Investors are piling into healthcare AI start-ups; there’s room for more

Jason Schoettler of Calibrate Ventures talks about the potential for growth in healthtech and how it's attracting plenty of venture investors with their eyes on the $800bn US healthcare industry.

“The robot doctor will see you now.”

If you expand the conception of robots to include artificial intelligence, such a future is already here.

During the pandemic, millions of people worldwide shifted their doctor visits online, using telehealth, text messaging and chatbots. These platforms often use natural language processing, a form of AI, to understand patient questions, direct the conversation, and provide diagnoses and treatment options.

VC Venture AI
Jason Schoettler, Calibrate Ventures

AI had already been making inroads in healthcare, but the pandemic accelerated this trend. The healthcare AI sector is expected to grow from just under $5 billion in 2020 to more than $45 billion by 2026. As an investor in the AI and automation sectors for the last 15 years, I’ve never seen a bigger focus on AI in healthcare than today, and other investors seem to agree. In the first quarter of 2021, US investors poured $2.5 billion into 111 healthcare AI start-ups, up 140 percent year-over-year, according to CB Insights.

With such huge potential for growth, the healthcare AI sector is attracting many venture investors who are hoping AI start-ups can make inroads into the $800 billion US healthcare industry. Start-ups such as Babylon Health, K Health and Spring Health, which use AI to offer more personalized care, have achieved lofty valuations. Many other biotech companies, such as PathAI, Paige.AI and Caresyntax, have raised hundreds of millions of dollars for their AI-powered disease detection and drug discovery platforms.

But although healthcare AI start-ups are in hot demand, there is still ample opportunity for investors in this space. That’s because we’ve only scratched the surface of what AI can accomplish in the huge and fragmented healthcare sector.

Below are four areas of healthcare AI where promising start-ups are emerging, and which hold outsized promise for investors.

Providing better preventive care

AI software can ingest vast amounts of data from medical databases, anonymized patient records, doctor notes, test results, insurance claims and other sources to make meaningful predictions. AI software can ‘learn’ from itself as data volumes increase, spotting potential problems for a patient down the road and suggesting preventive care strategies before true illness arises.

GNS Healthcare and Healx, for example, use AI to suggest optimal drug treatment programs for individual patients based on their unique profiles.

Unlocking value from datasets, anonymously

Hospitals, insurers, pharmaceutical companies, and researchers have a treasure trove of data, most of which remains under lock and key due to regulatory, privacy or legal concerns. However, sharing insights from this data would offer tremendous public good and potentially generate new sources of revenue for data owners. One way to ethically and safely share data is to anonymize it, cleansing it of private information.

One company using AI in this way is Diveplane, which my firm invested in. Diveplane has built AI and ML decision-making models that enable hospitals to turn private patient data into anonymous synthetic data twinsets that can be analyzed for clinical, diagnostic or research reasons. Duke Health and the UK’s Alder Hey Children’s Hospital are already using Diveplane to unlock value in their datasets.

Providing more accurate diagnoses 

One of the most exciting uses for AI in healthcare is to provide more accurate diagnoses. Traditionally, doctors review case notes, patient records and published research for diagnoses, but doctors are stretched thin and don’t always have the time to research all the potential conditions for each patient.

HealthTensor, another company my firm invested in, has built an AI platform used by doctors to provide more accurate diagnoses. It ingests a patient’s electronic medical record and drafts a physician’s note, including likely diagnoses, based on the most recent information in the EMR and from other anonymized clinical, treatment and diagnosis data. The doctor may then accept, edit or delete the note, giving them full control over the suggested diagnosis.

There are many other diagnostic opportunities for AI, such as better understanding of medical images – with Aidoc and Caption Health as two interesting start-ups in this space –or predicting disease progression based on remote patient monitoring data.

Providing better remote care

The pandemic has shown us all that virtual care can replace many in-person visits, and AI-enabled systems will improve remote care by ‘understanding’ individual patients’ symptoms, situations and potential conditions. By using natural language processing, chatbots can direct patients to care within minutes, either routing them to virtual visits or understanding that they need urgent in-person care.

AI is also being used in care and companion robots for home-bound or elderly patients, for cognitive interaction apps that help Alzheimer’s patients retain function, and in many other care capacities.

A couple of start-ups making great inroads in AI-assisted care are Alertive Healthcare and Optimize.health.

The US healthcare system is costly, disjointed and ripe for disruption, but has so far withstood widespread change. AI could finally begin turning the tide to make healthcare more affordable and accessible. AI-powered computers can ingest, parse and understand vast volumes of data, unlocking new ways for doctors to diagnose and care for patients, for patients to access personalized care and for hospitals to manage costs.

Investors looking to fund the next generation of start-ups that will truly make an impact on the healthcare system – in the US and around the world – should look to AI healthcare companies that are tackling specific challenges or segments of the overall market.

AI will not solve all of healthcare’s problems, but it’s a step in the right direction.