
Medtronic Chief Technology and Innovation Officer Ken Washington [Photo by Hardy Wilson for Medical Design & Outsourcing]
“We’ve got work to do to unlock the revolution of AI in medical devices,” said Medtronic’s chief technology and innovation officer.
Leaders at the world’s largest medical device company think artificial intelligence combined with medtech could enable the innovation needed for personalized and effective healthcare.
“We have to activate the future of healthcare with AI and digital connectivity and technology. And it’s not going to be easy. But we know what the reward is. The reward is the vision that many people have talked about for a long time, and we know it’s coming, but it’s not here yet. And that’s the vision of personalized digital wellness,” Medtronic Chief Technology and Innovation Officer Ken Washington said during the Design of Medical Devices Conference at the University of Minnesota this week.
Washington noted that the doctor’s office hasn’t changed much since the early 2000s. It may even be worse because the doctor is now focused on a computer screen displaying medical records rather than the patient. Meanwhile, industries where Washington used to work, such as retail, automotive and defense, have undergone major digital transformations.
Guiding healthcare through a digital transformation is not going to be easy, he acknowledged. “The vision for getting to this personalized healthcare is going to take steps.”
Medtronic already has several products that utilize AI: the GI Genius system for AI-enhanced colonoscopies, the Touch Surgery Enterprise for AI-enabled surgery, the AiBLE digital ecosystem for neurosurgery, the MiniMed 780G insulin pump system, and the Linq implantable cardiac monitor.
But Washington also sees challenges that need to be overcome if Medtronic and other medical device companies are going to enable AI to reach its full potential. Here are three that stood out from his talk:
1. Bringing order to the data
Washington said people tell him that Medtronic has an advantage in using artificial intelligence in healthcare because it is sitting on so much data from its devices. However, he compared the data to a big messy pile of Lego bricks. He described getting the data sorted as 80% of the work to make AI function around a medical device. For the AI to use it, those pieces of data need arranging and proper presentation. They need to be explained with a story.
“The data readiness work that we have to do is significant, but we know how to do it, and we just need to get on with it.”
2. Filling gaps in technological readiness
The biggest gap that Washington sees is the need to develop medical-grade, embedded tensor processing units. TPUs are the application-specific integrated circuits that Google developed for neural network machine learning.
“We’re talking to multiple chip companies about how to work with us to build out medical-grade reasoning and training chips. Nvidia has made a big step in that direction in building Holoscan. I’m a big fan of that platform. GI Genius is built on top of Nvidia Holoscan. So was Touch Surgery Enterprise. But we need to take it to the next level. They’re not motivated to make that happen because of the scale of our industry. … Being able to embed a better training chip into into a Linq device or a pacemaker or neurostimulator or a DBS module — I’d love to have that kind of capability available at scale at a reasonable cost so that we could build next-generation products and devices that are intelligent that have medical-grade devices in them.”
3. Regulatory hurdles
“There are all these complexities around payers and policy changes and the regulatory environment. Regulators are not yet ready to embrace some of these new technologies. We have to bring them along. This is why I’m encouraging all of my colleagues and my team to be active in the regulatory space as well as in the technical space.”