The Need for Speed: What Machine Learning Means for Healthcare’s Future - and Siri’s Brain

With its adolescent Jeopardy win, Watson started the machine-learning craze. Now it’s matured to the point that its brain-power is being used in healthcare settings, and Big Daddy’s planning to sell its stuff.

Stephanie Baum’s title is a great one: “Will Apple pick Watson’s brain for healthcare apps in partnership deal with IBM?” It’s a relevant question, and indicative of the place-at-the-table already reserved for machine learning in healthcare’s future.

As has been noted, there isn’t a much bigger Dynamic Duo in tech than former competitors Apple and IBM, so their recently announced partnership has provided a feast of jaw-dropping chatter. It wouldn’t seem like such a big-healthcare-deal if it weren’t for the fact that IBM sired Watson, and Apple is about to give birth to HealthKit. And the real kicker is that both have been sidling up to EHR giant Epic to complete the Digital Health Trifecta.

What’s the real scoop beneath the hype? Watson, of course. 

With its adolescent Jeopardy win, Watson started the machine-learning craze. Now it’s matured to the point that its brain-power is being used in healthcare settings, and Big Daddy’s planning to sell its stuff.  No wonder physicians are wringing their hands—wondering if they’re about to be replaced.

To try to grasp the long and short of it, let’s take a quick look at what the hubbub is all about.

Machine learning is gaining popularity in healthcare because of its ability to use existing mathematical models and apply them to new instances of an established concept in other data. This ability to automatically identify patterns in data is one of the major reasons for the potential of machine learning in healthcare settings—as well as its ability to fill in the gaps of expert knowledge, adjust for exceptions, and efficiently handle massive amounts of data.

Both machine learning and artificial intelligence (AI) are on the rise across the expanse of technology. A new Big Data analytics startup, Adatao, is a perfect example. As Christopher Nguyen, the former engineering director for Google Apps and now Adatao founder noted, “When you have enough memory and compute, a funny thing happens. It comes alive.”

You’ve got to admit there’s something creepy about that.

In addition to other components, Nguyen’s company is using an artificial neural network (ANN) to make it happen—which gives it the ability to “identify data objects on the fly in response to queries entered in plain English.” As Eric Knorr, editor in chief at InfoWorld notes, “Thanks to the cheap, enormous computing resources that are making new intelligent systems possible, we are now entering into a qualitatively different phase of computing.”

In “Machine Learning Floats all Boats on Big Data’s Ocean,” (another great title), James Kobielus, Senior Program Director for Product Marketing for Big Data Analytics Solutions at IBM, refers to machine learning as “the unsung hero that powers many of the most sophisticated Big Data analytic applications.” He notes that its core function—“enabling analytic algorithms to learn from fresh feeds of data without constant human intervention and without explicit programming”—is what makes it so. Since Kobielus hails from within IBM’s ranks, he’s undoubtedly quite familiar with young Watson’s rapid rise to fame.

When IBM partnered with Wellpoint in 2011 to apply Watson’s processing and analytical abilities to healthcare, the stage was set to explore the potential of tapping into the massive amount of data available within a variety of sources to deliver precise, evidence-based healthcare. Data processing and predictive analytics have come a long way since the partnership was announced, but it provided an initial framework for all others involved in such efforts since then: “

Watson’s ability to analyze the meaning and context of human language, and quickly process vast amounts of information to suggest options targeted to a patient’s circumstances, can assist decision makers, such as physicians and nurses, in identifying the most likely diagnosis and treatment options for their patients…Watson can sift through an equivalent of about 1 million books or roughly 200 million pages of data, and analyze this information and provide precise responses in less than three seconds. Using this extraordinary capability WellPoint is expected to enable Watson to allow physicians to easily coordinate medical data programmed into Watson with specified patient factors, to help identify the most likely diagnosis and treatment options in complex cases. Watson is expected to serve as a powerful tool in the physician’s decision making process.”

Indeed, the Cleveland Clinic was eager to join the Watson effort, and was the first institution to do so—with an eye on training Watson for use in clinical settings. The Clinic’s Lerner College of Medicine is working on two new technologies: WatsonPaths and Watson EMR Assistant. Though currently used for educational purposes only, the hope is that WatsonPaths will be able to help doctors diagnose patients and solve medical problems—due to its ability to understand spoken language and process vast amounts of research in mere seconds. Watson EMR assistant is geared toward helping with the analysis of electronic medical records. Memorial Sloan-Kettering Cancer Center has joined as well, and physicians at both are teaching Watson to recognize different types of cancers and treatments. According to IBM’s CEO Virginia Rometty, “This will change the face of healthcare. This is a new era of machine-human collaboration.”

With its efficiency and accuracy improving quickly, IBM recently announced the formation of the Watson Group, a division made up of 2,000 employees that will focus exclusively on the continued development of the supermachine’s abilities for a variety of applications both in healthcare and other industries—a move that coincides with the company’s revision of its 2015 “business analytics” goal from $16 billion to $20 billion. That’d explain why Big Daddy just announced it’ll be selling Watson’s stuff as Watson Discovery Advisor—a software-as-a-service product targeted at “helping research teams analyze vast troves of data to come up with new research ideas.”

IBM’s continued development of neuromorphic systems like Watson that mimic the human brain has led to the recent release of its brain-based, neurosynaptic chip—a project operated in collaboration with Cornell University and iniLabs, and partially funded by the Defense Advanced Research Projects Agency (DARPA). Breaking away from the previous von Neumann model, the chip is equipped with 5.4 billion transistors that can simulate 1 million neurons, 256 million synapses and 46 synaptic operations per second per watt—and requires much less energy than the average microprocessor to do it. According to Dharmendra Modha, IBM Fellow and Chief Scientist, Brain-Inspired Computing at IBM Research, “These brain-inspired chips could transform mobility, via sensory and intelligent applications that can fit in the palm of your hand without the need for Wi-Fi.”

IBM says synthesized cognition via neuromorphic systems could be used in all kinds of devices—including smartphones—to mimic the brain’s abilities related to perception, action and cognition. That includes the collection and analysis of raw data for a variety of purposes. As we look to the new Dynamic Duo, we can once again say that healthcare will never be the same—since Siri is now Watson’s little sister—and it looks like she’s about to get a new brain.