June 2023
Volume 64, Issue 8
Open Access
ARVO Annual Meeting Abstract  |   June 2023
Improving Predictive Models for Noncycloplegic Binocular Autorefractors
Author Affiliations & Notes
  • Thayjas Patil
    Drexel University College of Medicine, Philadelphia, Pennsylvania, United States
  • Pranav Sharma
    Drexel University College of Medicine, Philadelphia, Pennsylvania, United States
  • Heather Modjesky
    Conestoga Eye, Pennsylvania, United States
  • David Silbert
    Conestoga Eye, Pennsylvania, United States
  • Footnotes
    Commercial Relationships   Thayjas Patil None; Pranav Sharma None; Heather Modjesky None; David Silbert None
  • Footnotes
    Support  None
Investigative Ophthalmology & Visual Science June 2023, Vol.64, 1464. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Thayjas Patil, Pranav Sharma, Heather Modjesky, David Silbert; Improving Predictive Models for Noncycloplegic Binocular Autorefractors. Invest. Ophthalmol. Vis. Sci. 2023;64(8):1464.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : Handheld binocular infrared autorefractors are often used to screen children for amblyopia risk factors and are particularly useful in international medical mission work. To assess whether noncycloplegic devices could be used to prescribe glasses for children, a retrospective clinical study compared the noncycloplegic SPOT autorefractor readings to cycloplegic refractions performed during recent medical missions in the Andean region of Ecuador. Preliminary models were created to better predict a pediatric ophthalmologist’s eventual prescription from SPOT’s refractive values alone.

Methods : Values for sphere (sph), cylinder (cyl), and axis for both right (OD) and left (OS) eyes were measured from 305 patients, ages 3-24, who presented at Partners for Andean Community Health located in Ecuador. Each patient’s refractive error was analyzed using SPOT and were compared to a standard dilated refraction (WRX) and the final prescription (Rx). Each variable’s measurements were compared across different linear regression models: a linear model, a quadratic model, and a polynomial model to the power of 4 degrees. The R2 dictated the model of best fit. All analyses were conducted using RStudio 2022.07.2 in January 2023.

Results : All measurements had at least one model fit a range of data with good predictability. 3 models with the highest predictability were the Spot OD cyl vs final Rx (R2=0.60), the Spot OD cyl vs WRX (R2=0.76), and the Spot cyl OS vs the final Rx at an (R2=0.72). Preliminary refraction prediction model equations were also found to go from Spot to the final Rx for OD sph (y= -0.827+21.2x-4.38x2-3.09x3-0.0664x4), OS sph (y=-0.755+14.8x-13.3x2-6.02x3+5.45x4), OD cyl (y=1.79+22.3x-1.9x2-1.35x3-0.855x4), OS cyl (y=1.95+14.6x+9.96x2-7.67x3+0.0728x4), OD axis (y=95.7 +91.2x-40.4x2-162x3+50.3x4), and OS axis (y=87.5+80.8x+56.1x2-39.4x3+13x4).

Conclusions : Preliminary tests prove that the gap between the SPOT autorefractor, the standard dilated refraction, and the final prescription can be modeled and better predicted. Further analyzation of the bounds of the predictive ranges and improved outlier removal will increase the usability of the model. Continual improvement of these tools may allow the use of binocular infrared autorefractors in under resourced areas to provide a good starting point for pediatric glasses prescriptions and may allow training of non-ophthalmic providers to prescribe glasses.

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.

×