A few years ago I attended a virtual oncology congress as a speaker where one of the attending professors asked me about the IT infrastructure in medical practices. Knowing the medical reality, he inquired whether those existing computers and accompanying systems would allow the use of advanced technologies such as artificial intelligence (AI). Practically, he wanted to know how I think existing hardware should be upgraded to make it ready for AI.
To answer him, I made a seemingly unrelated analogy coming from the video game industry: cloud gaming. This approach, where the cloud handles all the processing power, has gained traction; bypassing the need for powerful PCs and consoles altogether. I told him that I believe the same fate awaits medical professionals and hospitals with limited IT infrastructure that still want to benefit from power-hungry technologies like AI.
And by now, reality has caught up with this prediction: CareCloud already uses Google Cloud to help power its own operational needs and enable generative AI and search functionality in its products.
By using Google’s offerings such as Vertex AI and Generative AI App Builder, the company is developing tools to bring physicians at smaller practices the same AI-enabled capabilities that are increasingly available at large hospitals and health systems.”
Physicians working in smaller practices could ask generative AI more complex clinical questions and gain deeper insights with evidence-based recommendations derived from analyzing datasets.
And it all makes sense. Rather than upgrading physical hardware, cloud computing technology can upgrade healthcare institutions, and provide access to advanced technologies; while circumventing the need for physical health IT upgrades. Such an approach has not been explored at large, but it holds promise for the future of healthcare as we shall see in this article.
From the moons of Mars to the cloud
For the majority of its existence, the video game industry relied on consumer hardware. The bigger the storage, the better the graphics cards and the faster the processors that gamers’ PCs and consoles could pack, the superior performance they would get, allowing them to play the latest games with more features. This bigger-better-faster approach extended past the gaming machines to the games themselves. This software evolved from being stored in kilobytes on floppy discs and cartridges to being burnt on multiple CDs and Blu-ray discs that store multiple gigabytes.
For instance, id Software’s iconic 1993 game Doom – where you patrol the moons of Mars – took up some 2.39 MB of storage and fit in only one floppy disc. The latest instalment in the series released in 2020, DOOM Eternal, boasts better visuals, and lets you roam the surface of Mars but has a file size of up to 50 GB.
However, in recent years, video games have increasingly bypassed storage media hardware altogether in favour of cloud storage. No need for physical copies, your games follow you wherever there’s an internet connection. Platforms like Epic Games Store and Steam serve as distributors for these digital versions of games, and digital sales from such platforms first outsold physical sales in 2020.
Now, even gaming systems are going digital with cloud technology. With services like Google Stadia and Amazon Luna, gamers don’t need high-end PCs or new consoles to play the latest games, which would traditionally require such devices. In fact, with a high-speed internet connection, they can play those demanding games like DOOM Eternal, on a supported, yet underpowered device they bought a couple of years ago. All this is made possible thanks to the power of the cloud, which can similarly give a boost to existing healthcare infrastructure.
Cloud magic
Cloud gaming works through cloud computing technology. Rather than having the user’s computer handle all the processing to play a demanding game, it’s done remotely via servers which provide the computing power to the user and relay it via the net, or “the cloud.” It’s basically having powerful, distant PCs do the computing, while you benefit from the same software output on an underpowered system. You might even think of it as a beefed-up Netflix. Cloud gaming subscribers choose the game they want to play on a supported device anywhere with a robust connection and the servers will run the game while the player remotely controls it.
It’s particularly handy for those who want to play a new game but can’t afford to buy a new console to play it on. They just use their cloud subscription on a device they already own and voilà, new software on older hardware. And that’s exactly how this technology can benefit healthcare.
WHO officers in Africa noted in a perspective article that unreliable infrastructure is among “the greatest challenges to the full realization of the benefits of [digital health]”. The continent is already a hotspot for digital health, but the infrastructure might limit the adoption of more advanced and helpful technologies like AI. These require newer systems to operate, which might not be so widely available. So, instead of fitting medical institutions with costly, top-of-the-line computers to run those, how about outsourcing the processing power and providing it on demand?
This is what’s possible with cloud computing, and it circumvents the whole health IT upgrade process in healthcare institutions with limited IT infrastructure. The technology isn’t totally foreign to healthcare; systems already employ cloud technology from Amazon Web Services and Microsoft Azure. This helps develop chatbots and health platforms to more securely manage the increased data capacity that comes with the digitisation of healthcare. In fact, health systems spent almost $40 billion in 2022 on cloud computing. However, these haven’t been employed at a large scale for demanding technologies like AI.
AI enters the game
In September 2020, a first step towards such an offering was made. Google Cloud partnered with the U.S. Department of Defense (DoD) to develop an AI-based digital pathology system. It involved augmented reality microscopes to help physicians, assisted by an AI to improve the accuracy of cancer diagnoses.
While this is just one example, potentials of the technology in healthcare are manifold. These range from optimising logistics to digesting medical and research data at unprecedented rates to assist professionals in decision-making. By offloading the computing power required to cloud services, more institutions can benefit from the technology.
Moreover, cloud infrastructure can provide more computing power than would be possible with local hardware. Thus, demanding analyses of large amounts of data can be done more efficiently, which will in turn help better train AI models to improve their accuracy.
The flexible resourcing aspect of cloud-based processing infrastructure also enables healthcare institutions to leverage AI’s potential in more ways than they could if they ran it in-house. They could use the technology to streamline tasks such as triaging with chatbots, analyse health tracker statistics and offer personalised recommendations; while physicians focus on tasks that require professional medical attention. All this will be possible without upgrading the local IT infrastructure.
Before elevating healthcare to the cloud
Just like gamers are now buying affordable processing power from the cloud to play high-performing games without spending much more on newer hardware, similar possibilities open up for healthcare institutions. But while hardware might not need upgrades, other factors will determine its successful implementation.
Similar to cloud gaming services, a high-speed internet connection will be a must, in particular for demanding AI processes. The WHO article further notes that unstable and expensive internet connectivity and unstable power supply can’t sustain advanced technologies. With 5G around the corner, one of these variables could be addressed.
For those medical professionals and hospitals with limited IT infrastructure willing to benefit from advanced technologies, cloud-based infrastructure presents an adequate solution. It holds the potential to boost the adoption of tools like AI and skip the need to upgrade IT infrastructure in medical practices. As such, preparing for such adoption by ensuring adequate internet connection and power supply will help healthcare reach the cloud faster.
The post Cloud Gaming Can Bring Artificial Intelligence To The Doctor’s Office appeared first on The Medical Futurist.