“Information is the oil of the 21st century, and analytics is the combustion engine.”

Peter Sondergaard, Gartner research

I have the very special privilege of giving a talk at an upcoming meeting at the Harvard/MIT Center for Regulatory Science and its annual Global Conference: Medical Devices and Digital Health. My talk will focus on leveraging artificial intelligence for medical devices. I will draw on an earlier opportunity that I had to be involved in a project of pediatric ventricular assist devices (VADs); there were myriad challenges in gathering data – both during and after the hospital stay.

Medical devices range from the traditional – such as blood pressure measuring devices – to cutting edge ones such as portable EKG devices. While class I are low-risk devices (non electric wheelchairs), classes II (CT scanner) and III (pacemaker) have higher risks. Several innovations in evidence generation for devices, including premarket approval and postmarketing monitoring, will be essential for the future in clinical trials of medical devices:

Multimodal artificial intelligence

The health data from sources such as genomic data, radiological data, and other real world data (RWD) sources will continue to increase exponentially in volume as well as in sophistication. As we mentioned a few weeks ago, RWD consists of data sources that are not in conventional randomized controlled trials. Just like healthcare data in general, however, there are challenges of RWD that include lack of standardization of quality and type of RWD, fragmentation of RWD sources that can lead to access issues, and accuracy and reliability of RWD, not to mention the privacy and security concerns.

Federated and swarm learning

For medical devices and clinical trials, it is essential for these devices to be able to communicate with each other in the form of federated or swarm learning so that all the devices can learn from each other. This wisdom of the devices as a group (vs the human crowd) can be facilitated with an edge node that will obviate the need for the traditional registries.

Embedded artificial intelligence

In the near future, primitive artificial intelligence tools in the form of neural nets (“tinyML”) will be embedded in the medical devices to generate insights into the performance and efficacy of these devices.

In addition to the aforementioned artificial intelligence strategies for medical devices, pediatric devices have additional challenges such as growth and development of the patients while the devices are being utilized.

We are very excited to welcome you to attend in-person the AIMed22 Annual Global Summit January 18th-20th, 2022, at the sublime Ritz-Carlton resort in Laguna Niguel, southern California. This summit promises to be the most exciting yet, with Drs. Eric Topol and Daniel Kraft among the keynote speakers. We are all very much looking forward to seeing and learning from each other in person for human-to-human conversations and networking at this event. Book now here.