Simon McIntosh-Smith, Professor of High-Performance Computing at the University of Bristol, discusses what the Arm-based Isambard supercomputer—and its peers—could help us achieve in the years ahead
As the first all-virtual Consumer Electronics Show (CES) approaches, digital health is close to the top of the agenda. With 31 of CES 2021’s sessions dedicated to the topic of digital health, it’s the 3rd most popular topic behind artificial intelligence (AI) and autonomous vehicles. It’s 5th on the list of exhibitor topics, too, with over 400 companies set to showcase their digital health technology at CES 2021.
Without COVID-19, the story may have been very different. At the beginning of 2020, Accenture’s 2020 Digital Health Consumer Survey revealed that growth in consumer digital health adoption had fallen off a cliff. Use of medical devices that collect information about our fitness, vitals and lifestyle had dropped from 33 percent in 2018 to 18 percent in March 2020 — the lowest in four years.
The pandemic, and the resulting global stay-at-home orders, demonstrated the enormous role that technology had to play—from keeping us in contact with our loved ones to connecting us with doctors via our devices through rapidly-deployed telehealth solutions. According to McKinsey, telehealth has surged under COVID-19.
As the heightened risk became clear for those with certain pre-existing conditions—including obesity, high blood pressure, asthma and diabetes—it has also been a wake-up call for many to take greater care of themselves.
Wellness vs healthcare devices
I don’t doubt that this lead many to dig in their bedside cabinets for discarded fitness wearables and other wellness devices in an effort to reclaim agency over their health. But while such devices are fantastic tools in providing reassurance and enabling the tracking of gains over time, there is a very significant distinction to be made between wellness devices and digital health devices with the certified accuracy required to collect clinically interesting data.
From previous experience trawling the halls of CES, a fair proportion of devices touted as ‘digital health devices’ are really wellness monitors. Notable exceptions have stood out for me in recent years, such as Gyenno’s smart spoon at CES 2016 that counteracts hand tremors from Parkinson’s Disease while collecting data about tremors in the cloud, or Respiro—a smart asthma inhaler I wrote about on Arm Blueprint back in November 2018 and which was featured as part of our Arm technology showcase at CES 2018. Then last year at CES 2020 we saw devices such as Add Care’s Glutrac wearable glucose monitor blur the lines between FDA-approved medical functionality and the kind of ergonomics previously only found in smartwatches.
It’s this kind of device that I very much hope we see more of at CES 2021 next week. When it comes to monitoring their health, consumers don’t need another fitness tracker—they need a digital health device they can use to monitor their health in the knowledge that the data being recorded is medically sound.
Sadly, building a medically sound digital health device is not as easy as it sounds—and much of the inherent complexity comes down to two interwoven factors: stringent regulatory standards and proprietary technology.
It’s traditionally been a whole lot easier to build a wellness device than it is to build a certifiable digital health device suitable for use in both home and professional healthcare environments.
Stringent standards and proprietary technology impede innovation
Within the healthcare industry, any instrument, machine, implant and in vitro diagnostic product (IVD) used in the diagnosis, prevention and treatment of diseases or other medical conditions is held to the most stringent standards. That applies to hospital-based digital health devices but also to any digital health device intended for home use provided to a patient by a medical professional.
The approvals process to take a medical-grade digital health device into the mass market presents a high barrier and approvals are also region-specific which adds to complexity and costs.
A key issue with many digital health devices is that the way they communicate to the cloud and present their data to the clinician is often proprietary. To date, many digital health solutions on the market use proprietary software in the device, cloud and app.
As a result, they rarely interface into clinical medical records or provide triage of normal and abnormal readings. Those that can be implemented on a technology level may still present an unnecessary learning curve. Telehealth may have become much more commonplace this year, but the fact remains that should a doctor wish to use data collected by a digital health device during a consultation, they would need to learn whatever proprietary interface that device comes with.
Furthermore, digital health devices may use any combination of Bluetooth, Wi-Fi and cellular to deliver data. These communication interfaces create sufficient complexity in communications, security and file systems. In many cases, digital health device manufacturers lack confidence when adding such connectivity to products due to data and security areas they are unfamiliar with.
The solution? A platform approach
The key issue here is that digital health devices are still being built from the ground up using proprietary technology, with software instructions executed directly on hardware—what we call bare-metal programming. This leaves little to no room for standardisation between devices.
By enabling device manufacturers to move from bare-metal programming to building upon established embedded platforms and operating systems such as Arm Mbed, we can ensure that the devices they build are inherently trusted, secure and unified in their communication of data.
Arm has been part of the mobile phone industry for 30 years, where platforms, standards and architectural choice has fuelled innovation and driven down costs. The consumer medical industry is now adopting the same approach, ensuring that new digital health devices are far more easily implemented into both a home and professional healthcare environment.
A platform approach means it’s also far easier, quicker and cheaper to build a digital health device—be it a consumer wearable or something designed only for clinical use. But it’s also vital in expediting the approvals processes associated with enabling new digital health devices and their data to be used by medical professionals.
Jumping the regulation hurdle with cloud
Companies such as Cambridge-based L2S2 are working with Arm to offer compliant reference hardware and firmware designs based on Arm Mbed. These enable device manufacturers to build digital health devices that are compliant with L2S2’s Managed Medical Device Cloud (MMDC), a cloud data platform that aggregates medical data from a range of home digital health devices without the need for those devices to have been certified or accredited by each medical facility.
It’s this kind of joined-up thinking that will enable a new wave of digital health devices, and I’m personally very interested to see whether this is visible in some of the new digital health devices being showcased at CES 2021. By choosing the right OS, device management solution and health data platform, digital health device manufacturers can enable rapid development and integration.
Perhaps the world of digital health technology is finally getting close to the holy grail of consumer-friendly and digitally enabled medical devices in which patient data is more secure, no longer siloed and innovation can keep pace with patient needs and outperform the highest regulatory challenges. It’s that which we need if we want to see real benefit in fighting viruses such as COVID-19 using home digital health devices.
Mbed: The trusted OS for IoT
Arm Mbed is a free, open-source Internet of things (IoT) operating system that includes all the necessary features to develop smart, connected products on Arm Cortex-M based hardware. It’s the platform you need to quickly develop trusted digital health solutions for both consumer and professional use.