OR WAIT null SECS
The technology’s promise extends to telehealth, interoperability, cybersecurity, and more.
When healthcare leaders discuss edge computing, the conversation tends to focus on its potential to improve the Internet of Things (IoT). Medical devices, they often say, can benefit from computing that takes place on the “edge of the network,” not in the cloud, saving time and managing large amounts of data. Although that is true, the promise of edge computing for healthcare is far greater, touching everyday care, telehealth, interoperability, and more.
Weisong Shi, PhD, a faculty fellow and computer science professor at Wayne State University in Michigan, researches edge computing and seeks to help Detroit hospitals use the technology. He recently spoke with Healthcare Analytics News™ about edge computing and its implications for medicine, laying out several paths that hospitals and healthcare decision makers can follow to leverage this advancement. He also described what edge computing accomplishes, how it does so, and why it is captivating health-tech leaders.
Cloud computing’s meteoric rise has resulted in many tech services and assets living in the digital realm, on cloud platforms owned by companies like Amazon, Microsoft, IBM, and Google, Shi says. Consequently, around 2015, the demand for edge computing began to increase. The popularity of the IoT, both in the workplace and at home, and its big data component have also contributed to the desire for more edge computing.
“The increase of this data is way faster than the increase of bandwidth and cloud storage,” he says. “That means it is impossible for them to be turning to the cloud all the time.”
Although edge computing is only taking off now, its techniques have been around for quite some time. Shi points to the early 2000s and even before, when web caching caught on as a means to accelerate content delivery. That technology took certain parts of a webpage, like its logo and icons, and stored them in a temporary internet folder on a user’s computer, which spurred quicker load times, he notes.
Edge computing is similar in that it entails stockpiling data on the edge of a network, with the goal of reducing how long it takes for users to access information. The technology may also allow users to get data during an outage, like when their internet connections go down, Shi says. Further, edge computing can bolster cybersecurity by limiting the flow of sensitive data to the cloud and maximize efficiency by sending only the information someone wants, when they want it, he adds.
The short answer: in many ways.
Imagine a not-so-distant future in which everyone wears an EKG sensor, Shi says, tracking heart rate, blood pressure, and even things like physical and social activity. Such a device would generate a tremendous amount of data—too much to leverage in any meaningful way if it were all destined for the cloud, he says. The personal nature of that data might also make users wary of sending it elsewhere. Edge computing, he says, could process that data on periphery of a home network, blasting only relevant results to the cloud.
“So, the doctor may say, ‘I need to see the EKG for the last hour,’ and you can then ask the edge to send only that data,” Shi says. Users may one day accomplish this task through voice assistants like Amazon’s Echo.
This sort of application may also advance chronic disease management, cutting down rehospitalization rates by enabling healthcare providers and other stakeholders to keep daily tabs on a patient, he says. What’s more, edge computing could help make available relevant data to resolve disputes between hospitals, insurers, and other parties when rehospitalization does occur, he says.
Telehealth companies and patients may also benefit from the technology. Edge computing stands to boost the ease with which digital video can be used to facilitate primary care appointments, Shi says. For developing countries, where healthcare is sometimes scarce, edge computing can be used on the network of an outfitted medical truck that visits isolated villages to connect residents to telemedicine services, he notes.
“But without a good edge computing platform, that would be very difficult,” Shi says. “Edge computing makes this a reality.”
Perhaps the most significant advance could be seen in emergency medical services (EMS). Edge computing can better enable providers to transmit crucial data from the ambulance to the hospital, saving time and arming emergency department teams with the knowledge they need to save lives. This arrangement could be particularly useful during large-scale disasters, Shi says.
“Today’s EMS is basically just transportation to home or the hospital,” he says. “Edge computing can be revolutionary for EMS.”
Looking at the EMS example, you might think that it’s easy: Just install a computer in the ambulance and send information via a typical network setup. But Shi ran experiments last year that suggested connectivity in cities is not always conducive to such a system. Despite using 4G LTE, they encountered connection problems when driving fast or in poor weather (2 common experiences for ambulances), Shi says. Edge computing could clear some of the obstacles that prevent this common-sense arrangement and various telemedicine initiatives from becoming commonplace.
The technology offers a similar solution for slow or down networks used by IoT devices, Shi says. As many Echo owners know, sometimes the cloud or WiFi is unavailable, which means their content is also out of reach. Edge computing keeps certain data accessible, even in cases like this, which could prove especially pivotal in healthcare, whose data are more consequential than, say, a party playlist stored on Amazon Music.
Depending on the user, edge computing may also help improve interoperability, Shi says. Its baked-in efficiencies could connect sharing parties more quickly and conveniently, shaving time and fighting headaches, he says.
As Shi makes clear, edge computing offers many solutions for healthcare and beyond. The technology also promises to power smart cities and their large camera-generated data sets, consumer electronics, and financial firms, to name a few.
Of course, edge computing does have its cheerleaders, and Shi says some appear a bit too ambitious. One speaker at a recent conference, for example, claimed that the goal of edge computing is to make all cloud services accessible within 30 milliseconds. “That might be too much,” Shi notes. “But the only way you can even approach that is through edge computing.”