We spoke with Dr Peter Thomas, Chief Clinical Information Officer and Consultant Ophthalmologist at Moorfields Eye Hospital, about the recent launch of Cascader Limited—a new medical technology company aiming to push the boundaries of eye and systemic disease detection through AI-powered retinal imaging. Backed by Moorfields Eye Hospital, the UCL Institute of Ophthalmology, and Topcon Healthcare, Cascader brings together clinical excellence, academic insight, and industry innovation. Peter Thomas is co-founder of Cascader, together with Professor Pearse Keane and Ege Ilicak.

Eye Diseases: More Common Than You Think

According to the World Health Organization (WHO), an estimated 2.2 billion people around the world are living with some form of vision impairment—over a quarter of the entire population.And the prevalence is growing. In the UK alone, the number of people with late age-related macular degeneration (AMD) is predicted to increase by 24.4%, primary open-angle glaucoma by 15.9%, and vision-impairing cataracts by 16.7% over the next decade.In the WHO’s 2019 World Report on Vision, they note that “those who live long enough will experience at least one eye condition in their lifetime.” And with a rapidly aging population, that “long enough” is increasingly relevant.By 2050, the number of people over 60 is expected to more than double—and, as we age, the risk of eye conditions like glaucoma and AMD climb. At the same time, so too does the likelihood of developing systemic diseases such as diabetes and hypertension—conditions that can directly impact eye health. It’s a double-edged sword: aging on one side, systemic disease on the other, each sharpening the rise in eye disease rates around the world.

When Sight Loss is Preventable But the System Can’t Keep Up

But despite the scale of the challenge, eye care departments, spanning both hospital-based ophthalmology practices and high-street optometrists, aren’t equipped to meet it. Over three-quarters (76%) of NHS ophthalmology departments don’t have enough consultants to meet the current demand—a pressure that will undoubtedly increase as the population continues to age. Ophthalmology is now the busiest outpatient specialty in the NHS, and many departments lack the staff, resources, and digital tools to cope with rising demand. Thousands of patients face long waits for diagnosis or treatment—delays that can lead to permanent, often avoidable, sight loss. In fact, according to the WHO, of the 2.2 billion visual impairment cases, 1 billion are thought to be preventable. This stark reality has steered the movement towards “prevention over cure,” but without the right infrastructure in place, the ambition risks falling flat. It’s like trying to construct a room in a house before building the foundations first.Peter says: “Ophthalmology became the largest outpatient specialty in about 2017—it overtook Trauma and Orthopedics. And we have an aging population; as you get older, things tend to go wrong with your eyes. It means there’s a very large burden on the health service.”

Introducing Cascader: A Company at the Intersection of Three Leading Titans

Confronting the issue head-on is Cascader—a newly-launched medical technology company aiming to build cutting-edge AI tools that can detect diseases in eye scans earlier and more efficiently than traditional methods. At the heart of the company lies a collaboration from three powerhouse institutions: Moorfields Eye Hospital, UCL Institute of Ophthalmology (IoO), and Topcon Healthcare. Alongside them, the company will be further supported by INSIGHT Eye Hub at Moorfields—the UK’s largest ophthalmic bio-resource, with over 35 million anonymized eye scans linked to clinical data. This multi-pronged network leverages strengths from three domains: (1) clinical expertise supplied by Moorfields Eye Hospital, a global leader in ophthalmology care, (2) academic “prowess” from UCL IoO, another global leader in ophthalmological research and teaching, and (3) Topcon Healthcare’s worldwide reach and technical capabilities. Peter says, “Each of these three partner organizations brings something to the table. The combination of them together is really powerful.” And, with INSIGHT Eye Hub’s additional support, the company will benefit from access to a huge reservoir of data—in essence, a rich “training ground” to support the development of high-performing AI models that detect diseases from a quick eye scan. Peter reflects on the hub as not only an “unparalleled resource in the ophthalmic world,” but one that is probably unparalleled in the non-ophthalmic world, too.

Bridging the “Implementation” Gap

But it’s not groundbreaking news that AI could provide the scaffolding for better eye care. Ranging from administrative to diagnostic support, autonomous tools could leave doctors with more time to spend on the things that matter: seeing and treating patients. So, with more and more medical tech start-ups popping up every day, we are left with a poignant question: Why are there not more of these AI solutions actually in the clinic?Oftentimes, we see AI models perform well at the research stage, but then fail to make it to clinical application—something Peter refers to as “the implementation gap.”“You generate these algorithms, you show that they work really well, but there’s no clear route to actually getting them out into the world,” he explains. “What we need is a focus on the next stage of that translational pipeline.” Addressing this problem, Cascader wants to carve out a “route through which that research work can be converted into a real-world clinical product that can be delivered in a safe way,” as Peter tells us—essentially, closing the gap between research and deployment for ophthalmological AI technology.

