AI Revolutionizing Eye Care Insights from Mark Daly
Mark Daly discusses the transformative role of AI in ophthalmology, highlighting its potential to enhance patient care and streamline physician workflows.
(Image credit: (c)InfiniteFlow/AdobeStock)
Mark Daly, the Chief Technology Officer at Digital Diagnostics, spoke alongside other experts during a session at the Association for Research in Vision and Ophthalmology (ARVO) annual meeting, which was held in Salt Lake City, Utah. The session was titled, "AI regulation and implementation in ophthalmology" and took place on Tuesday, May 6, from 8:30-10:15 AM.
Following the meeting, Ophthalmology Times connected with Daly to discuss the use of AI (artificial intelligence) in the field of ophthalmology and what the future might hold for its increasing role in medical care.
AI's Current Success in Eye Care The LumineticsCore Example
Mark Daly emphasized that AI is already making a significant impact in ophthalmology today. He pointed to LumineticsCore, an AI software device with breakthrough designation, as a prime example. This tool has successfully navigated the path from FDA clearance to reimbursement by Medicare, Medicaid, and major payors. Daly explained that LumineticsCore's success stems from its design, which directly addresses a critical need in eye care: the lack of patient access and the heavy burden on physicians.
He further highlighted that the entire product and its development process are rooted in a bioethical framework. This framework considers patient engagement, regulatory aspects, ethics, and economics as fundamental components for bringing a safe, effective, and trustworthy product to market. A key technical aspect in the regulatory journey involved understanding and quantifying performance and the variability in how physicians interpret images. Daly stressed the importance of collaboration between innovators and regulators to establish clear pathways, including consensus reference standards, for bringing new products to market as regulators adapt to rapid technological advancements.
Tackling Challenges for Ophthalmologists and Patients with AI
AI offers solutions to several pressing challenges in ophthalmology. For physicians, Daly noted the significant time spent reviewing images for high volume screening programs. After a long day in the clinic and managing numerous inbox messages, sifting through hundreds of diabetic retinopathy (DR) screenings to meet service commitments can be arduous. LumineticsCore can alleviate this burden, improving patient satisfaction with point of care results and reducing physician workload.
From the patient's perspective, the primary challenge is often accessing appropriate care. Daly shared that most LumineticsCore patients are currently identified in primary care settings. If the AI detects disease, these patients are then referred to an ophthalmologist. This means AI solutions like LumineticsCore are guiding patients who might otherwise not be tested into ophthalmology appointments. For many, this is their first visit to an eye specialist, presenting a valuable opportunity for connection and patient education.
The Future of Diagnostics AI for Early Detection and Prediction
Looking ahead, AI tools are poised to revolutionize early disease detection. Daly explained that current diagnostic imaging tests are designed for specific conditions in particular populations, with AI models being developed to replace human interpretation of these images. These AI models can enhance consistency and availability, as software does not experience fatigue and is not geographically limited. While these largely serve as direct replacements for human interpretation now, Daly expressed excitement about the future.
As large scale data collection on diverse populations becomes more common, de identified longitudinal records (tracking the same patient over time) can be used to train new models. These advanced models could identify conditions much earlier than current methods and potentially predict disease progression timelines. This predictive capability is especially crucial for planning interventions for conditions such as Age related Macular Degeneration (AMD) or Diabetic Macular Edema (DME).
Daly also highlighted the promising developments in using retinal imaging for the early screening of neurological conditions like Alzheimer disease. This area is actively being researched by his team and others, potentially opening new avenues for future AI products and applications.
Navigating the Evolving Regulatory Landscape for AI in Medicine
The regulatory environment for AI in medicine is dynamic worldwide, with regulators responding to new AI technologies. Daly noted that a new administration in the US is bringing significant changes to FDA strategy and goals. There is a consensus in the industry that current frameworks struggle to keep pace with the rapid rate of technological change. Consequently, changes and proposals are emerging from both the FDA and international bodies to address these challenges. For example, South Korea's Ministry of Food and Drug Safety (MFDS) recently announced guidance for devices using Large Language Models (LLM) and Large Multimodal Models (LMM) technology.
Daly anticipates that the market demand for AI innovation will compel regulators globally to provide clearer pathways. He believes everyone recognizes the potential impact of AI, even if the specific regulatory nuances for deploying the technology safely at scale are still being refined.
Practical Steps for Implementing AI in Ophthalmology Practices
For ophthalmologists looking to integrate AI tools into their practice, Daly offered several recommendations. Most importantly, he emphasized collaboration with clinical partners. This includes fostering relationships with endocrinology, primary care, and other settings that provide referrals. Understanding how to support larger system reporting and quality initiatives as a valuable contributor with unique quality metrics is key. He pointed out that AI reimbursement, currently more common in ophthalmology and cardiology, presents a unique opportunity to highlight when collaborating.
Another effective strategy is to study other systems that have successfully deployed AI, understanding their planning processes and partnerships. Daly warned that projects can be derailed if they lack an integral IT or clinical champion, advising to identify supporters early. Partnering with strong technologists who understand both workflow and data integration is essential. For instance, having a certified imaging informaticist on the team is almost a requirement for a successful technology program deployment.
Responsible AI Deployment Maximizing Benefits and Minimizing Risks
Daly believes there are numerous opportunities for AI tools to enhance the patient experience, improve quality, and increase care availability. Innovations are emerging daily across the healthcare journey, from appointment scheduling to understanding exam results and treatment education.
He asserted that the best value is achieved when automation addresses significant existing gaps in access to care. However, this is contingent on the technology having undergone proper processes and generated evidence to prove its safety, efficacy, and cybersecurity. The primary risks Daly has observed in this space involve inadequate monitoring of deployed systems for consistent performance and insufficient planning for workflow and integration aspects of a successful deployment.
