Back

The Mismatch Negativity compared: EEG, SQUID-MEG and novel 4Helium-OPMs

Gutteling, T. P.; Mattout, J.; Daligault, S.; Jung, J.; Labyt, E.; Schwartz, D.; Lecaignard, F.

2025-04-05 neuroscience
10.1101/2025.04.04.647154 bioRxiv
Show abstract

Magneto-encephalography (MEG) provides a higher spatial resolution than electro-encephalography (EEG) to measure human auditory responses. However, conventional cryogenic MEG systems (SQUID-MEG) suffer from severe technological restrictions limiting, for instance, routine clinical use. Fortunately, a new generation of MEG sensors, optically pumped magnetometers (OPMs), have been developed to bridge the gap, combining the wearability of EEG with the benefits of MEG signal acquisition. We aim to assess their potential for studying auditory mismatch processing. The auditory Mismatch Negativity (MMN) is a well-characterized evoked component observable using a passive oddball paradigm with two-tone sound sequences. It has been extensively described using both EEG and MEG and is part of many EEG-based clinical applications, such as the assessment of patients with disorders of consciousness. MMN is therefore a relevant candidate to evaluate OPM performance. We use recently developed Helium-OPMs, which are high dynamic range MEG sensors that operate at room temperature. We compare their performance with cryogenic SQUID-MEG and EEG in a passive frequency oddball paradigm. Results show a significant MMN across subjects in all modalities as well as a high temporal similarity between modalities. Signal-to-noise ratios were also similar, and detection of significant individual MMN (within-subjects) using the OPM system was equal or better than EEG. Given that the OPM system tested here is a prototype comprised of only five sensors, these results are a promising step towards wearable MEG that combines the advantages of MEG and EEG.

Matching journals

The top 2 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.