PLoS ONE (Jan 2020)

The effect of constitutive pigmentation on the measured emissivity of human skin.

  • Matthew Charlton,
  • Sophie A Stanley,
  • Zoë Whitman,
  • Victoria Wenn,
  • Timothy J Coats,
  • Mark Sims,
  • Jonathan P Thompson

DOI
https://doi.org/10.1371/journal.pone.0241843
Journal volume & issue
Vol. 15, no. 11
p. e0241843

Abstract

Read online

BackgroundThe measurement of body temperature has become commonplace in the current COVID-19 pandemic. Body temperature can be measured using thermal infrared imaging, a safe, non-contact method that relies on the emissivity of the skin being known to provide accurate readings. Skin pigmentation affects the absorption of visible light and enables us to see variations in skin colour. Pigmentation may also affect the absorption of infrared radiation and thus affect thermal imaging. Human skin has an accepted emissivity of 0.98 but the effect of different skin pigmentation on this value is not known. In this study, we investigated the influence of different skin pigmentation on thermal emissivity in 65 adult volunteers.MethodsA reference object of known emissivity (electrical tape) was applied to participant's skin on the inner upper arm. Tape and arm were imaged simultaneously using a thermal infrared camera. The emissivity was set on the camera to the known value for electrical tape. The emissivity was altered manually until the skin temperature using thermal imaging software was equal to the initial tape temperature. This provided the calculated emissivity value of the skin. Participants were grouped according to skin pigmentation, quantified using the Fitzpatrick skin phototyping scale and reflectance spectrophotometry. Differences in emissivity values between skin pigmentation groups were assessed by one-way ANOVA.ResultsThe mean calculated emissivity for the 65 participants was 0.972 (range 0.96-0.99). No significant differences in emissivity were observed between participants when grouped by skin pigmentation according to the Fitzpatrick scale (p = 0.859) or reflectance spectrophotometry (p = 0.346).ConclusionThese data suggest that skin pigmentation does not affect thermal emissivity measurement of skin temperature using thermal infrared imaging. This study will aid further research into the application of thermal infrared imaging as a screening or bedside diagnostic tool in clinical practice.