From Vision to Whole-Body Insight

But stretching beyond the field of ophthalmology, Peter hopes that Cascader can build AI tools to detect systemic disease too—from the same eye scans you might get at a local optician. “The eye is transparent,” says Pete. “You can look from the outside and see right to the back of it. It’s the only place in the body where you can see the human vasculature. You can also see the optic nerve, a part of the central nervous system. So it’s the only place where you can look in and you can essentially see the brain.”That visibility makes the eye “window” into broader health. Changes in the blood vessels, retina, and optic nerve can reflect what’s happening elsewhere in the body—from blood pressure and glucose levels to neurodegenerative changes. This is the principle behind oculomics: the idea that imaging the eye can reveal clues about conditions ranging from Alzheimer’s and cardiovascular disease to diabetes and even schizophrenia. Many of these changes are too subtle for humans to identify—even trained specialists. But AI algorithms excel at spotting patterns invisible to the human eye.Peter draws on biological sex as a key example, when a research paper from Moorfields and UCL in 2021 showed that AI could distinguish between male and female eyes—something that humans cannot do. “I’ve looked at an awful lot of eyes, and it’s never occurred to me that that was something that I could look for,” he says.

Reimagining Diagnostics With One Eye Scan at a Time

This idea of uncovering biological hidden traits by simply “gazing into the eyes”, might sound like a novelty, but it holds real clinical promise. Compared to traditional imaging techniques, scanning the eye is quicker, non-invasive, and can be done at a fraction of the cost—offering a much faster, easier, and cheaper diagnostic route. Consider taking a Magnetic Resonance Imaging (MRI) scan of the brain: there’s a lot you can tell from it, but MRI machines are few and far between, expensive to run, and in high demand for treating acute conditions. On the other hand, an Optical Coherence Tomography (OCT) machine—the ones that capture eye scans—are much cheaper, quicker, and there are several OCT machines on most high streets across the UK. While it costs an average of £20–£80 per OCT scan, for an MRI, this can soar up to £1,500.“It’s so easy to look at the eye,” explains Pete. “You can take the scan on a conscious patient in 30 seconds. So it’s not just that we can see all these things by looking at the eye, it’s also that there’s such ready availability of the imaging modalities.” With these advantages, it’s easy to imagine a future where a visit to your local optician could flag your risk of both serious ocular and systemic disease, perhaps years before symptoms appear. [caption id="attachment_5457" align="aligncenter" width="641"]

MRI vs OCT scans comparison.

Conclusions

While the potential to detect systemic disease is thrilling, the most “reachable” application of AI in ophthalmology still lies in eye diseases themselves.In the initial stages, Cascader’s work will center around diagnosing macular degeneration and other diseases that affect the macular. With high-performing and clinically validated AI models, Peter hopes that optometrists could accurately identify those who need urgent care, and filter out those who don’t, just from an initial OCT scan.“We have this potential that we can use the scans that get performed all the time in the high streets to detect disease at a very early stage—perhaps earlier than what previously have been clinically obvious,” says Peter. But he also hopes that, in future, Cascader will be able to focus more on the oculomics aspect of eye scans: “However, there’s there’s also that potential to say ‘this patient looks like they’re at high risk of a stroke in the next six years,’ and put together a diagnostic package to work out what they’re overall risk is. It’s tremendously promising.” By developing and, more importantly, deploying AI systems across eye care practices, healthcare could be diffused from overburdened hospitals onto the high street. This not only supports the drive towards preventative medicine, allowing earlier disease detection, it also bolsters the shift towards community-based care—freeing up doctors’ time at the busy institutions (i.e., the hospitals) to focus on patients who need them most. And, although Cascader is only at the beginning of its journey, it marks a promising step toward moving ophthalmology and oculomics-based AI from the research “bench” and into the clinics—the “bedside.” These tools may just be the “needle” that deflates the ballooning pressure on healthcare systems.Peter concludes: “There’s a huge amount of research, a huge amount of papers published, a huge amount of promise from medical AI and eye care. And the real challenge now is bringing that to the clinical frontline.”