Machine learning + wearable medical devices = a healthier future for all

By Anne-Lindsay Beall, SAS Insights Editor

It was 1945 when top US scientist Vannevar Bush foresaw a future world where machines would begin to think and learn. It seemed like science fiction then, but today, things as common as your Google search results are a product of machine learning. Netflix uses machine learning to provide personalized movie recommendations. eHarmony uses it to predict love. Banks use it for cybersurveillance and to monitor for fraud and abuse.

Machine learning affects our lives more and more every day, and yet there’s a lot of confusion about what it is, what it can do and why you should care.

Heather Lavoie, Chief Operating Officer at Geneia, speaking at The Premier Business Leadership Series, explained: “Simply put, machine learning refers to the techniques used to construct and study algorithms that can learn from data. The algorithms operate by building models based on inputs and using that to make predictions or decisions rather than following explicitly programmed instructions.”

Heather Lavoie, Chief Operating Officer at Geneia, spoke about machine learning at The Premier Business Leadership Series
Heather Lavoie, Chief Operating Officer at Geneia, spoke about machine learning at The Premier Business Leadership Series.

Seeing patterns and saving lives

At Geneia, a health care technology and consulting company, they use big data along with machine learning to help health care organizations deliver better patient care at a lower cost.

“Geneia brings data in from a lot of different sources,” said Lavoie. “Clinical data, lab data, physiological data, actuarial data, consumer data. We’ve built Theon, a unified platform to integrate the data and allow us to apply machine learning techniques.”

Using machine learning, Geneia can match and determine missing values, as well as perform principal component analysis and look at patterns in the data – clusters that help them see trends and causality.

“Machine learning allows us to see patterns in the data that we couldn’t see before. And with wearable medical devices and sensors, we can bring in so much more information,” said Lavoie. “Air-quality information, humidity, number of times they’ve opened the refrigerator, used the toilet, etc., and we are able get a more robust picture of an individual and trends across populations.”

Applying machine learning to all of this data allows Geneia to be more accurate – to predict earlier and intervene sooner than they could in the past with clinical assessments and lab values – especially from continuous data sets that show changes much more quickly.

“For example, if your grandmother has suffered congestive heart failure and is recuperating at home, wearable medical devices can quickly catch changes in her weight or pulse oxygen levels or respiration rate. We can bring all this data into the Theon application and to detect and alert earlier than ever before when health might be deteriorating,” said Lavoie.

As Lavoie points out, there aren’t enough health care providers for our aging population, and there’s an urgent need to find new, unconventional ways to care for patients.

"Machine learning combined with wearable medical devices will help us improve the health of our population, increase longevity and allow patients to convalesce where they’re happiest – at home with their families.”

The future of machine learning

While the use of machine learning is becoming more common across industries and organizations of all sizes now, it will become “ubiquitous and essential for businesses to compete,” said Lavoie.

And while we’ll need to recognize the limitations of even well-trained models, Lavoie envisions applying machine learning to the synthesis of previously unrelated data sets (environmental, socioeconomic, biometric, financial, genomic, agricultural, etc.). That will likely result in breakthroughs even Vannevar Bush couldn’t have foreseen.


Read More

Get More Insights


Want more Insights from SAS? Subscribe to our Insights newsletter. Or check back often to get more insights on the topics you care about, including analytics, big data, data management, marketing, and risk & fraud.

Back to Top