June 2021
Volume 62, Issue 8
Open Access
ARVO Annual Meeting Abstract  |   June 2021
Physiology-enhanced data analytics to evaluate the role of ocular hemodynamics in glaucoma progression
Author Affiliations & Notes
  • Omar Ibrahim
    Electrical Engineering and Computer Science, University of Missouri, Columbia, Missouri, United States
  • Alon Harris
    Ophthalmology, Icahn School of Medicine at Mount Sinai, New York, New York, United States
  • Daphne Zou
    Electrical Engineering and Computer Science, University of Missouri, Columbia, Missouri, United States
  • Nicholas Mattia Marazzi
    Electrical Engineering and Computer Science, University of Missouri, Columbia, Missouri, United States
  • James Keller
    Electrical Engineering and Computer Science, University of Missouri, Columbia, Missouri, United States
  • Brent A Siesky
    Ophthalmology, Icahn School of Medicine at Mount Sinai, New York, New York, United States
  • Alice Chandra Verticchio Vercellin
    Ophthalmology, Icahn School of Medicine at Mount Sinai, New York, New York, United States
    Universita degli Studi di Pavia, Pavia, Lombardia, Italy
  • Ryan Zukerman
    Ophthalmology, Icahn School of Medicine at Mount Sinai, New York, New York, United States
  • Giovanna Guidoboni
    Electrical Engineering and Computer Science, University of Missouri, Columbia, Missouri, United States
  • Footnotes
    Commercial Relationships   Omar Ibrahim, None; Alon Harris, AdOM (C), AdOM (S), AdOM (I), Luseed (C), Luseed (I), Oxymap (I), Phileas Pharma (S), Phileas Pharma (I), Qlaris (C), Qlaris (S), Qlaris (I), QuLent (I); Daphne Zou, None; Nicholas Marazzi, Duned (E); James Keller, None; Brent Siesky, None; Alice Chandra Verticchio Vercellin, None; Ryan Zukerman, None; Giovanna Guidoboni, Foresite Healthcare LLC (C), Gspace LLC (I)
  • Footnotes
    Support  NSF-DMS (1853222/2021192)
Investigative Ophthalmology & Visual Science June 2021, Vol.62, 2578. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Omar Ibrahim, Alon Harris, Daphne Zou, Nicholas Mattia Marazzi, James Keller, Brent A Siesky, Alice Chandra Verticchio Vercellin, Ryan Zukerman, Giovanna Guidoboni; Physiology-enhanced data analytics to evaluate the role of ocular hemodynamics in glaucoma progression. Invest. Ophthalmol. Vis. Sci. 2021;62(8):2578.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : Many studies have identified ocular hemodynamics as relevant in glaucoma; however the specific link between vascular biomarkers, blood pressure (BP), intraocular pressure (IOP), and the progression of open angle glaucoma (OAG) remains uncertain. Herein we address this issue via a novel physiology-enhanced approach to analyze data from the Indianapolis Glaucoma Progression Study.

Methods : 115 OAG patients were evaluated twice yearly over 5 years. Functional and structural OAG progression was monitored by visual field testing, optical coherence tomography and Heidelberg retinal tomography. A validated model of the retinal circulation (Guidoboni et al 2014) was used to simulate 8 individualized hemodynamic outputs (Fig.1) based on 4 inputs (SBP, DBP, IOP, HR) measured on each patient. In the retinal model, changes in IOP and BP induce changes in transmural pressures and, therefore, vascular resistances. The model is capable of predicting hemodynamic variables of physiological interest that cannot be measured directly, thereby producing a physiology-enhanced dataset with 12 instead of 4 variables for each patient and visit. The 12 baseline variables for each patient were then fed into an unsupervised machine learning model, Fuzzy c-means (FCM) clustering algorithm to identify patient clusters with similar ocular hemodynamics. The relationship between clusters and OAG progression was analyzed statistically with p-values<0.05 considered statistically significant.

Results : The FCM clustering algorithm generated 4 clusters based on the 12 variables and were plotted based on IOP and mean arterial pressure (MAP = 2/3*DBP+1/3*SBP), Fig. 2. A statistically significant difference was found between the occurrence of OAG progression in Cluster 4 (n=35; 20 progression, 15 no progression) when compared to the other clusters (n=80; 62 progression, 18 no progression), p=0.026. Further, HR and Rv in Cluster 4 were found to be markedly lower than the other clusters (Fig.2).

Conclusions : Our analysis suggests that physiology-enhanced data analytics based on mathematical modeling and unsupervised machine learning of ocular hemodynamics may be effective in glaucoma risk stratification. The integration of modeling, machine learning and the clinical outcomes of glaucoma patients has the potential to improve OAG screening and treatment paradigms.

This is a 2021 ARVO Annual Meeting abstract.

 

Table of abbreviations

Table of abbreviations

 

Cluster properties in MAP-IOP (a), HR (b), Rv (c)

Cluster properties in MAP-IOP (a), HR (b), Rv (c)

×
×

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.

×