AI has the power to revolutionize industries from finance to cybersecurity to transportation and healthcare. The latter is particularly important because it has a direct impact on human health and longevity. At the Fortune Brainstorm Health conference earlier this month, health experts such as the Mayo Clinic CEO Dr. Gianrico Farrugia and Zebra Medical Vision CEO Eyal Gura discussed how AI and robotics are already disrupting healthcare today.
For example, Zebra Medical Division has AI solutions that improve diagnoses with programs that detect breast cancer in mammograms and fractures in X-rays, and also help the admin-side of care by automating triage decisions.
The promise of AI in healthcare is exciting, but to fully feel the revolution, it must be accompanied by a software revolution.
The Digital Healthcare Revolution Relies on Medical Devices
At the heart of this healthcare revolution are connected medical devices. There's no breast-cancer detecting AI without the mammography machine that collects and shares the data, and no fracture-sensing algorithm without connected x-ray machines supplying x-rays. As of yet, medical device software isn't keeping pace with the changing business model of healthcare.
As hospitals and clinics invest in connected medical devices, they often find insufficient software that fails to anticipate all use cases of that medical device. They either rely on limited native apps, or find external apps that may be insecure, inaccurate, or poorly written.
In other words, while connected hardware is proliferating, the software is still stunted and often inadequate for many end-uses of a given device. Companies like Zebra Medical Vision are making an impact, but is it large enough to meet the full demand?
Surely not. And that isn't Zebra Medical Vision's fault. The medical device industry has the greater power to change the software landscape in health. It's time they took a page from the smartphone industry by opening up their device APIs to third-party apps.
If medical device manufacturers opened up their devices, then hospitals, clinics, and private practitioners could shop for the best app that suits their needs. On the flipside, software developers like Zebra Medical Vision would have instant access to more and more machines for which they may develop apps. And, just as important, it would become easier for new medical software developers to bring their apps to market.
Importantly, they would do so in an app store where all apps meet FDA regulations for medical device software and would be assured of quality and security. This is central to the medical device app store's success.
The Fortune Brainstorm Health Conference dedicated an entire panel to medical data security and concerns that AI can learn biases and bake them into diagnoses and care. A medical device app store would impose conditions and rules for software and AI that can be updated and disseminated across the industry. A central rulebook would go a long way in solving data privacy and security concerns.
Embracing a Software Developer Platform in Healthcare
A platform is not to be confused with a data cloud. The data itself is valuable, but it isn't sufficient to deliver what medical practitioners need. They need applications built on top of that data. The platform will unlock the value of patient data by putting data-analyzing apps at the doctor's finger tips, not as raw data but as actionable insights and recommendations.
In a way, the medical device industry will take a page from the history of smartphones. In the late 1990s and early 2000s, cell phones competed on quality and product specs, until Apple's and Google's app stores changed the nature of the competition. Before the innovation of the App Store (which is an app development platform), Blackberries were the leading phone on the market. Spec wise, Blackberries were great phones, and their proprietary apps were well reviewed, especially Blackberry Messenger. Yet, once Apple and then Android opened their app stores to third-party developers, Blackberry couldn't compete, because consumers didn't want to be limited to Blackberry's suite of products. Eventually, Blackberry was forced to open its own app store, but it struggled to gain traction in a market that had already left it behind.
App stores can also impose rules around data collection and usage, code specifications, and other terms of service (ToS) and use. In the medical device industry, software must be compliant with FDA rules. An app platform built with the FDA's regulations at the center of its ToS (or better yet, with the FDA's cooperation and counsel) could perform an important security and quality assurance function by only accepting apps that meet the FDA's regulations.
In the process, medical software developers will become familiar with how medical software must be structured. In the same way that Apple taught developers how to develop for smartphones, the medical device app store teaches developers how to develop for medical devices, and FDA rules should be at the core of that.
AI is a powerful tool that can reshape the healthcare industry. However, without an equally powerful distribution method, it will be a powerful tool in the hands of too few.