Joseph Rubinsztain, MD, CEO & co-founder of ChronWell, a digital health and therapeutic company focused on digestive diseases.
Medicine remains part art, part science. It’s important we acknowledge there is still much to learn about human disease, so we can improve faster with new tools to learn from experience and act intelligently. Modern automation paired with artificial intelligence offers enormous opportunities to accelerate learning and deliver evidence-based personalized healthcare at scale.
By extending the reach of clinicians, AI-powered technologies are playing a vital role in the digital transformation of healthcare. Specifically, innovation in AI and robotic process automation can improve outcomes and reduce cost via holistic, personalized and patient-facing interventions based on well-accepted medical practices. In medicine, clinicians will evolve to utilize modern technologies as robotic extenders to ensure millions of patients receive the highest standard of care at a reasonable cost. It’s an incredibly exciting time to pioneer this field.
Healthcare’s AI Applications
People are most familiar with applied AI for common tasks such as natural language processing and computer-aided visual detection in specialties like radiology to assist with stratification and diagnoses. AI has also experienced growth in other advanced fields like psycho-social and genetic profiling, real-world sensors and conversation engines, paving the way for precision medicine and digital healthcare workers. Digital care coordinators, dietitians and coaches for the masses are just around the corner.
Taking a step back, it’s important to note that advanced technologies like AI and, more specifically, machine learning, actually work counter-intuitively by finding formulas that explain real-world results in a known population. These formulas identify the variables that most impact the health outcomes of a given patient group and can be used to create personalized care plans. This level of understanding can be used in prescription digital therapeutics, allowing providers and insurers to (finally!) deliver optimal treatment with marginal labor cost to large populations. In other words, a single prescription can trigger and execute a cascade of orders optimized for each patient that are fulfilled through an automated supply chain.
The Right Rx: Gear, Gadgets And Behavior
Remote monitoring is nothing new: From diabetic to cardiac patients, sensors and software help care teams keep an eye on conditions outside the clinic, but this information is rarely used to identify opportunities and deliver interventions before a medical event occurs. The FDA is acting proactively to bridge the gap between medical and consumer-facing technology through an accelerated software certification program for digital therapeutic platforms (software as a medical device platforms).
Smartwatches have sensors to monitor physical activity, sleep, blood oxygen, cardiac activity and even glucose levels. These watch monitors are becoming smarter and offer a “magical user experience” in the smallest of packages. The leap to medical-grade sensors paired with platforms certified to provide evidence-based recommendations tailored for each patient is a logical one. If the wearer’s blood glucose spikes often, weight is rising and the latest laboratory results show tissue damage, further testing and a visit to the doctor should follow without human intervention. This is AI at work in medicine.
Taking the technology a step further, consider a patient who is obese and afflicted by fatty liver disease, high blood pressure and high cholesterol. The clinician will prescribe diet and exercise — typically with little patient support — and request a follow-up visit in six months. Imagine, instead, prescription AI that tracks motion and pressure in a shoe to detect exercise and weight trends and adjusts recommendations considering medical and socio-economic information for each patient. This “magical experience” is the only way to reach 64 million people who suffer from fatty liver. The progress of medical technology has been substantial over the past decade, and I can’t wait to see what the next five years in ethical AI applications will bring.
Interconnectedness And Curability In Medicine
To harness the power of data in medicine, industry leaders must actively support sharing of existing medical information. The U.S. government has provided guidance for data sharing, but the information remains stored in silos. Still, data extraction and quality are key to reaching full data liquidity and must be approached without bias and through smarter tools. Data sharing must also be done ethically and with patient consent, noted most recently last month by the WHO’s report on the global use of AI for the public interest.
Digital therapeutic products derived from best practices and new insights from machine learning offer a key opportunity that has yet to be realized: optimal effective treatment of patients with common conditions that are difficult to control. The reality is that treatment for our most common and curable conditions — illustrated by the obese patient above — is not well reimbursed under fee-for-service models. For clinicians to deliver the best outcomes at the lowest possible cost they will have to “hire” digital workers to address diet, monitoring and therapy, to name a few — to support major lifestyle changes. Advanced technologies will enable smarter, more effective, personalized treatment that provides true continuity of care, changing the future of wellness and medical treatment.