fbpx

Investing in AI Healthcare Companies and Stocks

How Can Artificial Intelligence Help Healthcare?

Throughout the years we’ve been watching AI percolate throughout a diverse range of industries, from fintech and AI-powered investment banking to smart farming technologies. As we noted in our earlier articles on trends in artificial intelligence, healthcare is where machine learning and deep learning will pave the way to a new era of better drugs and smarter healthcare management.

Machines are becoming as good as humans in their ability to interpret medical images. About 40% of healthcare providers reportedly use some form of AI-powered, computer-assisted diagnostics like chatbots or apps that offer personalized health advice based on a patient’s data, biometric inputs from wearables, or the rich datasets contained in today’s electronic healthcare records (EHRs). 

While the investment opportunity for artificial intelligence has largely belonged to venture capitalists up until now, the rapid growth of AI-powered healthcare solutions means retail investors can now build an investment strategy around this emerging technology. Forget IBM and Dr. Watson – we’re talking about real pure-play investments in AI healthcare companies.

The AI Healthcare Investment Thesis

With more than 3,000 “AI startups” around the globe focused on using machine learning, deep learning, and other AI technologies to solve problems, investment opportunities have been limited to a handful of computer vision stocks, a few pick-and-shovel investments in AI chips, and a handful of companies that provide AI-powered solutions to healthcare organizations. One reason that the healthcare industry gets so much attention is because of its size.

In 2018, the global healthcare market reached a value of $8.5 trillion, with $3.65 trillion spent in the United States alone. There are very few trillion-dollar industries out there. If you’re able to capture even the tiniest percentage of that number, you can easily have a billion-dollar business.

There are also the feel-good aspects around investments that help democratize health care, improve the quality of life, and help save lives. Healthier people live longer, so more healthcare resources will be required to take care of tomorrow’s elderly. In the long run, investors can do well by doing good.

How Will Artificial Intelligence in Healthcare Transform Clinical Experiences?

Healthcare is one of the top industries impacted by artificial intelligence for other reasons as well. The global healthcare sector is not only massive but it’s also complex, full of inefficiencies, and has loads of patient data readily available in shoddy form. Imagine what AI systems can achieve in a sector where old-school data warehousing still hasn’t been mastered. Add to that the hefty price tag associated with anything medical (in the US the average hospital stay costs $10,700 and the average doctor earns $294,000 a year) and even tiny incremental efficiencies will make a huge difference to the bottom line. Among the many potential benefits, AI applications promise to improve patient outcomes, shorten drug development and clinical trials, and empower preventive and precision medicine. Besides the feel-good aspects of democratizing health care, the resulting efficiency gains are translated into hard dollars for investors in AI healthcare companies.

How to Invest in AI Healthcare Technologies?

The IBM Watson Thesis

One potential investment thesis we looked at involved focusing on large corporations like IBM. For example, IBM’s Watson has seen considerable resources allocated to it, though nothing seems to be happening there at the moment. The Watson Health division recently appointed a new leader who is doing the typical come to Jesus talk about getting back to basics, doubling down, and being laser focused on execution. It remains to be seen what – if anything – will come from this. (With that said, we do hold IBM as one of two technology stocks in our dividend growth investing (DGI) portfolio) Maybe IBM’s new leadership will be more transparent with shareholders about its AI healthcare strategy – if there is one – rather than a scattershot approach that seems based on generating publicity.

The Medical Device Maker Thesis

Another possible investment thesis for AI healthcare might surround medical device companies that are developing hardware and software solutions that capture tremendous amounts of healthcare data. All this data can be fed to machine learning algorithms which can then help with decision making or provide predictive analytics to help shape precision medicine, truly personalized healthcare for the masses. Medical device giants are actively dabbling here: Medtronic with its move into surgical robots and Stryker with its adoption of robotics and 3D printing. We hold both as part of our DGI strategy, but neither would be considered pure-play AI stocks. We needed to take a different approach.

