Apple, acquisitions, and adherence: Inside IBM's Watson Health unit

After its first six months in existence, is IBM's Watson unit for the medical and life sciences industries in good health?

IBM is raising Watson's profile in the healthcare world. Image: IBM

IBM's cognitive computing effort Watson may still be best known for winning Jeopardy! back in 2011, but since then, Big Blue has been working to raise its profile in the business world.

After reportedly making just $100m from Watson in the three years that followed its quiz show win, IBM doubled down on the cognitive computing tech. In January last year, IBM set up the Watson Business Group with an investment of $1bn, aimed at recasting the natural language learning system as the motor behind cloud analytics services for the pharmaceutical, retail, banking, media, and education industries.

In April this year, IBM stepped up its Watson efforts once again with the launch of Watson Health, its first vertical offering involving the technology. That the company targeted the healthcare market initially is no surprise -- two of the first pilots of the technology involved oncology hospital Memorial Sloan Kettering Cancer Center and health insurer Wellpoint.

According to Dr Shahram Ebadollahi, VP of innovation and the chief science officer at IBM Watson Health Group, the unit was set up in part in response to the changing nature of healthcare data.

Read this

Will Apple bring developers to IBM's Watson ecosystem?

IBM was hoping to build a mobile developer ecosystem for Watson, but that takes time. It's quite possible that Watson can piggyback on Apple's iOS developer base.

Read More

"We know that data in healthcare is becoming increasing available and it's at the point where managing it and making sense of it is becoming problematic. If you look at the so-called determinants of health -- things that have a bearing on people's health and wellbeing -- 10 percent of them are clinical data, things that get captured in a patient-physician interaction, 30 percent are genomics related data, and more than 60 percent are the so-called exogenous data -- things that happen to people outside a regular care setting," he told ZDNet. "Behavourial, social, environmental, and non-regular clinical type of data."

Among the drivers behind the growth in exogenous data is the spread of health-related connected hardware. Not only is there more data being gathered by fitness bands such as the Fitbit, or the smartphone platforms like Apple HealthKit, Google Fit, or Samsung's S Health, there's also more and more medical kit -- insulin pumps, pacemakers, blood pressure cuffs, and so on -- that's capable of recording data about its user.

"The data around activity, nutrition, other things that could be captured but don't get captured in your physician check-up have a lot of important things to say. Take a diabetic person: they probably see their physician once every 90 days, but what happens in between those visits has a lot of bearing on what direction they're going to go. This data is becoming important, especially in managing chronic conditions," Ebadollahi said.

Healthcare providers will need to gather that data, combine it with other data sets such as electronic patient records or hospital reports, and add it to their analytics systems like Watson Health Cloud to enable doctors, as well as patients themselves, to get information they can use to manage their health, he added.

But with so much exogenous data being generated on a great many diverse devices and platforms, there are obvious difficulties in getting that information out of the silos it's in and into the system the healthcare provider wants to use.

Over the six months since its inception, the Watson Health group has struck a number of partnerships with companies in the healthcare field, from medical device manufacturer Medtronic, to Johnson & Johnson, to latterly Apple - and its that last partnership IBM is hoping can help tackle the silo problem.

"IBM has a strong partnership with Apple that goes back a year and a half or so and that is about the enterprise enablement of Apple devices and apps, but in Watson Health, now this partnership goes into HealthKit and ResearchKit-type things -- how to interface with those kind of data sets, how to bring them into Watson Health Cloud, and providing services for the provider institution that needs that kind of thing for their patients."

Acquisitions and target markets

In the six months since its inception, Watson Health has ramped up significantly -- it got a new headquarters and a new GM recently, and has launched a number of new cloud services aimed at the healthcare industry. The headcount at the unit has doubled by "at least two orders of magnitude," IBM's Ebadollahi said. "It's growing very fast."

It's alsos acquired several businesses, including medical imaging company Merge for $1bn, which will enable the unit's analytics systems to be put to work on X-rays and other medical images as well as written material.

Watson Health has also snapped up data company Explorys, and health management software firm Phytel. Explorys connects to healthcare providers' systems to bring their data into the cloud and run analytics on it, for example, for risk stratification and care coordination for large populations of people, while Phytel's software lets clinicians manage individuals' care.

"Explorys provides a macro view of the population whose conditions you're trying to manage... and if Explorys is about a macro view of population, Phytel is a micro view: you know this person is at risk for their diabetes condition or heart failure condition, the platform enables the provider to deliver care to them and do outreach," Ebadollahi said.

Curam, which makes administration software and was bought by IBM back in 2011, has also been moved under the Watson Health umbrella. Initially a social services software product, under Watson Health, Curam has been put to work looking at how social factors can impact healthcare.

"5,000 hours per year people go about their daily lives and don't see physicians. What happens to them in those 5,000 hours has a lot of bearing on their health and wellbeing," Ebadollahi said.

The rationale behind the new additions was to give Watson Health products the ability to manage health data from single-individual level to population level, he added.

Machine learning face-off: Microsoft uses Band to show what its Watson rival is capable of

Plug-and-play machine learning data models that can instantly analyze information and provide intelligent insights? Yep, Microsoft has 'em, it says.

Read More

And, despite its name, Watson Health is looking beyond the healthcare industry: earlier this month it added new life sciences products including the IBM Watson Health Cloud for Life Sciences Compliance, which lets researchers put regulated workloads, such as preclinical trials, onto a cloud infrastructure.

As well as drug discovery and post-market analytics, IBM is targeting the pharma industry with its partnership with US pharmacy chain CVS Health, which is centred on the issue of adherence -- how well patients follow the treatment regimes prescribed by their doctors.

"On average, people make probably nine or ten visits to such a pharmacy store when they don't go to their doctor every month. You have an opportunity to address issues like adherence. One of the things we're working with CVS Health on is the issue of adherence -- how to make sure that pharmacist or the person in a retail clinic... one of the things they're working on is how they can help that customer or patient stay adherent to their medications. If they don't, all the money [healthcare organisations spend on drugs] is going to waste, and they're not getting better. You can extrapolate this to all sorts of uses across life sciences."

With its first six months now behind it, the main concern for Watson Health is integrating IBM technologies both old and new, Ebadollahi said.

"It's a fast-growing organisation with multiple parts. The priority is bringing all the pieces together and taking these new technologies -- our cognitive technologies, the technologies coming from our research division -- adding them to the capabilities that came through the acquisitions," he said.

Read more about IBM's Watson