Neural Noise

fNIRS signal quality and hair/skin characteristics for inclusivity

2 min read

  • The impact of hair and skin characteristics on fNIRS signal quality has been quantified to support enhanced inclusivity.1
  • The methods improve deployment in diverse populations and are directly applicable to BCI and neuroimaging pipelines.1 1

Gardner updates

  • A Nature study quantified the impact of hair and skin characteristics on fNIRS signal quality for enhanced inclusivity and deployment in diverse populations. 1

Weekly enrichment (2026-07-20)

  • The source study (Yücel et al., Nature Human Behaviour 2025, 9:2651–2668) quantified how hair and skin characteristics, head size, sex, and age affect fNIRS signal quality in n = 115 participants.2
  • Measurements used two continuous-wave NIRSport2 systems (NIRx) with 16 sources and 16 detectors plus short-separation channels, recording at 760 and 850 nm at a 10.2 Hz sampling rate; long-separation channels were ~30 mm and short-separation ~8 mm.2
  • Skin pigmentation was measured objectively as a Melanin Index via reflectance spectrophotometry (0–99 scale), and hair was characterized with high-resolution trichoscopy (shaft thickness range 42–88 μm, follicular-unit counts, and FIA texture from fine to coarse).2
  • Signal quality (Corrected Signal Mean and Scalp Coupling Index) showed significant negative Spearman correlations with hair shaft thickness (ρ ≈ −0.30) and skin pigmentation (ρ ≈ −0.35, p < 0.01) after Benjamini–Hochberg correction; higher pigmentation corresponded to a roughly 11-fold reduction in corrected signal, and coarser hair to a large negative fold change.2
  • The authors provide actionable recommendations: a standardized metadata table for reporting hair/skin phenotype, plus cap/optode configuration, hair-management, and data-collection strategies to reduce bias across diverse participants.2
  • Context: perspective work frames poor fNIRS/EEG contact for coarse, curly, or protective hairstyles and darker skin as “structural racism in neuroimaging,” disproportionately excluding Black participants.3
  • Context: a review of 87 fNIRS papers (2017–2022) found only ~2% reported race/ethnicity, 0% reported skin tone, and 3% reported hair type, versus >90% reporting gender—evidence of systematic under-reporting.4
  • Context: hardware fixes such as brush optodes that thread through hair raised study success rates to nearly 100%, cut setup time roughly threefold, and improved activation SNR by up to 10× in participants with dense hair.5
  • Context: best-practice work on Afro-textured hair reports that dedicated capping techniques improved signal quality by about 50% on average, with the largest gains in anterior channels.6

Footnotes

  1. https://news.google.com/rss/articles/CBMiX0FVX3lxTE1DZXQtZENYU213VVN0VmJIUVhEQzVWQ2w5R0J4b3h4blc5WXdPcjhSZFlUWXZ2WF9wc0VnWHBkY1Njam9mSVJMblZyRGZxbGJELTFpNHJKcy1pYVo3RVlB?oc=5 ↩ ↩2 ↩3 ↩4

  2. https://www.nature.com/articles/s41562-025-02274-7 ↩ ↩2 ↩3 ↩4 ↩5

  3. https://doi.org/10.1016/s2215-0366(22)00079-7 ↩

  4. https://doi.org/10.3389/fnins.2023.1086208 ↩

  5. https://doi.org/10.1364/boe.3.000878 ↩

  6. https://doi.org/10.1002/dev.70134 ↩

Graph View