Author: Ulya Khan, Chief Operating Officer, Physicians Interactive
Who is Dr Watson? Most people know Watson™ as IBM’s Jeopardy!-winning artificial intelligence supercomputer that premiered about a decade ago. Not long after, IBM’s ambitions for Watson were steered toward having it evolve into a medical genius—hence the name, Dr Watson.
These days, IBM boasts that Watson can be used in practically any industry. Their website proclaims, “Which industry will IBM Watson help transform next? Well, which industry are you in?” Essentially, Watson could be used for anything—from crafting state-of-the-art inventions to designing cutting-edge architecture.
Dr Watson has read countless medical textbooks, cover to cover, along with the massive PubMed and Medline databases of medical journals—in their entirety—and thousands of patient records from Memorial Sloan Kettering Cancer Center (MSKCC), a treatment and research center in Manhattan. (As of 2014, U.S. News & World Report ranked MSKCC as the “#1 cancer hospital” in the country.)
Connecting the dots
In science, there are countless seemingly unrelated data points. It’s difficult to put together those complex relationships or to realize that those relationships even exist—but, not for Watson. Watson has a discovery capability called “Discovery Advisor” that can visualize relationships in information from millions (perhaps trillions) of data points it has absorbed from textbooks, online articles, human interactions, and more; information that could take researchers decades or even a lifetime to process.
Scott Spangler, a lead IBM researcher, said, “Watson … is a powerful tool that understands nuances in human language. It also provides the ability for researchers to visually see connections in their data they would not have considered … and can reduce the time researchers need to formulate hypotheses, increasing precision, and accelerating the pace of discovery from years to months and months to just days.”
Here’s how IBM describes how Dr Watson could work in practice
A physician would describe symptoms and other related data points to Dr Watson. The patient’s medical record would be given to Dr Watson, which can then identify the key pieces of information and comb through the patient’s data to find relevant facts, such as medications, family history, blood work results, and comorbidities. Dr Watson can even order further tests to rule out (or rule in) possibilities. It combines all this data, then forms and tests its hypotheses by comparing it to immense data stores (eg, clinical trial findings, electronic medical record data, digital medical notes, peer-reviewed research, treatment guidelines).
After all this data processing and testing, Dr Watson would provide treatment choices and rate its confidence for each option.
The Watson supercomputer may soon be the best doctor on the planet
As you might’ve guessed, Dr Watson is capable of storing far more medical info than a human doctor; what’s more is that it can detect insights where no one would’ve ever thought to look. As scientists at IBM continue training Dr Watson to harness its vast stores of knowledge so as to use the supercomputer in real-world medical decision-making scenarios, even the sharpest human minds are hardly a match against Watson’s diagnostic abilities. And, Dr Watson is getting smarter every day.
Unlike humans, its conclusions are all objective and evidence-based—free of biases and subjectivities. Watson can generate hypotheses, evaluate the fortitude of those hypotheses, and learn from others and itself. Its capabilities are beyond simply storing data; IBM calls it “cognitive computing”.
Here’s my question…
Watson is now being used in some clinical settings and a handful of studies have shown that the computer is better at diagnosing disease than humans because of the information it has at its disposal. Let’s take away the stigma of alternative care models as inferior to seeing a doctor in a clinic. What if a patient doesn’t have access to this doctor for whatever reason?
My question is this: If we know that these supercomputers can diagnose disease just as well (or better) than most humans, why would we deny access to care for those who have no access to a live physician, if we could use these amazing computer-based solutions?
Reasons for Dr Watson
1. Doctor shortage: With the shortage of young doctors coming out of medical schools, there is a market need. Watson would address supply-and-demand problems among growing patient populations and a shrinking market for general practitioners.
2. Access: If a patient anywhere in the world has access to a computer or phone, Dr Watson is on call for them.
3. It’s accurate, objective, and consistent.
4. Availability: Dr Watson is always available. Dr Watson is never sick, on hiatus, vacationing, etc.
5. Price difference between seeing a human doctor vs Dr Watson: It was, of course, very expensive to build and train Dr Watson. However, the cost of getting diagnoses and finding answers using Watson is minimal. For example, if a regional or national health system utilized Dr Watson, obviously the original purchase price would not be cheap. Yet, once it’s in place, there is hardly a cost to keep it going.
Dr Watson could change our world
It’s not a matter of replacing human doctors. It’s a matter of doctors working in conjunction with Watson. Bottom line—allowing access to medical consultations that don’t require an in-person examination would bring unparalleled access to healthcare to those who might not have it.
If we gave free access to Dr Watson worldwide, can you imagine how it would change that world along with the overall population health status? Watson connects the dots from trillions of points of data. What we discover from this can lead to the breakthroughs that progress our medical world and save millions of lives. When this amazing advanced technology exists today, why aren’t we leveraging it more? I think it’s time.