Emergence of non‐artificial intelligence digital health innovations in ophthalmology: A systematic review

Abstract

The prominent rise of digital health in ophthalmology is evident in the current age of Industry 4.0. Despite the many facets of digital health, there has been a greater slant in interest and focus on artificial intelligence recently. Other major elements of digital health like wearables could also substantially impact patient-focused outcomes but have been relatively less explored and discussed. In this review, we comprehensively evaluate the use of non-artificial intelligence digital health tools in ophthalmology. 53 papers were included in this systematic review -25 papers discuss virtual or augmented reality, 14 discuss mobile applications and 14 discuss wearables. Most papers focused on the use of technologies to detect or rehabilitate visual impairment, glaucoma and agerelated macular degeneration. Overall, the findings on patient-focused outcomes with the adoption of these technologies are encouraging. Further validation, large-scale studies and earlier consideration of real-world barriers are warranted to enable better real-world implementation.

Actualización en inteligencia artificial y oftalmología

“…Su aporte al mundo de la retinopatía, en su estudio de 32 OCT, permitió concluir que las imágenes proveen hasta 87.5% de sensibilidad para la patología antes descrita, con 16 casos de control previamente analizados, y especificidad también altamente representativa de 87.5%. [30][31][32][33] Los diseños de software que emplea el sistema de OCT permiten aportar información, no sólo en el caso de diabetes como se documentó en las series asiáticas, sino también de enfermedades como glaucoma, mediante el diseño de algoritmos capaces de detectar la pérdida de capa de fibras nerviosas, antes de ser clínicamente manifiestas. Además, se han diseñado programas más allá de la campimetría periférica automatizada, los cuales también tienen la capacidad de detectar defectos en el campo visual y, por ende, facilitar el diagnóstico de dicha enfermedad previo al daño estructural irreversible.…”

Welcoming artificial intelligence into ophthalmology

“…Using almost 100 000 articles published between 2000 and 2021 and identified in the PubMed database, the team developed an algorithm that used year, title, authors, abstract, medical subject headings or MeSH, and journal to predict which articles would be most highly or most lowly cited. Finally, the interesting article by Tseng et al 11 ‘rebutted’ the current major focus on AI with a detailed description of real‐world digital health technologies that do not involve AI. Their systematic review—a first in the field—focused on virtual or augmented reality, self‐monitoring applications and wearables, and covered 53 articles that highlighted many innovations.…”