AI Can Detect Pediatric Eye Diseases From Mobile Photographs

TOPLINE:

Artificial intelligence (AI) can detect eye diseases in children with high accuracy using mobile photographs.

METHODOLOGY:

• This cross-sectional study, conducted at Shanghai Ninth People’s Hospital in China from October 1, 2022, to September 30, 2023, involved 476 children diagnosed with myopia (n = 251), strabismus (n = 180), and/or ptosis (n = 171).
• The researchers used 1419 images taken using a smartphone to train a model based on deep learning.
• They assessed the sensitivity, specificity, and accuracy of the model.

TAKEAWAY:

• The model showed high sensitivity and correctly identified myopia, strabismus, and ptosis in 84%, 73%, and 85% of cases, respectively.
• It also correctly identified individuals without myopia, strabismus, and ptosis (specificity: 76%, 85%, and 95%, respectively).
• The model showed good accuracy (80%-92%) and high performance (83%-94%) for all three conditions.

IN PRACTICE:

“These results suggest that it can assist families in screening children for myopia, strabismus, and ptosis, facilitating early identification and reducing the risk of visual function loss and severe problems due to delayed screening,” the authors wrote.

SOURCE:

The study was led by Qin Shu, MD, of the Shanghai Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China; Jiali Pang, MS, of the Shanghai Artificial Intelligence Laboratory, Shanghai; and Zijia Liu, PhD, of the Institute of Image Communication and Network Engineering at Shanghai Jiao Tong University, Shanghai. It was published online on August 6, 2024, in JAMA Network Open.

LIMITATIONS:

The single-center design and small sample size of the study may limit the generalizability of the findings. Only one photograph per participant was used, which may restrict the capabilities of the algorithm. The sample sizes for the three conditions varied, potentially affecting the sensitivity of the model in detecting strabismus.

DISCLOSURES:

The study was supported by the National Natural Science Foundation of China; Shanghai Science and Technology Innovation Action Plan Biomedical Technology Support Special Project; Target Commission of China Hospital Development Institute, Shanghai Jiao Tong University; Shanghai Jiao Tong University School of Medicine High Peak Plateau Double Hundred People Plan; Shanghai Rising Stars Young Medical Talents Cultivation Program; Shanghai Municipal Commission of Health Excellence Research Program; and the Undergraduate Training Program on Innovation, Shanghai Jiao Tong University School of Medicine. The authors declared no conflicts of interest.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.