Comparative Validation of PlusoptiX and AI-Optic Photoscreeners in Children with High Amblyopia Risk Factor Prevalence

Abstract

Purpose: In 2022, an inexpensive multi-radial infrared photoscreener, the AI Optic was released in a similar format as the 2012 PlusoptiX a-12 but utilizing artificial intelligence (AI) for online, central image interpretation. We studied them because no prior comparative validation concerning amblyopia risk factors and particularly refractive error has been done.

Patients and methods: Children from a pediatric ophthalmology practice had AI Optic and PlusoptiX-a12 photoscreen concomitantly during comprehensive examination with precisely measured strabismus and refraction. Validation to AAPOS 2021 and 2013 guidelines was evaluated with receiver operating characteristic (ROC) curves while refractive estimates were compared by the ABCD ellipsoid univariable technique.

Results: In 200 ethnically diverse children aged 1-18, 148 were 4 years or older, 35% had developmental delays, and 2/3 had amblyopia risk factors (ARF). Area under the ROC curve (AUC) targeting AAPOS 2021 refractive plus strabismus for age ≥4 years was 0.58 for AI Optic and 0.74 for PlusoptiX while for children <4 years, AUC was 0.53 for AI Optic and 0.72 for PlusoptiX. For 134 comparable sphero-cylinder refractions, the ABCD Ellipsoid median (interquartile ranges) for AI Optic right eye 2.53 (1.54, 4.01) and left eye 3.05 (1.83, 5.00) did not approximate actual refraction as well as by PlusoptiX right eye 1.88 (1.12, 2.86) and left eye 2.10 (1.26, 3.04) Mann Whitney z=3.7 right and 4.2 left, p<0.001 each. AI Optic via central "AI" reading gave sphero-cylinder and referral estimates in all but 16 of 200 high risk children whereas Plusoptix had 25 inconclusives. On the other hand, with inconclusives scored as a refer, Plusoptix outperformed AI Optic in terms of ARF validation and refractive estimate.

Conclusion: Plusoptix provided more valid amblyopia and refractive screening than the cheaper AI Optic. Clinics must weigh cost versus performance, and central data sharing before selecting one of these vision-saving devices.

Keywords: amblyopia; amblyopia risk factor; photoscreening; sphero-cylinder refraction.

Figure 1

The AI Optic (left) and the PlusoptiX a-12 photoscreeners.

Figure 2

The receiver operating characteristic (ROC) curves comparing AI Optic to the Plusoptix a-12 in children over 4 years of age targeting the 2013 AAPOS amblyopia risk factors for strabismus and refractive error with inconclusive results counted as a “refer” and Instrument referral criteria “Arnold, ABCD” chosen indicated by the green arrows.

Figure 3

Receiver operating characteristic (ROC) curves targeting AAPOS 2021 refractive and strabismic amblyopia risk factors comparing school age (left column) with preschool (right column). Inconclusive results (i) are either counted as a “refer” (I = refer) or excluded from calculations of sensitivity and specificity. Instrument referral criteria “Arnold, ABCD” chosen indicated by the green arrows.

Figure 4

The ABCD ellipsoid compares spherical cylindrical refractions from Plusoptix and AI Optic with actual (cycloplegic) refraction. Top metric box plots with interquartile ranges (yellow and purple) highlights values of 1 correspond to one line visual acuity blur (grade A) while value of 2 implies 3 logMAR lines blur (grade B) and value of 3 6 logMAR lines blur (grade C). Bottom metric shows proportion with each grade (colors purple, Orange and green) blur and those that are a poorer spectacle match than grade C (teal color) (more than 6 lines visual acuity blur).