For several years now, millions of people have grown used to waking up each morning to their wristwatches offering an insight into their health, with features as advanced as letting them know if they’re at risk of heart failure. 10 years on from the first smartwatch boom, patient-led health monitoring is par for the course.
You’re no doubt familiar with Fitbits, Garmins, Apple Watches and other such devices. But these mass-market wearables are simply the most visible aspect of a bigger technological union. Healthcare systems are becoming enmeshed with a model of software design called edge computing (EC).
Understanding edge computing
EC is not a technology itself, but a model of software architecture. Rather than a cloud network, where devices send information to be processed in far-off data centres, the computational journey takes place locally.
Devices collect data and then communicate with server devices nearby. These servers process that data themselves, before uploading it to cloud servers. This approach slashes latency, allowing information to be harvested and processed in real-time.
You may not have heard of EC before, but you’ve probably heard of the Internet of Things (IoT). While the two are not exactly synonymous, EC forms the basis for many of the most-used IoT systems. Sensors on devices like your doorbell pick up information and ping it over to servers housed in nearby objects to make useful sense of it.
So far, so technical. But the implications for holistic healthcare provision are vast.
Healthcare on the edge
As devices shrink in size and cost, EC's footprint in healthcare is expanding beyond bedside medical monitoring.
EC has made home-based care more responsive and effective. But more significantly, it's widened the health applications of wearable devices in and out of the home.
Many of these technologies have been in place for decades, while others are newer. We can think of them as making up an Internet of Medical Things (IoMT).
Broadly speaking, there are three use categories for wearable EC technology in healthcare.
- Harm prevention and maintenance of health
- Mass-market, wearable products encourage healthy behaviours in an ageing, more obese population.
The Fitbit model of physical self-monitoring is a useful tool for general weight management. But wearable devices can also help mitigate acute health challenges. For example, when the Apple Watch detects hard falls, it gives wearers the option to call the emergency services.
- Patient management
Wearables give patients with chronic ailments more autonomy in managing their conditions.
For instance, patients with chronic obstructive pulminary disease (COPD) are commonly prescribed pulmonary rehabilitation exercises. Wearable sensors have been found to enrich this process. Meaningful data empowers patients and helps clinicians be more accurate.
IoMT wearables are also proving increasingly useful in motor therapy for stroke patients.
- Disease management
EC devices offer clinicians a powerful tool to remotely monitor and treat patients with serious conditions.
IoT-enabled Wearable ECGs allow cardiologists to keep an eye on patients’ heartbeats.
Diabetics now benefit from hybrid systems for detection and treatment. Smart insulin pumps adjust delivery according to data provided by an accompanying monitor. Patients can read all this data through a synchronised mobile app.
Clearly, the medical uses of wearable devices have long eclipsed counting your 10,000 daily steps. But we’ve probably barely scratched the surface.
The shape of things to come
It’s become a hoary cliché when writing about tech to predict that artificial intelligence (AI) will revolutionise this or that sector. But in healthcare, the hype may well be justified. AI and machine learning (ML), are fusing with the IoMT to expand healthcare systems’ capacities. The activating ingredient? Data, and lots of it.
As EC links up the built environment, it creates vast agglomerations of data. There are countless objects that, if made ‘smart’, could become important data sources for healthcare and public health. These include clinical devices like stethoscopes and pulse oximeters. But they also include everyday objects like gym equipment. Put these together and you have a network generating millions of data points in real-time.
The potential uses of these data are even further reaching than their sources, not least for public health. AI applications could use this mass of deidentified, real-time data to identify or even predict the next epidemic. Medical authorities would have a bird’s eye view of regional health discrepancies. Aggregated, segmented data could also help pharma companies gauge real-world treatment efficacy.
At the patient level, self-updating data profiles for individuals would revolutionise prevention and early detection of illnesses. Blood pressure reaching dangerous levels? Your phone will tell you before it gets out of hand. A Covid infection brewing in a clinically vulnerable patient? Their wearable device will identify the key signs in time for the patient to be put on a course of antivirals.
The holy grail would be a closed-loop system, capable of monitoring and treatment, for every common condition.
Early signs are promising. As mentioned earlier, this technology already exists for diabetics. Microsoft have now developed a prototype wearable device for Parkinson’s sufferers. This watch monitors wrist tremors and activates vibrating motors to steady arm movement. This allows the wearer to write legibly.
The rate of development in sensor technology will determine these projects' success.
Sense and sensor ability
Every wearable device on the market will sink or swim on the strength of its sensors. The IoMT is only possible thanks to exponential leaps in sensor technology over recent decades.
Broadly speaking, we can classify three types of sensor technology.
- Skin based: A new generation of textiles can measure biopotentials like heart rate and temperature through the epidermis. Similar sensors can even be incorporated into tattoos to measure blood pressure.
- Biofluids based: Other sensors measure body secretions like sweat, saliva, tears and urine. These can take the form of polymer-based wristbands or even underwear.
- Drug delivery: Devices such as smart bandages deliver medicine transdermally. For instance, a ring-shaped ocular insert secretes prostaglandin at regular intervals into glaucoma patients' eyes. This removes a common barrier to treatment courses: patients forgetting to take their eye drops.
The most successful wearable products will exhibit seriously impressive capacities for monitoring many different biological parameters. But they’ll also have to be comfortable. Even the highest-spec devices will flounder if the user experience is too invasive.
Privacy concerns will also determine these products’ fates. In 2018, thousands of units of Medtronic’s MiniMed insulin pump were recalled by the FDA due to potential cyber-security flaws. Devices that leave patients’ insulin flows at the mercy of hackers will struggle to gain a foothold.
The winners in the relentlessly innovative healthtech sector will be those who can bridge all these priorities.