Three key factors have been driving this transition: Consumer demand, COVID-19, and the increasing sophistication of the technology itself. According to research from analyst GWI, while consumers still consider fitness tracking the leading reason for owning a wearable, nearly half of all owners also expect their wearable to monitor their health. One in five of us also value its ability to help doctors manage our health.
In March 2020, the U.S. Food and Drug Administration (FDA), a body ensuring the safety and efficacy of medical devices, issued a new policy allowing manufacturers of FDA-cleared non-invasive, vital sign-measuring devices to expand their use so healthcare providers could use them to monitor patients remotely. This included wearables that measure body temperature, respiratory rate, heart rate, and blood pressure.
This saw the rapid development and deployment of numerous 'ehealth' wearables. These devices provided a means to alert authorities about the likelihood of COVID-19 infection without requiring individuals to present in person for diagnosis by already stretched medical facilities. The wearables could also generate rich data sets that enable medical professionals to offer personalized health decisions, make earlier diagnoses, and ensure prescription adherence for many other health care purposes.
The medical profession is widely supportive of the potential of wearables to support their work to enable better diagnoses and ongoing treatment. The question, though, is who will sift through and analyze the massive amount of data such devices collect for doctors to make clinical sense of it?
The answer lies in the ability of wearable technology to do heavy computational work on behalf of the medical profession. Today's wireless SoCs and SiPs that power advanced wearables are well equipped to support the machine learning (ML) algorithms needed to wade through huge volumes of data and rapidly establish anomalies that could indicate underlying health issues. In so doing, the chips have the potential to allow wearable developers to explore ways to selectively provide doctors with only the information they need to make rapid, accurate clinical decisions and ignore extraneous 'noise'.
Edge Impulse is a specialist in TinyML, a scaled-down form of ML suitable for IoT edge devices. Edge Impulse's partnership with Nordic Semiconductor allows developers to benefit from easy-to-use ML features as standard on Nordic's nRF52 and nRF53 Series multiprotocol SoCs and its cellular IoT-based nRF9160 SiP for medical and other applications where relaying data directly to the Cloud via the cell network offers obvious advantages.
Embedding ML and big data capabilities on these tiny wireless chips will enable wearable devices to move beyond basic wellness or step-counting applications toward multifunction medical solutions. They could also equip healthcare providers with informed yet selected diagnosis, treatment, tracking, and disease prevention data.
There is also a growing body of research to demonstrate that when AI makes better decisions than humans—by solving problems differently from us, recreating human trial and error, and then checking our work over and over again—it will also make our own decision-making better by exposing us to different ways of processing and understanding data. Such research offers intriguing possibilities for the AI- and ML-powered wearables of the future by allowing them to support medical professionals and improve their decision-making.