Five Areas of Healthcare Leveraging AI

In order to find pure-play stocks in the AI healthcare space, we started by identifying areas of healthcare where artificial intelligence already has been making an impact. None of this “we’re planning to disrupt X Y or Z” stuff, but actual revenues being generated. Our research led us to the following five areas of healthcare AI is having the most impact on:

  • Natural language processing
  • Medical imaging and diagnostics
  • Drug discovery
  • Big data analytics
  • Personalized medicine.

The focus areas listed above provide a growing opportunity for both institutional and retail investors.

Natural Language Processing

If you don’t know much about natural language processing, we’ve got you covered with our guide on Investing in Natural Language Processing. Simply put, much of our interaction with intelligent computer systems today involves communicating through text or voice command. Think about all of the clever (or not) chatbots that immediately start pestering you the moment you land on a new webpage. Or all the times smart voice assistants like Siri or Alexa respond (or not) to some trivial piece of trivia. These are all examples of a discipline of artificial intelligence known as natural language processing (NLP), which refers to the ability of the machine to read language and turn it into structured data. Under NLP there are two additional categories:

  • Natural Language Understanding (NLU): the understanding of human language by computer.
  • Natural Language Generation (NLG): the ability of machines to write things like reports and sports stories.

While NLP is the umbrella term used here, much of the technology behind the ability of virtual assistants, for example, to answer healthcare questions actually relies on NLU, which has come on the scene much more recently. As you can imagine, it’s extremely important to transcribe what a doctor is saying with 99.99% accuracy (no software is perfect). NLP technology has advanced to a point where it’s now able to provide critical support functions that help creates efficiencies in healthcare. From transcribing paper documents to helping schedule appointments, AI algorithms are what the hospital of the future will be built on.

Medical Imaging and Diagnostics

One of the area of healthcare that’s quietly being transformed by AI is medical imaging. Deep learning algorithms can interpret medical images better than humans now. All imaging technologies stand to be improved – X-rays, ultrasounds, MRIs, and other types of scans – by AI algorithms that can offer a diagnosis or even prognosis. It was wishful thinking that a behemoth like IBM would dominate in this emerging market. It’s really the startups that are the future of medical imaging, particularly in oncology, where AI is becoming at least as good doctors in spotting tumors.

Tempus is also using deep learning techniques to predict which tumors might respond favorably to immunotherapy treatment.
A startup called Tempus is using deep learning techniques to predict which tumors might respond favorably to immunotherapy treatment – Credit: Tempus

Medical imaging has become such a large market that startups are now carving out niches for themselves. Some use machine learning to sharpen images, or employ natural language processing to extract information from health records for more precise diagnostics. Others focus on medical imaging big data for research purposes. A startup called Vara Healthcare provides an end-to-end workflow that starts from when a medical image is taken and ends when a diagnosis has been made. Another company called MaxQ is looking to gain competitive advantage by partnering with industry giants like IBM Watson and GE Healthcare. (We were surprised to see MaxQ pull its IPO filing in June 2019. The company remained in private hands since then.) We’ve covered at least 26 companies with FDA-approved AI algorithms, five of which we highlight in our piece on 5 Small Global Stocks With FDA Approved AI Algorithms. HeartFlow uses AI to analyze imagery and create 3D models of the heart to diagnose heart disease, but their attempt at going public using a SPAC failed.

Medical imaging became one of the largest markets for AI healthcare use cases, but other diagnostic applications are also catching up. A number of AI startups are working on solutions for blood testing, urinalysis, and big data algorithms for early detection of diseases like breast cancer (Volpara Health is an Australian stock in this space). Retail investors will want to check a promising pure-play stock in this space, Renalytix AI but you’re probably best served waiting until they get FDA approval and achieve meaningful revenues.

Drug Discovery

Developing new drugs isn’t for the faint of heart. One oft-quoted study found that it takes more than a decade and upwards of $1.3 billion for a pharmaceutical company to bring a drug to market. AI algorithms can help jumpstart the process by rapidly identifying promising molecules out of billions of iterations. Some companies are developing AI solutions to expedite clinical trials, which are the most expensive phase of the process. And a few of the more than 200 startups working on various aspects of AI-driven drug discovery do it all. The use of machine learning in drug discovery and development is really the latest tool in the digitization of the pharmaceutical industry. The optimal way to discover drugs now is a strategy that’s become known as computational drug discovery, where advances in computing, sequencing, and modeling, are helping to accelerate the drug discovery process.

