Most promising applications of AI in healthcare
We are excited by the continuing application of artificial intelligence to healthcare. While this is a wide subject matter and there is certainly a lot of “hot air”, we think there are some great investment opportunities.
On the consumer side, we have already seen how consumer data and wearables can inform and help individuals track their own wellness. From improved sleep quality to fitness tracking to home monitoring of heartrate for cardiovascular patients, applications have led to better health, better information and reduced costs for some groups.
However, from an investor standpoint, consumer digital health businesses introduce several challenges. First, the regulatory position is often grey as regulators catch up with these new forms of digital diagnostics and treatments, resulting in an unhelpful uncertainty for entrepreneurs and investors. Second, there is increasing concern from healthcare providers, insurers and governments that the trackers will increase the numbers of “worried well,” placing strained healthcare systems under even more pressure. Again, this provides uncertainty as these bodies consider how to address this issue. This is then combined with the challenges of building a consumer business in an environment where there are an ever-increasing number of competitors. Altogether this means a lot of care needs to be taken in analysing these opportunities.
On the other hand, as access to high quality large and diversified data sets becomes a distinguishing factor in providing superior care, businesses with interesting (but compliant) data collection strategies may make compelling investments. This could include companies with novel methods to collect data, ways to include underrepresented groups or new ways to contextualise or structure the data.
Also promising are direct AI applications in the healthcare delivery system: helping care givers make better treatment decisions and improving their diagnostic powers. With serious illnesses such as cancers, an accurate and early diagnosis is key. We have been investigating companies that use AI across large data sets to ensure earlier diagnosis and better chances for effective treatment. For example, our latest investment, Optellum, has designed AI tools that significantly improve the speed and accuracy of diagnosis of lung cancer from lung scans. One thing that sets Optellum apart is the focus on diagnosis rather than detection. We have seen plenty of companies working on technology to disrupt radiology by enhancing detection capabilities, but we don’t think this goes far enough. A diagnostic element is critical. Gaining acceptance by the medical community of a diagnostic tool is not an easy feat, but for software companies that achieve this, there is significant value to capture.
We have also been seeing a multitude of teams use AI for research and new drug discovery. In one of the latest large-scale deals, Atomwise launched an AI-powered drug screening program with Enamine in June. Atomwise will use its capabilities in AI to simulate interactions of 10 billion compounds in pursuit of treatments for childhood cancers. This follows the Insitro and Gilead deal signed in April to combine computational modeling and wet lab experiments to speed drug discovery. Other smaller players such as Recursion Pharmaceuticals are expanding into white spaces that have traditionally not been economically attractive to Big Pharma. Recursion argues that AI and automation make it possible to study rare diseases in a way that is efficient and economically viable. Finally, companies such as Sensyne Health and Flat Iron are using access to large real-world patient data sets to enable clinical trial design and potentially identify precision patient cohorts.
As we monitor technology companies in healthcare, we look for groups that have achieved important milestones, making progress with regulatory buy-in, clinician buy-in and validation via credible peer reviewed publications. As an investor community, we often lose sight of the challenges specific to tech companies in the healthcare sector. Traditional communication routes that work for other products (especially consumer products) simply won’t work due to high regulation and even higher cynicism after the Theranos scandal.
However, we are galvanised by the promise of several young companies that we have seen. With its strengths in pharmaceutical research and the strong backbone of the NHS, the UK remains an extremely interesting place to be looking for promising new healthcare investments.