Developing customized NIRS-EEG for infant sleep research: methodological considerations

Louisa K Gossé; Paola Pinti; Frank Wiesemann; Clare E Elwell; Emily J H Jones

doi:10.1117/1.NPh.10.3.035010

Developing customized NIRS-EEG for infant sleep research: methodological considerations

Neurophotonics. 2023 Jul;10(3):035010. doi: 10.1117/1.NPh.10.3.035010. Epub 2023 Sep 25.

Authors

Louisa K Gossé¹, Paola Pinti¹, Frank Wiesemann², Clare E Elwell³, Emily J H Jones¹

Affiliations

¹ Birkbeck, University of London, Centre for Brain and Cognitive Development, London, United Kingdom.
² Research and Development, Procter & Gamble, Schwalbach am Taunus, Germany.
³ University College London, Department of Medical Physics and Biomedical Engineering, Biomedical Optics Research Laboratory, London, United Kingdom.

Abstract

Significance: Studies using simultaneous functional near-infrared spectroscopy (fNIRS)-electroencephalography (EEG) during natural sleep in infancy are rare. Developments for combined fNIRS-EEG for sleep research that ensure optimal comfort as well as good coupling and data quality are needed.

Aim: We describe the steps toward developing a comfortable, wearable NIRS-EEG headgear adapted specifically for sleeping infants ages 5 to 9 months and present the experimental procedures and data quality to conduct infant sleep research using combined fNIRS-EEG.

Approach: $N = 49$ 5- to 9-month-old infants participated. In phase 1, $N = 26$ (10 = slept) participated using the non-wearable version of the NIRS-EEG headgear with 13-channel-wearable EEG and 39-channel fiber-based NIRS. In phase 2, $N = 23$ infants (21 = slept) participated with the wireless version of the headgear with 20-channel-wearable EEG and 47-channel wearable NIRS. We used QT-NIRS to assess the NIRS data quality based on the good time window percentage, included channels, nap duration, and valid EEG percentage.

Results: The infant nap rate during phase 1 was $\sim 40 %$ (45% valid EEG data) and increased to 90% during phase 2 (100% valid EEG data). Infants slept significantly longer with the wearable system than the non-wearable system. However, there were more included good channels based on QT-NIRS in study phase 1 (61%) than phase 2 (50%), though this difference was not statistically significant.

Conclusions: We demonstrated the usability of an integrated NIRS-EEG headgear during natural infant sleep with both non-wearable and wearable NIRS systems. The wearable NIRS-EEG headgear represents a good compromise between data quality, opportunities of applications (home visits and toddlers), and experiment success (infants' comfort, longer sleep duration, and opportunities for caregiver-child interaction).

Keywords: EEG; data quality; functional near-infrared spectroscopy; infant; sleep; wearable neuroimaging.