June 2023
Volume 64, Issue 8
Open Access
ARVO Annual Meeting Abstract  |   June 2023
Identifying pericyte cell loss in mouse models using automated image analysis
Author Affiliations & Notes
  • Cammi Valdez
    Department of Natural Sciences, Northeastern State University, Tahlequah, Oklahoma, United States
    Harold Hamm Diabetes Center, The University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma, United States
  • Anne Martin
    Department of Natural Sciences, Northeastern State University, Tahlequah, Oklahoma, United States
  • Madison Whitekiller
    Department of Natural Sciences, Northeastern State University, Tahlequah, Oklahoma, United States
  • Lauren Wilcox
    Department of Natural Sciences, Northeastern State University, Tahlequah, Oklahoma, United States
  • Dustin Baucom
    Department of Natural Sciences, Northeastern State University, Tahlequah, Oklahoma, United States
  • Joshua Butcher
    Department of Physiological Sciences in the College of Veterinary Medicine, Oklahoma State University, Stillwater, Oklahoma, United States
    Harold Hamm Diabetes Center, The University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma, United States
  • Footnotes
    Commercial Relationships   Cammi Valdez None; Anne Martin None; Madison Whitekiller None; Lauren Wilcox None; Dustin Baucom None; Joshua Butcher None
  • Footnotes
    Support  Genentech Career Development Award for Underrepresented Minority Emerging Vision Scientists (ARVO)
Investigative Ophthalmology & Visual Science June 2023, Vol.64, 1253. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Cammi Valdez, Anne Martin, Madison Whitekiller, Lauren Wilcox, Dustin Baucom, Joshua Butcher; Identifying pericyte cell loss in mouse models using automated image analysis. Invest. Ophthalmol. Vis. Sci. 2023;64(8):1253.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : In the normal retina there is tight regulation (1:1) of pericytes and endothelial cells (EC’s) in the capillaries, however in diabetic retinopathy (DR), the ratio shifts (1:4) due to pericyte loss. Due to ambiguity in pericyte and EC morphology in mice, it has been estimated that 20-30% of retinal capillary cells are indeterminate. We developed an automated image analysis program to identify and quantify pericytes and EC’s in the mouse retinal vasculature to increase reliability, reduce bias, and minimize analysis time.

Methods : Mouse retinas were elastase digested to isolate the retinal vasculature. The tissue was then stained with hematoxylin and Periodic Acid Schiff and imaged with light microscopy. Images were analyzed with our automated program.
Using CellProfiler, an open-source image analysis software, we created two pipelines with a series of modules to identify microvascular cells. Cells cutoff from view or overlapping were removed from analysis in our pipeline. Using Python, we developed an algorithm to remove any cells associated with vessels larger than capillaries. Manual analysis of individual cell images (n=2000) to categorize cells as either pericyte or EC was used for machine learning.
Manual and automated analysis were compared for time and cell identification and quantification.

Results : Image analysis time (n=50) was significantly decreased (p=0.007) with our automated program (33.3 minutes) compared to manual analysis (426.3 minutes, n=3). Pipeline 1 had no significant difference in EC identification and quantification compared to manual analysis, however there was a significant difference in pericyte detection (n=20, p=0.018). Pipeline 2 had significant difference in EC detection (n=20, p=2.7x10-6) and no significant difference in pericyte identification and quantification compared to manual analysis. Removal of cells associated with vessels larger than capillaries was successful, however remains low throughput. Through machine learning, several qualifiers were determined to allow for cell categorization as pericyte or EC in mouse retinal microvasculature.

Conclusions : Based on criteria from our analysis, we were able to create a threshold to distinguish between pericytes and EC’s in mouse microvasculature. Our findings suggest that with the combination of pipeline 1 and pipeline 2 we developed, we can reliably identify and quantify mouse retinal pericytes and EC’s in an automated fashion.

This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.

×
×

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.

×