Zimmer Biomet uses novel Azure-based AI model to predict individual patient gait speed recovery outcomes after joint replacement surgery

The medical technology company Zimmer Biomet has a long record of supporting orthopedic patients and clinicians through technical innovations. The company’s mission to alleviate pain and improve quality of life spurred exploration into how technology can empower clinicians and patients before, during, and after orthopedic surgeries. Zimmer Biomet’s patient engagement app, mymobility, helps patients who have had hip and knee replacements understand the recovery process and assess whether their mobility and pain levels are typical for the post-operative period. 

The mymobility app is part of Zimmer Biomet’s broader ZBEdge™ ecosystem, a suite of digital health tools, robotics, and implant technologies. ZBEdge offers hospitals and ambulatory surgery centers access to pre-, intra-, and post-operative technologies through an integrated product experience. Its data is aggregated, analyzed, and translated into actionable clinical insights that can inform surgeons through the platform. Each tool was built by Zimmer Biomet to unlock this information and deliver it seamlessly into the workflow of today's surgeons and care teams.

The company believed that better information about a patient’s recovery could notify clinicians of slow progress or unusual symptoms. Zimmer Biomet needed to develop a broad AI framework with customized algorithms to analyze the data and help patients and clinicians understand specific aspects of their recovery progress.

Developing a robust AI strategy for healthcare analysis with Microsoft Azure

As the team at Zimmer Biomet began developing a Microsoft Azure–based AI strategy for their business, their top priority was to create a responsible framework that would protect patient data and comply with medical privacy requirements. “It’s all about the data now,” says Mark Brincat, Senior Director of AI and Advanced Analytics at Zimmer Biomet. “To have any chance of building something that is sustainable, you must have an AI strategy. Establishing the right digital foundation, particularly in a healthcare setting, is key to success.”

In surveying the available technology, Zimmer Biomet decided that choosing Microsoft would make it possible for the company to build a sustainable and secure AI solution. “We carried out a thorough process, talking to companies and working out what approach we wanted to take,” says Michael Griffiths, Machine Learning Engineering Associate Director at Zimmer Biomet. “We had the freedom to decide which cloud provider we were going to use, and we saw that Microsoft was a good fit all around.”

Zimmer Biomet wanted to build an environment where it could not only explore and develop ideas but also deploy and manage products. The team began to use Azure Machine Learning to develop algorithms and models for data analysis, with a focus on being able to explain to surgeons why the data models delivered particular results. “We wanted our products to include clear explanations,” says Karl Surmacz, Associate Director, Data Science, at Zimmer Biomet. “When our models offer a prediction to a clinician, they also provide the reasons behind that prediction.”

Another goal of the Zimmer Biomet team was to fully automate internal company tasks, reducing manual work and making it possible for the team to easily share information. The company turned to Azure DevOps to automate tasks so that employees could access and use the AI environment quickly. To integrate new and future AI-assisted products, Zimmer Biomet uses Azure Kubernetes Service (AKS). “AKS helps us to have a scalable platform,” says Griffiths. “We can add data from products into it, and that environment will meet the resource needs because it’s backed by the Azure-provided infrastructure.”

Providing near-real-time insight into patient gait recovery and supporting early interventions

With the AI framework in place, Zimmer Biomet began conceptualizing new products and services, such as the orthopedic industry’s first near-real-time prediction model for knee and hip replacement. mymobility collects gait data and patient-reported data, which WalkAI uses to predict which patients are on track with their recovery and which are expected to have a lower gait pace (compared to similar patients) 90 days after a knee or hip replacement. “The data can really contextualize a patient’s gait recovery, either to reassure them that they are on track or to encourage them to be a little bit more active,” says Surmacz.

This groundbreaking innovation allows clinicians to gain a personalized view of each patient’s recovery and determine which individuals would benefit from additional follow-up. “We have enough data to get more personalized and predictive and start to move away from a very reactive market,” says Brincat. “Instead of just checking in with patients at 6 and 12 weeks to see how they’re doing, surgeons can add digital monitoring to their care plan and be more targeted in their interventions.”

Envisioning future innovations powered by Zimmer Biomet’s AI framework and Microsoft Azure

Zimmer Biomet continues to innovate new medical products and services based on its AI framework, creating opportunities for surgeons to help patients achieve better outcomes through smart technology and data analysis. The company’s confidence in the security and reliability Microsoft provides allows employees to focus on creating products to help clinicians and patients. “We’re really keen to make sure that what we build is scalable and flexible for all the different types of products that we are deploying, now and in the future,” says Griffiths. “Microsoft is there, making sure that the underlying infrastructure is up and running all of the time.”

Zimmer Biotech credits its AI framework with helping it revolutionize the industry, and the company is seeking new ways to improve patient outcomes through objective data. “With WalkAI, we’ve created the first AI-based predictive analytics model in orthopedics,” says Ivan Tornos, President and Chief Executive Officer of Zimmer Biomet. “We have a relentless focus on investing in innovation that will enable our customers to deliver best-in-class care for the patients they serve and breakthrough technologies like WalkAI help move our mission forward.”

Patients must have Internet access and a text-capable mobile device or a compatible smartphone to use mymobility; not all smartphone app features are available with web-based version. Not all patients are candidates for the use of this product and surgeons should evaluate individually to determine which patients are appropriate for therapy at home.

Patients must have a compatible smartphone and undergoing a hip or knee replacement to use WalkAI™.

MEDICAL DEVICE DISCLAIMER. [Microsoft product(s)/service(s)] (1) [is/are] not designed, intended or made available as a medical device(s), and (2) [is/are] not designed or intended to be a substitute for professional medical advice, diagnosis, treatment, or judgment and should not be used to replace or as a substitute for professional medical advice, diagnosis, treatment, or judgment.