AI may help address challenges associated with home monitoring of chronic ocular diseases
Click Here to Manage Email Alerts
Key takeaways:
- Machine-learning algorithms show promise in monitoring ophthalmic conditions.
- Clinical trials and economic evaluations are necessary to determine the role of artificial intelligence in home monitoring.
Artificial intelligence has potential to facilitate accurate, frequent and larger-scale home monitoring of chronic ocular diseases when safely integrated into workflows, according to a review published in Current Opinion in Ophthalmology.
“Monitoring health conditions in the home environment is useful for chronic diseases that require rapid intervention or frequent data sampling to decrease risk of irreversible vision loss, particularly in [age-related macular degeneration] and glaucoma,” Tiarnán D.L. Keenan, MD, PhD, from the National Eye Institute, and Anat Loewenstein, MD, from Tel Aviv University, wrote. “In addition to the possibility of early detection and intervention, potential advantages include convenience, decreased cost and intensive monitoring of clinical trial participants.”
According to Keenan and Loewenstein, artificial intelligence (AI) may address challenges associated with home monitoring, including the generation of large volumes of data from frequent testing. Ophthalmic information collected via home-based monitoring includes functional, biometric and imaging data.
One home-monitoring method for visual acuity is the FDA-cleared Visibly Digital Acuity Product, a web-based, self-guided software application for individuals aged 22 to 40 years. Others include smartphone-based acuity tests such as the Portable Eye Examination Kit Acuity application and OdySight.
Researchers noted that computer simulations indicate that weekly home monitoring of visual fields may detect glaucoma progression earlier than 6-month in-office visits, despite unreliable compliance and accuracy.
Two FDA-cleared approaches for home monitoring of preferential hyperacuity are ForeseeHome (Notal Vision) and the myVisionTrack application (Vital Art and Science), which have been used for early detection of progression to neovascular AMD.
An approach for ophthalmic imaging at home is the Notal Home OCT System, which includes the SD-OCT device and the Notal Health Cloud. Scans taken by individuals are analyzed by the Notal OCT Analyzer (NOA), an AI-based software application that detects and quantifies intraretinal and subretinal fluid. Studies have demonstrated high agreement between automated NOA and manual detection and quantification of fluid.
Keenan and Loewenstein noted that while prospective randomized trials, real-world studies and economic evaluations are still needed to determine the role and value of AI in home monitoring, machine-learning algorithms show promise in detecting progression to neovascular AMD and interpreting OCT self-imaging data.
“Overall, artificial intelligence has the potential to facilitate more accurate and larger-scale home monitoring, if it is integrated safely, efficiently and transparently into workflows,” they wrote. “Importantly, this should be alongside human physicians, instead of replacing them, and artificial intelligence outputs should be transparent and verifiable wherever possible.”
Collapse
Click Here to Manage Email Alerts