Сахарный диабет (Mar 2024)

Automated analysis of retinal microcirculation in type 1 diabetes mellitus

  • Yu. N. Yusef,
  • M. H. Durzhinskaya,
  • V. G. Pavlov,
  • D. V. Petrachkov,
  • I. B. Gurevich,
  • V. V. Yashina,
  • A. T. Tleubaev,
  • V. V. Fadeyev,
  • I. V. Poluboyarinova,
  • A. E. Goldsmid,
  • R. А. Karamullina,
  • D. V. Lipatov,
  • M. V. Budzinskaya

DOI
https://doi.org/10.14341/DM12931
Journal volume & issue
Vol. 27, no. 1
pp. 41 – 49

Abstract

Read online

BACKGROUND: The paper is dedicated to the assessment of the retinal microvasculature in patients with type 1 diabetes mellitus (DM) with various features of the clinical course and different stages of diabetic retinopathy (DR). Automatic analysis of optical coherence tomogram angiograms (OCT-A) was carried out with specially developed software that provides the ability to estimate quantitative vascular parameters.AIM: The purpose of the study was to assess diagnostic accuracy of clinical parameters and imaging biomarkers in type 1 diabetes using a new algorithm for OCT-A analysis.MATERIALS AND METHODS: The study involved 186 people (365 eyes) with type 1 diabetes. The analysis of the OCT-A parameters was performed with a specially developed software. The range of studied parameters included: foveal avascular zone (FAZ), vessel area density (VAD), skeletonized vessel density (VSD), vessel diameter index (VDI), vascular curvature index (VCI) at the level of superficial (SCP) and deep (DCP) retinal capillary plexuses in the macular region. A correlation between the involvement of OCT-A biomarkers and age, degree of DM, increased glycated hemoglobin (HbA1c) level, stage of DR, and maximally corrected visual acuity (BCVA) was analysed.RESULTS: A significant dependence of all quantitative OCT-A parameters on the age of and duration of diabetes (p<0.05) was revealed. An increase in FAZ SCP (K=0.788, p=0) and DCP (K=0.764, p=0.03); decrease in VAD SCP (K=-0.476, p=0) and DCP (K=-0.485, p=0); VSD SCP (K=0.692, p=0) and DCP (K=0.713, p=0); an increase in VDI SCP (K=0.698, p=0) and DCP (K=787, p<0.01), as well as an increase in the VCI SCP (K=0.735, p=0) and DCP (K=0.694, p p=0). An inverse relationship was found between HbA1c level and VAD SCP (K=-0.636, p=0) and DCP (K=-0.619, p=0.05) were identified as well as a direct relationship with VDI DCP (K=0.717, p<0.05). The influence of the HbA1c level on other parameters was not confirmed (p>0.05). The presence of correlation between BCVA and FAZ DCP (K=-0.728, p=0), as well as VSD DCP (K=-0.754, p=0) was proved.CONCLUSION: As a result of a comprehensive analysis of clinical data and imaging biomarkers, a number of patterns that have diagnostic value in diabetic retinopathy were identified.

Keywords