Nvidia and Johnson & Johnson to build AI applications for surgery

Johnson & Johnson on Monday announced it is working with Nvidia to develop and scale new artificial intelligence applications for surgery. 

J&J’s MedTech unit and Nvidia plan to integrate AI within devices and platforms from pre-op to post-op to help ensure that surgeons have access to all the information they need, Nvidia’s vice president of health care Kimberly Powell said. For instance, the companies are using AI to analyze surgical video and automate the time-consuming documentation required after a procedure. 

“There’s an ability to use all the sources of data inside an operating room, whether it’s your voice, or whether it’s the video coming from a camera inside the body, or elsewhere, to take advantage of the generative AI moment that we’re in,” Powell told CNBC in an interview. 

The MedTech unit at J&J creates tools and solutions for conditions such as heart failure, kidney disease and stroke, and its technology is used in more than 75 million procedures each year, the company told CNBC. Powell said Nvidia has worked in medical devices and imaging for more than a decade. 

Shan Jegatheeswaran, vice president and global head of digital at J&J MedTech, said just one minute of surgical video is equivalent to roughly 25 CT scans, so having the compute power and infrastructure to annotate and share those videos widely will be powerful for surgeons.

In the short term, he said de-identifying and enhancing the video can help educate and train surgeons. In the long term, analytics can be layered on top of video to provide real-time decision support. More accessible surgical video means residents will not have to solely depend on the insight and availability of the more experienced physicians at their institutions. 

“Think about athletes. They look at game tape, and they get better over time as they look at themselves,” Jegatheeswaran told CNBC in an interview. “That’s sort of the starting point. That’s the holy grail in the short term.”

Powell said the collaboration is in the “early innings,” and many applications will take time to fine-tune and implement safely. However, she said nondiagnostic use cases such as automating paperwork will help save surgeons time and make a difference “right out of the gate.”

“I think all of us as patients should get really excited about the fact that this kind of technology is going to be able to enter in and be within reach of all the clinicians and all the hardworking nurses and all the health-care staff,” Powell said. “They’re going to have the very best tools and information at their disposal.”