Developing Innovations and Solutions for Healthcare’s Future: A Conversation with AMD’s Subh Bhattacharya

“Our team is responsible for growing our presence in this space … and also developing and evangelizing innovations and solutions for the future,” says Subh Bhattacharya during his discussion with Clarius CEO Ohad Arazi on the Medical Imaging for All podcast. Read on for more highlights.

For Subh, his twenty-five-year career in the technology industry started in computer engineering. Before moving into the healthcare space, he designed microprocessors for Intel, then expanded into product management, marketing, and servers. After a “sharp move towards the healthcare side,” Subh built a modular healthcare information systems platform, and spent some time in pharmaceuticals. All this to say that when Subh joined AMD, one of the largest semiconductor companies in the world as the Lead of Healthcare & Sciences at AMD, he had the background in how to build the devices that health care practitioners need the most.

Building Today’s Medical Devices

One of the questions Subh gets a lot is « What is AMD, a semiconductor company, doing in healthcare? » At the very heart and brain of medical devices and diagnostic equipment are the chips and processors made by AMD that make them work. They create the power behind MRI and CT imaging, ultrasound imaging, surgical robotics, AEDs, and defibrillators, as well as image processing, visualization, graphics, and other technology.

But to provide quality devices that can deliver cutting-edge patient care, Subh explains that there are a number of considerations that device makers need to consider. As devices become smaller, how do you also keep them running faster? How do you keep them secure? How do you keep them processing in real-time? « A lot of innovation then has to go into that, not only to have strong processing, but do it in a very space-, power-, and heat- efficient manner. Is that also a big constraint that you see in this space? Absolutely,” he explains. “These devices that are at the edge are smaller appliances. Performance over power is what typically I would say is one of the most important measures now. » This includes innovating on what machines can do at the chip level to facilitate real-time speed, performance, and capabilities.

Integrating Artificial Intelligence

One of the biggest advances in recent years is AI, which is finding a broad variety of use cases across healthcare. In addition to its ability to offer assistance, it’s also relieving radiologist fatigue. « For example, what we see with diagnostic assistance, x-rays, huge volumes — something like 2 billion scans a year. And there have been studies showing that it’s not the question of expertise, it’s a question of fatigue and not enough radiologists. So actually doing some sort of ROI, region of interest selection of x-rays using AI as an assistant is definitely a huge benefit,” he says, explaining how AI is “cutting down hours and hours … AI gives the results in a few minutes. It’s just amazing.”

AI has a number of applications on the surgical side as well. For example, AI can sound an alarm if you’re cutting a blood vessel by mistake, « or if a surgeon is looking at all these images during the surgery and you use AI to choose the right filter. A task that does not need to be done by humans and it definitely doesn’t pose a risk towards it. »

However, for AI to continue expanding its influence, it needs more data to improve its accuracy. « It’s still a little bit difficult when it comes to people trying to do research and having access to large amounts of data. And one of the things you have to do is to increase the accuracy,” he states. “So increase the accuracy means basically, it has to be the volume of data, volume of good, curated, labeled data sets » to train AI so that it can increase its ability to assist and grow its use cases.

The Future of Healthcare

Advances in healthcare technology will continue to develop, as will AI integration, but for Subh, the future of healthcare lies in protecting what that technology can do. « One of the areas that I feel we’re still a little bit behind is cybersecurity and implementation of cybersecurity, all the way from the hardware level to the network level. And then lack of regulations or lack of enforcement there has created a little bit of a mismatch there as well. »

That also includes more consideration around the convergence of IT and OT as more medical devices and equipment appear. Subh explains that « there is going to be a lot more technology developed from the node perspective there in terms of security, safety, availability at that level. And when you see the equipment makers, their internal customers are the service providers who are doing some level of system integration for hospitals and IT are also defining what it is needed. And their requirements may or may not be matching in terms of what they want from the security perspective. »

What is AMD doing to increase security in their devices? « We’re starting to address it from the root of trust and from the hardware from the base level,” he explains. “There are many ways to do encryption/decryption from the hardware perspective. We have a Security Engineering Center of Excellence within AMD which addresses specifically each market, providing a solution that is more end-to-end or bottom-to-top. »