In “The Automated Farm” we wrote about the promise we see for automation in agriculture. We share similar optimism for data’s ability to reshape healthcare—with key differences.

Like farming, medicine is big business. In the United States, healthcare spending accounts for a staggering 18 percent of GDP. It’s the country’s largest industry by many measures, and all signs point to continued growth for our lifetimes.

Like farming, healthcare faces enormous challenges that data technology is well suited to help solve. And like in farming, they are enormous challenges with far-reaching societal implications.

In other words, it’s an area ripe for the data-driven revolution.

AI carries the potential to address low-hanging fruit—identifying patients who may have difficulty paying their bill—to more complex problems like determining how to treat people based on their individual biology.

Venture capitalists have taken note: Rock Health reports that digital health companies recorded a record $8.1 billion in investments in 2018—a 42 percent jump from the year before—and are on track to surpass that number this year.

Start with data

While many of the data-driven solutions in agriculture begin with the physical (automated watering, warehousing, delivery, etc.), in healthcare they begin with data.

Healthcare is a comparatively late adopter of the digital data world. The industry data-rich, but data for too long has been trapped in old systems. Adoption has accelerated in recent years, propelled by technology.

Throughout the healthcare system, vast amounts of data are dispersed in complicated systems. From medical charts and financial reports in doctor’s offices and hospitals to heart-rate readings from wearables and blood glucose meters, healthcare data is everywhere. AI is capable of helping the industry sort through these data troughs and make sense of them; a big opportunity is breaking through these “data silos” to extract comprehensive insights.

Founders can set their sights on two categories of data, each with enormous potential to both lower out-of-control healthcare costs and improve health outcomes.

The first category is operational: data that helps health systems run more efficiently. This can mean anything from using blockchain “med tokens” for payments to identifying patients who are likely to skip out on a bill.

The second data category is medical: everything from doctor’s notes to data you collect on yourself. AI has the power to pull together disparate data, analyze medical images, augment diagnoses, and add value to personal data.

Companies like Viewics, which was acquired by Roche last year, are innovating new ways to apply AI to solve healthcare’s data problems. The company slurps data from a large number of sources—from medical data to financial data—to give medical staff and back-office staff new insights.

We are just scratching the surface of the potential of neural networks and artificial intelligence in a healthcare setting. 

Dependable detection

Beyond helping humans make sense of existing data, automation is proving capable of enhancing care as well—acting as a doctor’s assistant of sorts. Some of the earliest wins have come in detection.

Computers powered by AI and deep learning now outperform dermatologists at recognizing skin cancers in blemish photos. They detect other types of cancer, too. AI proved more adept than cardiologists in detecting arrhythmias in EKGs. In a paper published in Nature Medicine earlier this year, scientists reported building a AI system that diagnoses common childhood conditions—from influenza to meningitis—after processing the patient’s symptoms and medical history.

Cloud DX, out of Canada, uses AI to analyze the audio waveforms of a cough, which allows it to detect asthma, tuberculosis and pneumonia. WinterLight Labs, also out of Canada, is developing machine-learning software that can detect early-stage Alzheimer’s by analyzing snippets of a patient’s speech.

Companies are building similar technologies to automatically detect signs of illness, cancer, and disease in X-rays, MRIs, and eye scans.

An eye on innovation

We invested in Spect because we saw the promise of a data moat around optical imaging. We also saw the opportunity to make significant headway against the leading cause of blindness among working-age Americans: diabetic retinopathy. 

A traditional eye exam is burdensome and expensive, creating hurdles for the hundreds of millions of diabetes patients worldwide who are at risk of this condition. With Spect, patients take a photo of their eye from anywhere, and a retinal specialists aided with AI examines the photo. The result is a faster, less expensive exam that proves to be more accurate than the traditional method without AI. Best yet: The system improves over time, as Spect’s data moat grows.

Heart murmurs

If you want to get an idea of where medical treatment is headed, Dr. Patrick McCarthy is a good person to ask. He’s a heart surgeon at Northwestern Medicine and heads the Center for Artificial Intelligence, which opened last year.

“Artificial intelligence is the next frontier in breakthrough medicine,” Dr. McCarthy told Chicago Health. “Twenty or 30 years from now, a surgeon like me may be doing parts of operations, but a robot using artificial intelligence will be doing parts of it as well.”

McCarthy and his team are conducting clinical trials using artificial intelligence to help detect cardiac abnormalities. One of these clinical trials is Eko, a digital stethoscope bolstered with AI technology. Held to a patient’s chest, Eko records the electrical signals of the heart for 15 seconds and then uses algorithms to detect heart murmurs and rhythm irregularities.

At the same time Arterys, one of our portfolio companies, is creating the artificial “brains” that will power intelligent systems of tomorrow’s operating rooms. Using AI, advanced computer vision algorithms and 3D rendering, the company stitches together two-dimensional images from traditional medical imaging machines and creates 3D models of a patient’s heart and lungs that accurately depict the patient’s real-time blood-flow. This enhanced view makes radiologist more productive and accurate, providing them smart assistant for diagnosis. 

Accurate computer modeling is also a crucial step toward robot-assisted surgery. And, indeed, the surgical robots have arrived. 

Companies at the forefront of the data-driven revolution happening in healthcare are hard at work solving an impressive array of problems. With recent advances in human health and genetics, people have unprecedented access to information about their health risks and possible interventions. Computing these diagnoses and recommendations requires data spanning genetics, physiology, and lifestyle

We invested in Blockdoc because we see that this new wealth of medical data brings with it the need to use new technology to preserve people’s privacy. Opportunities for innovation are around every corner.

Personalized medicine

Innovations like these can turn our healthcare system on its head, transitioning from reactive to preventive medicine.  We know that if you can measure health outcomes, you can improve them; and that better health outcomes carry profound implications for society. We also know that the sooner you can detect medical issues the more effective—and less costly—the treatment can be.

Technology makes it possible to evolve to a system of predictive care. Rather than waiting until something is wrong to see the doctor, one day the doctor may call on you after noticing something is off—even before you notice the first symptom.

AI, blockchain, predictive analytics and other high-tech tools will continue to revolutionize healthcare. Properly harnessed, this will result in precision medicine: healthcare that’s calibrated to your individual biology and delivered with speed and accuracy. 

The change will be as dramatic as the transition from payphones to cell phones. A doctor’s visit in the future will look and feel very different than it does today.

To deliver on the promise of the data-driven revolution, companies must focus on the quality of the data inputs. There are few industries where accuracy counts as much as healthcare.

Isaac Kohane, a biomedical informatics researcher at Harvard Medical School, offers advice for those looking to join the revolution:

“I think that all our patients should actually want AI technologies to be brought to bear on weaknesses in the healthcare system,” she told Smithsonian Magazine. “But we need to do it in a non-Silicon Valley hype way.”

About the Authors

Ash Patel and Mike Marquez are co-founders of Morado Venture Partners, a seed venture capital firm dedicated to capturing investment opportunities in emerging technologies and data-fueled businesses.