A high-tech demonstration of a surgical procedure at a hospital in the small city of Nieuwegein in the Netherlands holds the promise of reducing operations backlogs, as well as making them more affordable. It resonates from the besieged National Health service in the United Kingdom to the crowded wards of South Africa.
At the St Antonius Hospital, a non-academic Dutch teaching facility, Gadget was allowed to view the live use on a patient of an AI system for a “transcatheter edge-to-edge repair” (TEER), a procedure to repair a leaking heart or mitral valve through a small cut, avoiding major surgery.
The operation used the Philips DeviceGuide, a system that combines live X-ray and ultrasound imaging to help doctors track the position and direction of a repair device inside the beating heart. The system moves away from flat, two-dimensional imaging towards a three-dimensional view during procedures.
It also allows doctors to perform complex heart procedures faster, with the aim of improving access to “minimally invasive” treatment for patients who may be too frail for open-heart surgery.
“For me, the biggest benefit is efficiency,” said Dr Martin Swaans, a prominent heart specialist at St Antonius who was involved in the development and use of the technology. He is a world leader in interventional imaging, which uses advanced imaging technology to guide complex heart procedures.
He told Gadget: “It improves communication within the team and it forces us to make a detailed procedural plan before we start. Instead of simply beginning the procedure and deciding everything along the way, we first sit down together, analyse the anatomy and define exactly where we want to position the device. Those planned trajectories are then displayed during the intervention.”
He said the system improves image quality during the procedure, helping the team maintain the best possible view while the device is being guided.
Dr Leo Timmers, who also performs mitral valve repair procedures at St Antonius, said the DeviceGuide gives doctors clearer direction during navigation.
“From my perspective as the operator, it’s immediately obvious where I need to steer,” he said. “The improved image quality also allows me to control the procedure more precisely. Overall, the procedure becomes faster, more efficient and ultimately safer for the patient.”
Procedure times have fallen sharply since 2009. Repairs that once took three to five hours had already been reduced to about an hour before DeviceGuide. Now, many can be completed in 30 minutes with the AI-guided system, depending on complexity.
Swaans said: “Much of that comes from eliminating additional imaging checks. Previously, I would repeatedly verify whether both leaflets (heart valve flaps) had been captured correctly. Now I can already see that while advancing the device.”
The AI system relies on two additional technologies form Philips: the Azurion image-guided therapy platform and EchoNavigator, which fuses live X-ray and echocardiography images. Deep-learning AI technology automatically tracks an implant through the beating heart, using both imaging methods.
“In the early development phase the AI occasionally became confused because it had only been trained on relatively limited data,” said Swaans. “Today, with the commercial release, that almost never happens.
“Occasionally, if another device is already present inside the heart, recognition can become more challenging. However, the system is now very good at recognising exactly which device is attached to the catheter. The AI has matured enormously.”
DeviceGuide does not remove the need for specialist staff or advanced equipment, but the system could help more hospitals perform complex procedures with greater consistency.
Philips says DeviceGuide represents a shift from AI as a diagnostic support tool to AI as part of live treatment. That shift could help doctors complete procedures faster, reduce the strain on specialist teams and expand access to minimally invasive care for patients who may otherwise have limited options.
Özlem Fidanci, Philips global head of international markets, said the company was preparing AI technologies for use across markets, but access would depend on local requirements.
“Not everything is relevant everywhere,” she said. “We are preparing our innovations on AI technologies, and we are bringing all our technologies everywhere as much as possible, based on local regulations and requirements.
According to Fidanci, healthcare systems are facing rising demand from ageing populations, chronic disease, staff shortages and delayed access to diagnosis and treatment.
“Today, the greatest need we see is for healthcare that is more sustainable, more accessible and more affordable.”
* Arthur Goldstuck is CEO of World Wide Worx, editor-in-chief of Gadget.co.za, and author of “The Hitchhiker’s Guide to AI – The African Edge”.