Perspectives on (AI for) Digital Health through Startup Trends
The Healthcare Information and Management Systems Society (HIMSS) has some fantastic coverage on many of the exciting developments in the digital health space. Recently, they released a detailed table on different companies that went through fundraising, giving us a sense of up and coming trends. The table is pretty lengthy, with 66 companies(!!) spanning areas as diverse as doctor bookings, data infrastructure, and telemedicine. We summarize some of the varied efforts and high-level themes we saw below.
Simplifying Healthcare for Patients
Several of the biggest companies in the table are focusing on building friendlier, patient-focused products to increase ease and accessibility of care. They’re enabling patients to schedule appointments online, order in-home delivery of prescription medications, and discover quality healthcare providers. Many are reminiscent of consumer-focused technology transformations in areas like restaurant reservations (OpenTable), online shopping (Amazon), and business discovery (Yelp). One unique challenge of bringing consumer-focused experiences to healthcare is aligning incentives for who pays and who benefits among providers, payers and patients. For example, Vim matches patients with providers who fit their specific needs, and promoting value-based care through their platform for payers and providers is a key part of their business. Another major challenge is managing user trust and security, especially when handling sensitive health and prescription data. However, done right, the room for possible benefits are immense. Across areas including billing, referrals, appointments, and prescriptions, these products have the potential to redefine the incredibly complex interface of health systems for patients.
Examples include Babylon Health, Capsule, Nurx, Vim, and Luma Health.
Data Generation and Preparation for AI Analysis
From medical imaging to medical records to clinical transcriptions, there’s a lot of excitement about future AI applications in healthcare. Today, many of these companies begin by building individual models: they train models on static, anonymized datasets collected from a few payers or providers, and deploy these models in one hospital at a time. These deployments all rely on critical core infrastructure – such as data anonymization and security, integration with existing data systems, continuous monitoring and compliance, and display on end-user interfaces like PACS systems and EHR software.
The development of high-quality, shared infrastructure will be crucial to enabling developers and subject matter experts to more quickly and safely build software applications for healthcare. Infrastructure-as-a service is the fastest growing segment of the public cloud market today. Much like infrastructure-as-a-service is accelerating the development of software companies, companies building health data infrastructure will help enable the development of a data-driven health system. The following set of startups is focusing on building this next generation of health data infrastructure.
Examples include Synyi, MDClone, and BrightInsight.
Continuous Monitoring for Chronic Illnesses
Another large group of startups are aiming to build smart devices that enable continuous monitoring, especially for managing chronic health conditions. This union of two big technology trends, smart devices and artificial intelligence, is part of broader initiatives in healthcare to move from treatment to prevention and to move care from the clinic into the day-to-day of patients. Many larger companies, like Apple (with the Apple Watch) are also trying to make headway into this space. These devices are creating new, previously unrecorded streams of data that can potentially enable a host of new monitoring and prevention applications. For example, a recent study from Apple and Stanford showed data from the Apple Watch could be used to identify atrial fibrillation during regular use. Because these companies own the data generation pipeline, they can immediately offer useful monitoring functionality to consumers, while simultaneously collecting data with potential for more advanced prediction and intervention. By enabling better preventative care and habit-building, these smart devices could help manage the chronic, life-time conditions that currently plague the American healthcare system.
Examples include One Drop, Hello Heart, and Myia Health.
AI as a Service
An emerging group of companies are critically reliant on the many exciting recent breakthroughs in AI capabilities. Examples range from automated clinical note transcriptions to evaluation of medical images. Such uses offer the potential to reduce burdens on doctors and help with improvements in quality, accessibility and efficiency. We also see a nascent trend using AI systems to target novel signals (perhaps uninterpretable by human experts) that may enable entirely new medical workflows. One example from the list is Aural Analytics, which uses novel vocal biomarkers to track the progression of neurological diseases. Another example from researchers at Google is predicting cardiovascular risk factors from retinal fundus images that were previously not thought to be present. Such applications may be more challenging to deploy, requiring thorough validation, close collaboration with regulators, and comprehensive clinical trials to assess their impact on outcomes, but they hold the potential to bring about fundamental changes in how diseases can be diagnosed, monitored, and treated.
Examples include Kherion, Robin Healthccare, Huimei Healthcare, and Aural Analytics.
Telemedicine and Telehealth Services
Finally, there’s a group of startups offering telemedicine services that range from pre-recorded videos for behavior health coaching to video calls with remote doctors. The rise of more reliable and immersive telecommunications software is enabling the growth of remote work in many fields, and healthcare may be an especially important beneficiary. By widening the services within reach for patients and providers, telemedicine promises to enable more timely and equitable access to health services. Patients can access the knowledge of physician, trainers, and specialists from within their own home, minimizing potentially difficult transportation and logistics. Within hospitals, the same technologies are improving tele-consultations with specialists, which is particularly important for supporting providers in more under-resourced areas.
Examples include Ginger, Maple, Health Recovery Solutions, and Kaia Health.
Conclusion
This isn’t an exhaustive coverage, but highlights some of the larger trends and categories that jumped out at us from the startup fundraising ecosystem. One of our main takeaways from this analysis was the radically different ways digital technology is making its way into healthcare. It was fascinating to see the ways in which the rise of digital health parallels the growth of digital technology in other sectors, while presenting its own unique opportunities and challenges. We’re really looking forward to seeing more such innovations in the months and years to come!