Automatic retinal oximetry

Area of Science:

• Ophthalmology
• Biomedical Engineering
• Medical Imaging

Background:

• Non-invasive measurement of hemoglobin oxygen saturation (SO(2)) in retinal vessels is crucial for diagnosing and monitoring various ocular and systemic diseases.
• Existing methods for retinal oximetry can be invasive or lack reproducibility and sensitivity.
• Developing an automated and reliable system for retinal oximetry is essential for clinical applications.

Purpose of the Study:

• To measure hemoglobin oxygen saturation (SO(2)) in retinal vessels using an automatic spectrophotometric oximeter.
• To evaluate the reproducibility and sensitivity of this novel oximetry device.
• To establish a reliable method for assessing in vivo retinal oxygenation.

Main Methods:

• Utilized specialized software to automatically identify retinal blood vessels in fundus images captured at four wavelengths.
• Calculated optical density ratios (ODRs) for each vessel and derived a linear model relating SO(2) to ODR.
• Assessed reproducibility through repeated measurements and tested sensitivity by inducing systemic hyperoxia in healthy volunteers (n=16) by altering inhaled oxygen concentration.

Main Results:

• The automated software significantly improved reproducibility, with mean standard deviations of 3.7% for arterioles and 5.3% for venules across five repeated measurements.
• The derived calibration model was SO(2) = 125 - 142 * ODR.
• Significant differences in SO(2) were observed between normoxia and hyperoxia in both arterioles (96% vs. 101%, P=0.0027) and venules (55% vs. 78%, P<0.0001).

Conclusions:

• The developed automatic retinal oximeter is a reliable and user-friendly tool.
• The device demonstrates high sensitivity to changes in hemoglobin oxygen saturation in response to altered inhaled oxygen concentrations.
• This technology holds promise for non-invasive assessment of retinal oxygenation in clinical settings.