MedTech: Rare Disease Diagnosis and Beyond

Written by Dr Peter Fish, CEO Mendelian.

Healthcare services globally are under more pressure to serve growing populations in new and innovative ways. The development of medical technology (medtech) has moved beyond clinical environments and into our homes, as remote patient monitoring and advanced technology become an expected norm. To help healthcare services to thrive (rather than just survive), cutting-edge technology, including machine learning and data science, has an essential role.

Healthcare staff are the backbone of these services and much of the new technology that is having a tangible, positive impact is not aiming to replace human expertise, but instead to support doctors, clinicians and specialists, helping them to do their jobs more efficiently. Artificial intelligence-based solutions in particular are set to have a huge impact.

The Impact of Rare Diseases

One such area that is starting to see the benefits of AI integration is rare disease diagnosis. Individually rare but collectively common, rare diseases are estimated to affect 350 to 400 million people globally – or around 1 in in every 17 people. In the US, that figure rises to 1 in every 10, according to the National Organization for Rare Disorders (NORD).

Sadly, children are disproportionately affected, accounting for around 50% of those living with these diseases. Furthermore, there are now thought to be as many as 11,000 known rare diseases – and five to ten new conditions are described in medical literature every week.

Often, these illnesses present with a range of complex and seemingly disconnected symptoms and the route to diagnosis and treatment can be long and frustrating for all involved. In the US, the average diagnosis time is around 4-5 years, though in some cases, it can take up to a decade. In the UK, the average diagnosis takes 5.6 years. Those are years of interrupted or altered life for patients, frustration for doctors, and ultimately, huge amounts of money spent in “best guessing” a way to an answer.

Working Towards a Solution

Understandably then, medtech companies are putting a great deal of effort into accelerating the identification process for rare and hard-to-diagnose diseases, because there’s very little upside to the costly, traumatic process as families and healthcare professionals experience it now.

For instance, Mendelian, a UK clinician-led medtech company, has developed MendelScan – a system that uses AI to enable doctors to identify patients who may have unrecognised rare and hard to diagnose diseases, and to do it much earlier than traditional, non-AI systems have been able to do.

It integrates with existing primary care clinical systems, scanning patient health records against detection algorithms for rare disease symptoms. It captures disease features from electronic health records across a patient population, then patients are matched to published diagnostic criteria for hundreds of rare diseases.

​Once a potential rare disease is detected, cases are validated by Mendelian and then flagged to the doctor, with a detailed report explaining the condition and recommending treatment pathways. Doctors can then decide the best way to help each patient by combining their own clinical expertise with the insights from MendelScan. The technology works in the background of existing healthcare systems, meaning there is no additional software for clinicians to learn to use.

While the focus here is in diagnosing rare diseases through the application of AI, artificial intelligence-based technology is also being applied across the whole health sector. It is able to help detect heart disease earlier, as reported by the British Heart Foundation, and is a central plank of the UK’s National Optimal Stroke Imaging Pathway. And there’s even some evidence that an AI chatbot could be useful in getting patients with sexually transmitted diseases the help they need, without any of the social stigma that sometimes comes with being examined by a flesh and blood doctor.

The evolution of AI from a science-fiction concept, often vilified, into a vital part of 21^st century medtech has not been entirely straightforward, and AI is still in relative infancy in terms of adoption by healthcare practitioners and organizations. But as more applications see AI cutting treatment times, diagnosis periods, and ultimately healthcare budgets, its adoption is likely to increase.

The Evolution of Artificial Intelligence in Healthcare

AI-based solutions first entered the world of healthcare in the 1970s, with the introduction of MYCIN – a program that helped practitioners identify bacteria causing blood infections by integrating specialist knowledge into a consultation system. The technology would then draw on this knowledge to offer explanations of conditions and recommendations on treatment, based on the symptoms and context provided by a doctor.

AI research and development continued in the sector throughout the 1980s and 1990s, improving the efficiency of healthcare procedures by collecting and processing data faster; mapping out in-depth DBA research; incorporating widespread electronic health records; and developing tools that assisted medical practitioners.

The Future of Healthcare Innovation

Over the next few years, we’re likely to see a range of AI-based solutions being trialled to assist clinicians – from patient interaction to diagnosis, management and referrals. For the immediate future, the continued overhaul of administrative aspects such as scheduling, communication, management and patient education is likely to remain a focus. Beyond that, the next decade will likely see a shift toward this type of technology having more direct clinical impact. We can expect progress in areas such as remote monitoring, health management, objective symptom collection, disease diagnosis and drug discovery, for example.

Eventually, technology advances are likely to become part of the fabric of healthcare services, helping overhaul old systems. The leveraging of analytics to mine significantly untapped reserves of clinical data will aid clinicians, with the ultimate outcome being saving and prolonging more lives. While separate technologies will, and already do, contribute value alone, the true potential lies in the synergy of multiple advances across the entire patient journey.

Written by Dr Peter Fish, CEO Mendelian.