phantom emulating 50 equivalent current dipoles (ECDs), measured by both a SQUID (Kanazawa Institute of Technology) and an OPM system (HEDscan) in the same shielded room at NYU Abu Dhabi. Drive currents of 10 and 100 µA varied signal-to-noise ratio; sources were reconstructed with the Sarvas formula and scored against CT-calibrated ECD positions. Placing OPMs closer to the source gave a peak signal 3-4× larger than SQUID (6.6 vs 2.2 pT), yet OPM localization error was about three times larger: 3.48 ± 0.58 mm versus 1.07 ± 0.17 mm at 10 µA, and 2.77 versus 0.74 mm at 100 µA. Controlled analyses showed this gap was not explained by signal-to-noise ratio, channel count, or spatial coverage; a 90-channel SQUID subset still reached 1.56 mm, and only ~40 SQUID channels were needed to match the 90-channel OPM accuracy (3.13 mm), implicating OPM sensor-array calibration as the dominant factor. A dry phantom thus offers a reproducible, physiology-free benchmark for cross-technology comparison.
Keywords: magnetoencephalography; SQUID; optically pumped magnetometers; dry phantom; source localization; sensor array scalability
Poster (PDF)
The poster is written in Typst; the compiled A0 PDF is committed alongside the source in this repository.
Download the poster
Source files (Typst source, figures, and the reusable NYUAD poster template) live
in the repository under
docs/source/7-meg-class-talks-demos/talks/samba-meg-2026-poster/ and
docs/source/_templates/typst/.
Reference
Oyama, D.; Zaatiti, H. Phantom-Based Approach for Comparing Conventional and Optically Pumped Magnetometer Magnetoencephalography Systems. Sensors 2025, 25(7), 2063. https://doi.org/10.3390/s25072063