A groundbreaking study has unveiled the potential of artificial intelligence (AI) to revolutionize the early and accurate detection of pediatric eye diseases. Researchers from Shanghai Jiao Tong University School of Medicine developed a deep learning-based model that can accurately identify eye conditions like myopia (short-sightedness), strabismus (misaligned eyes: one eye turned in a different direction from the other), and Ptosis (drooping or sagging upper eyelid), using simple smartphone images.1
The study was conducted between October 2022 and September 2023. The criteria for participant selection was that each participant had a diagnosis of one of the pediatric eye illnesses and was at least eighteen years old. The study analyzed over 1,400 facial photographs from nearly 500 children and adolescents.
In this study, Li and colleagues aimed to develop a deep learning-based model that predicts myopia, strabismus, and ptosis in children and adolescents at home using photos from mobile smartphones.
The study included 476 patients, mostly young children (62.82% were under 6 years old) and slightly more females (47.27%). To build the AI model, researchers used 1419 photos, including 473 pictures taken from both eyes to identify strabismus and 946 pictures taken from one eye to identify myopia and ptosis.
Myopia, strabismus, and ptosis are common pediatric eye problems that pose a significant risk to visual health, overall well-being, and childhood development.2 Early screening and identification are critical for successful disease management, but delays are commonly noted due to the need for in-hospital visit with an ophthalmologist.3
The AI model demonstrated impressive performance, accurately detecting all three pediatric eye diseases with comparable accuracy in both boys and girls.
Because AI has reduced the requirement for in-person screening, it has revolutionized the identification of eye problems throughout medicine, including ophthalmology.4
The AI model was able to accurately identify strabismus, ptosis, and myopia in a group of patients. Out of the 180 patients with strabismus, the model correctly identified 73% of them. For ptosis, the model correctly identified 85% of the 171 patients. And for myopia, the model correctly identified 84% of the 251 patients. Overall, the model showed good accuracy in diagnosing these three conditions.
The model was able to accurately identify the correct disease in most cases. The model's accuracy was very strong for ptosis (0.92 [95% CI, 0.91–0.93]) followed by strabismus (0.80 [95% CI, 0.79–0.82]), and myopia (0.80 [95% CI, 0.78–0.81]).
The AI model performed comparably when it came to identifying these pediatric eye disorders in both male and female participants. However, its accuracy in identifying the eye conditions varied depending on the age of the child. The model was able to identify the correct disease in most cases accurately. It was especially good at identifying myopia in older children and adolescents (ages 13-18), but not as good at identifying myopia in young children (ages 0-5). The model was also less accurate at identifying strabismus in young children compared to older children.
These results suggest that AI prediction models utilizing smartphone photographs may identify eye diseases in children and adolescents, providing a handy and early diagnostic tool for families to use at home.
Investigative team, led by Lin Li, MD, PhD, and Jie Xu, DHM, department of ophthalmology, Shanghai Jiao Tong University School of Medicine.
Overall, the AI model detected all three pediatric eye illnesses with a relatively high degree of sensitivity across a range of age and sex groups. These findings suggested that this model holds immense promise for families and healthcare providers.
Early detection of these conditions is crucial for preventing vision loss and ensuring optimal child development. With such AI-powered tools, parents can now monitor their children's eye health conveniently at home, reducing the need for frequent in-person appointments with ophthalmologists and lowering the likelihood of delayed visual loss.
The researchers emphasized the potential of this technology to improve global health standards by enabling more equitable access to eye care services. According to Li and associates, the model helps to identify potential issues early, which can help take proactive steps to protect the vision of future generations.
References
Shu Q, Pang J, Liu Z, et al. Artificial Intelligence for Early Detection of Pediatric Eye Diseases Using Mobile Photos. JAMA Netw Open. 2024;7(8):e2425124. doi:10.1001/jamanetworkopen.2024.25124 https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2822029
Kandel H, Khadka J, Goggin M, Pesudovs K. Impact of refractive error on quality of life: a qualitative study. Clin Exp Ophthalmol. 2017;45(7):677-688. doi:10.1111/ceo.12954 https://pubmed.ncbi.nlm.nih.gov/28370795/
Garcia SSS, Santiago APD, Directo PMC. Evaluation of a Hirschberg Test-Based Application for Measuring Ocular Alignment and Detecting Strabismus. Curr Eye Res. 2021;46(11):1768-1776. doi:10.1080/02713683.2021.1916038 https://www.statpearls.com/point-of-care/29513
Dong L, He W, Zhang R, et al. Artificial Intelligence for Screening of Multiple Retinal and Optic Nerve Diseases. JAMA Netw Open. 2022;5(5):e229960. doi:10.1001/jamanetworkopen.2022.9960 https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2791807
Rehash/Dr. Aditi Bakshi/MSM
