September 2016
Volume 57, Issue 12
Open Access
ARVO Annual Meeting Abstract  |   September 2016
Hyperspectral snapshot camera for functional imaging of the retina
Author Affiliations & Notes
  • Jason Dwight
    Bioengineering, Rice University, Houston, Texas, United States
  • Robert Earl Coffee
    Ophthalmology, Baylor College of Medicine, Houston, Texas, United States
  • Michal E. Pawlowski
    Bioengineering, Rice University, Houston, Texas, United States
  • Christina Y. Weng
    Ophthalmology, Baylor College of Medicine, Houston, Texas, United States
  • Tomasz Tkaczyk
    Bioengineering, Rice University, Houston, Texas, United States
  • Footnotes
    Commercial Relationships   Jason Dwight, None; Robert Earl Coffee, None; Michal Pawlowski, None; Christina Weng, None; Tomasz Tkaczyk, None
  • Footnotes
    Support  None
Investigative Ophthalmology & Visual Science September 2016, Vol.57, 3744. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Jason Dwight, Robert Earl Coffee, Michal E. Pawlowski, Christina Y. Weng, Tomasz Tkaczyk; Hyperspectral snapshot camera for functional imaging of the retina. Invest. Ophthalmol. Vis. Sci. 2016;57(12):3744.

      Download citation file:

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

  • Supplements

Purpose : Two wavelength-based systems have previously been used to estimate oxygen metabolism in retinal diseases. The technology requires extensive calibration procedures to provide accurate results. We present the Image Mapping Spectrometer (IMS) which utilizes 40 spectral channels and is based on the absorbance spectrum’s shape instead of two wavelength absorption ratio. As a result we are able to reduce the need for calibration prior to oximetry. Using IMS, we performed retinal oximetry in patients with age-related macular degeneration (AMD) and compared results with those of a two wavelength approach.

Methods : The IMS is a snapshot hyperspectral imager capable of acquiring a 350x350x40 (x,y,l) datacube in a single exposure time with no scanning (Fig. 1). The system was attached to a top port of Topcon TRC-50ex fundus camera. Oxygenation of a vessel was calculated based on absorbance data measured at 40 spectral channels. Oximetry was performed on the vein and artery for each subject. For comparison we also computed a two wavelength oximetry using data extracted from the hyperspectral cube acquired by the IMS.

Results : Using the 40 spectral channel method, the average oxygen saturation was estimated. Fig. 2 demonstrates the LSQ fit absorbance curves of the artery and vein corresponding to related oxygenation values. Values for the 3 patients in the artery and vein were similar, averaging 96% (SD = 7%) and 45% (SD=9%), respectively (Fig. 2a, 2b). Note that for all three patients, data was obtained from similar locations in the retina. The oximetry map of a patient with AMD is shown in Fig. 2c.

Conclusions : The use of IMS is a promising new approach for real time retinal oximetry. We obtained physiologically relevant data consistent with previous publications. The advantage of hyperspectral oximetry is that it is based on spectrum shape and does not require the extensive calibration procedures necessary for two wavelength oximetry. The spectral shapes related to oxygenation allow automatic real-time acquisition of self-calibrated oxygenation maps. In comparing the hyperspectral method and two wavelength method, we verified that the latter provides unrealistic values (130-140% oxygenation) if applied without prior experimental calibration and is incapable of distinguishing between veins and arteries.

This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.




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.