Application and performance of artificial intelligence in screening retinopathy of prematurity from 2018 to 2024: a meta-analysis and systematic review
Journal Title: Eye Science - Year 2024, Vol 1, Issue 3
Abstract
[Purpose] Artificial intelligence (AI) significantly enhances the screening and diagnostic processes for retinopathy of prematurity (ROP). In this article, we focused on the application and performance of AI in detecting ROP and distinguishing plus disease (PLUS) in ROP. [Methods] We searched PubMed, Embase, Medline, Web of Science, and Ovid for studies published from January2018 to July 2024. Studies evaluating the diagnostic performance of AI with expert ophthalmologists' judgment as a reference standard were included. The risk of bias was assessed using the QUADAS-2 tool and QUADAS-AI tool. Statistical analysis included data pooling, forest plot construction, heterogeneity testing, and meta-regression. [Results] Fourteen of the 186 studies were included. The pooled sensitivity, specificity and the area under the curve (AUC) of the AI diagnosing ROP were 0.95 (95% CI 0.93-0.96), 0.97 (95% CI 0.94-0.98) and 0.97 (95% CI 0.95-0.98), respectively. The pooled sensitivity, specificity and the AUC of the AI distinguishing PLUS were 0.92 (95% CI 0.80-0.97), 0.95 (95% CI 0.91-0.97) and 0.98 (95% CI 0.96-0.99), respectively. Cochran's Q test (P< 0.01) and Higgins I2 heterogeneity index revealed considerable heterogeneity. The country of study, number of centers, data source and the number of doctors were responsible for the heterogeneity. For ROP diagnosing, researches conducted in China using private data in single center with less than 3 doctors showed higher sensitivity and specificity. For PLUS distinguishing, researches in multiple centers with less than 3 doctors showed higher sensitivity. [Conclusions] This study revealed the powerful role of AI in diagnosing ROP and distinguishing PLUS. However, significant heterogeneity was noted among all included studies, indicating challenges in the application of AI for ROP diagnosis in real-world settings. More studies are needed to address these disparities, aiming to fully harness AI's potential in augmenting medical care for ROP.
Authors and Affiliations
Rui Liu, Guina Liu, Xiaoshuang Jiang, Fang Lu
Keywords
Related Articles
- EP ID EP768756
- DOI 10.12419/es24082501
- Views 5
- Downloads 0
Research and prospects on the visual mechanisms and treatment modes of amblyopia
Amblyopia is a neurodevelopmental vision disorder resulting from abnormal visual input during the critical period of visual development, such as strabismus, uncorrected anisometropia, high refractive errors, and form dep...
Advances in understanding conjunctival goblet cell traits and regenerative processes
Conjunctival goblet cells are of great significance to the ocular surface. By secreting mucins—particularly MUC5AC—they play a pivotal role in stabilizing the tear film, safeguarding the cornea from environmental insults...
Prevalence of idiopathic thoracic scoliosis in children and adolescents with superior oblique palsy: a cross‑sectional study
[Objective] Strabismus is associated with a higher prevalence of idiopathic thoracic scoliosis (ITS) in children. This study aims to investigate whether superior oblique palsy (SOP) poses a higher risk for developing ITS...
Interaction of ductal obstruction and glandular dropout in the pathogenesis of meibomian gland dysfunction
Meibomian gland dysfunction (MGD) manifests through two main clinical presentations, characterized by the meibomian gland (MG) ductal obstruction or acinar dropout. While previous research has predominantly associated MG...
A novel technique to identify and remove the conjunctival lymphangiectasia using indocyanine green gel
In this case report it describes a novel technique using indocyanine green gel for the surgical excision of conjunctival lymphangiectasia (CL). CL was found on the nasal side of left eye in a 50-year-old male. Surgical e...