July 2018
Volume 59, Issue 9
Open Access
ARVO Annual Meeting Abstract  |   July 2018
Imaging the human iris: a hyperspectral approach
Author Affiliations & Notes
  • Luca Di Cecilia
    Dept. of Engineering "E. Ferrari", University of Modena and Reggio Emilia, Modena, Modena, Italy
  • Francesco Marazzi
    Dept. of Engineering "E. Ferrari", University of Modena and Reggio Emilia, Modena, Modena, Italy
  • Luigi Rovati
    Dept. of Engineering "E. Ferrari", University of Modena and Reggio Emilia, Modena, Modena, Italy
  • Footnotes
    Commercial Relationships   Luca Di Cecilia, None; Francesco Marazzi, None; Luigi Rovati, None
  • Footnotes
    Support  None
Investigative Ophthalmology & Visual Science July 2018, Vol.59, 5866. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Luca Di Cecilia, Francesco Marazzi, Luigi Rovati; Imaging the human iris: a hyperspectral approach. Invest. Ophthalmol. Vis. Sci. 2018;59(9):5866.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose : Hyperspectral (HS) imaging is a promising optical technique that allows the detection of both spatial and spectral information in a single data acquisition. Here we investigate the capability of a prototype HS instrument to measure the iris spectral reflectance in vivo. Moreover, we evaluate the intra- and inter-session repeatability of quantitative HS measurements of the human iris.

Methods : Eight healthy volunteers (all males, average age 27 ± 2 years) participated in the study. Irises colors were classified using the grading system of Franssen et al [Grading of Iris Color with an Extended Photographic Reference Set, J Optom 2008; 1:36-40]. Each HS acquisition (duration: 4125 ms) consists of 22 images (HS cube) captured between 480 – 900 nm, in 20nm intervals. Each acquisition was repeated 6 times. The instrument was calibrated using a NIST white reflectance standard, to ensure quantitative and repeatable measurements. Reflectance was analyzed across the 22 spectral bands in an annular region of interest concentric to the pupil. Repeatability was assessed from images captured in 3 sessions, at the same exposure time, with realignment of the iris and refocusing between each session.

Results : Inter-subject variability was highest at short wavelengths (Fig. 1) and reflectance exhibited opposite trends in the visible and near infrared. This might be the result of melanin scattering. The influence of water absorption is detected above 850 nm. Repeatability was highest for lighter irises and for wavelengths from 520 to 900 nm. In that range, the coefficient of repeatability (95% of confidence interval) for measurements between sessions was ±10% and ±5% for dark and light irises, respectively. The within-session repeatability was ±7.5% and ±4%, respectively.

Conclusions : Automated in vivo HS imaging of the human iris appears to be reliable and reproducible. This technique could potentially be exploited for monitoring iris pigmentation changes with time induced by prostaglandin analogues. In addition, the iris spectral reflectance could be correlated with the development of certain ocular pathologies.

This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.

 

Figure 1: Calculated spectral reflectances for the eight volunteers’ irises. Iris color is reported according to Franssen scale. Figure 1.B shows the region of interest where the spectral reflectance is calculated (at λ = 580nm) for a volunteer classified as 13.

Figure 1: Calculated spectral reflectances for the eight volunteers’ irises. Iris color is reported according to Franssen scale. Figure 1.B shows the region of interest where the spectral reflectance is calculated (at λ = 580nm) for a volunteer classified as 13.

×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×