Evaluation of a retinal oximetry image quality indicator in patients with cataract

Oliver Niels Klefter*, Jesper Høiberg Erichsen, Mathias Møller Hansen, Lars Morten Holm, Sveinn Hakon Hardarson, Einar Stefánsson, Line Kessel

*Corresponding author for this work
2 Citations (Scopus)


PURPOSE: To evaluate a new automated retinal oximetry image quality indicator with cataract as a clinical model.

METHODS: Sixty-one eyes in 61 patients were imaged by the Oxymap T1 Retinal Oximeter at baseline and 25 eyes were also examined 3 weeks after cataract surgery. Image quality (0-10 on a continuous scale) was compared with standardized AREDS cataract grading and Pentacam lens densitometry. Associations with retinal oximetry measurements and visual acuity were examined.

RESULTS: Image quality correlated with total, nuclear and posterior subcapsular cataract grades (ANOVA, p < 0.05), tended to be associated with lens densitometry and it improved from 4.3 ± 1.4 to 5.7 ± 1.0 (p < 0.05) after cataract surgery. Very low image quality, below 3, led to vessel detection failure in retinal oximetry images. Higher image qualities were linearly associated with higher measured retinal oxygen saturations (r = 0.52 in arteries and r = 0.46 in veins; p < 0.001).

CONCLUSION: Retinal oximetry image quality deteriorated with increasing cataract density and improved after cataract surgery, supporting its use as a measure of optical clarity. The numerical quality indicator demonstrated a threshold below which images of poor optical quality should be discarded. Image quality affects the estimates of retinal oximetry parameters and should therefore be included in future analyses.

Original languageEnglish
JournalActa Ophthalmologica (Online)
Issue number3
Pages (from-to)312-317
Number of pages6
Publication statusPublished - 2024


  • cataract
  • image quality
  • retina
  • retinal imaging
  • retinal oximetry
  • Oxygen
  • Retinal Vessels
  • Cataract/diagnosis
  • Humans
  • Quality Indicators, Health Care
  • Oximetry/methods


Dive into the research topics of 'Evaluation of a retinal oximetry image quality indicator in patients with cataract'. Together they form a unique fingerprint.

Cite this