Credit: Paper on Computational Chemistry for Drug Discovery

Machine learning takes even more of the guesswork out of the equation, predicting everything from potential toxicities to how a drug candidate might interact with different body tissues and chemistry. The many startups developing computational drug discovery solutions have different takes on the problem. BigHat Biosciences foucses on biologics. Some, like Turbine, focus on specific diseases like cancer. Others, like Insilico Medicine, try to look at a broader group of problems like age-related disease and human longevity. A third group of companies including BenevolentAI, Verge, NuMedii, and TwoXAR have a holistic approach, trying to re-imagine the entire drug discovery process to work as efficiently as possible. Investors need to watch out for red flags though. Not all aspiring companies will be able to deliver on their promises. Leaders in this space right now include Atomwise and Recursion, with funding continuing to pour into AI drug discovery companies. Retail investors should check out our article titled, Comparing Four AI-Powered Drug Discovery Stocks. Six months after that piece was published, all four were trading at 50% off. Probably the least risky of the lot would be Schrodinger given their pick and shovel business model while Recursion, Exscientia, and AbCellera are both appealing to those who value potentially large rewards down the road. We’re avoiding Relay Therapeutics and BioXcel Therapeutics.

Big Data Analytics

It’s no secret that an AI algorithm is only as good as the data you feed it. In the last few years, plenty of companies have come along that apply artificial intelligence to the trove of digitized data that has been created. These data science companies are able to provide insights into everything from financing and trading to selling more sodas. In industries that haven’t seen a lot of operational innovation over the years, like healthcare, there’s lots of low-hanging fruit for healthcare data analytics companies. Big data paired with AI can help shrink down the size of medical devices or power apps that help manage our healthcare. Some companies offer enterprise systems with learning algorithms that ingest millions of data points to enable value-based care, a system that rewards healthcare providers based on outcomes rather than sheer volume under the old fee-based paradigm. Other companies combine big data algorithms with sensors to allow patient monitoring inside hospitals or help keep track of patients’ progress remotely, something that falls under the label of digital healthcare.

Value-Based Care vs Fee-Based Care in medical system.
Credit: University of Illinois at Chicago

Personalized Medicine

The concept of personalized medicine goes back thousands of years. Ayurvedic medicine, practiced by nearly all of India’s 1.3 billion people, is probably the most well-known example. Practitioners believe that each person has a highly individualized constitution called prakruti. Your body type or prakruti offers a guideline of what you should eat and affects your susceptibility to certain diseases. Western science has just added a layer of technology by employing AI, big data, and genetic testing to provide personalized recommendations on leading a healthy life.

This concept is not only applied in clinical environments. Consumer apps are now aplenty ranging from mental health applications to personal training and holistic wellness.

Market map of wellness tech startups from CB Insights.
Credit: CB Insights

How Can Retail Investors Invest in AI Healthcare Stocks?

So, you’re probably wondering which stocks you can invest in that might provide exposure to the exciting themes we’ve talked about today. We’ve got you covered. After years of researching these themes, we were able to identify seven pure-play AI healthcare stocks that anyone can invest in. Each company has its own specific business segments, applications, and strategy that are relevant to the market niche they serve. They’re also all relatively new. Six out of seven are very recent IPOs. Five out of seven are in the early stages of dynamic growth, acquiring market share rapidly, and consequently posting heavy losses. Most of these companies are offering a scalable product to a well-defined target market.

Sounds exciting? We created a detailed 55-page analytical report on these seven pure-play stocks. It’s only available to our readers who help support our work financially. If you want access to this report, you’ll need to purchase our reasonably priced Nanalyze Premium Annual subscription.

Not ready yet? That’s okay too. Sign up for our weekly newsletter. We’ll keep you up-to-date on AI investing along with other disruptive technologies out there that ought to be on investors’ radar. No politics, no B.S., no corporate buzzword bingo. Click here to sign up for Nanalyze Weekly.