The Role of Edge Computing and 5G in Healthcare
The role of technology in healthcare has changed the way we approach diagnosis, patient care and treatment. Patients now have access to the best diagnostic tools, advanced treatment procedures and minimally invasive surgeries resulting in less pain and quicker healing.
Edge computing and 5G promise to accelerate innovation and make healthcare delivery faster, cheaper and better. With the pandemic overwhelming an already overburdened healthcare system, it couldn’t have come at a better time. Many people do not want to go to a hospital or diagnostic center for fear of contracting COVID-19. This delays their chances of an early diagnosis and subsequent recovery. In addition, the steady rise in the elderly population is placing increased stress on elder care facilities around the world.
The number of people needing quick, efficient care far exceeds the number of providers available to care for them. Data from the World Health Organization shows that we are facing a global shortage of 7 million health workers. That’s the bad news. The good news is that the rapid advancement of technologies like edge computing and 5G is making it easier to introduce solutions that can make up for the shortfall in manpower and transform the way healthcare is being delivered.
Let’s look at some healthcare use cases of edge computing and 5G in action:
Wearables
Wearables can continuously monitor blood pressure, heart rate, body temperature, blood oxygen levels and more.
This data is pushed to the nearest edge server and is processed locally at that edge location, minimizing latency and increasing processing speeds. Doctors can use this information to evaluate the health of their patients in real time.
In Hospitals
Radar-enabled bed sensors monitor vital signs such as heart rate, respiration rate and blood pressure — all while alerting caregivers when normal limits are breached.
Sensors in beds track when and how long people sleep. This data acquired from sleep patterns can detect early signs of illness.
Until recently, hospitals have had a centralized architecture with their data stored in the cloud. Various smaller clinics and health centers connect to a central location to store and process data.
Edge computing provides hospitals with the benefit of storing data at the nearest edge location and making it available for processing quickly. This has an added security advantage since data is stored locally and not transmitted over long distances to the cloud, which reduces the risk of someone hacking the data midway.
Point-of-Care Diagnostics and Telehealth
On-demand healthcare in the form of mobile point-of-care diagnostics brings healthcare to people in both urban and rural areas.
Together with vital signs, data specific to critical illnesses such as diabetes and cardiovascular diseases is pushed to the closest edge server. The data can be processed, analyzed and transmitted to doctors at remote locations within a matter of minutes.
This ready availability of health data has led to the growth in telehealth apps resulting in demands for increased capacity at service providers’ sites.
Edge computing helps developers spin up additional compute and storage capacity quickly, meeting urgent demands and streamlining resources.
Ambulances
Ambulances can now do much more than just ferrying patients to the hospital and back.
Technology embedded inside point-of-care screening devices and high-definition video can transmit data from vital signs and other health parameters over a 5G connection, back-hauled over a last-mile edge provider’s network, back to a central monitoring station.
Paramedics and emergency medical responders can then collaborate with doctors to stabilize the patient before they are rushed to the hospital, while emergency room personnel can get ready for the patient to arrive.
AI
No discussion about the future of healthcare can be complete without artificial intelligence. From chronic diseases and cancer to radiology and risk assessments, we can use AI to transform patient care and diagnosis.
Here are two examples of how AI is transforming early detection and diagnosis.
Melanoma Detection
Melanoma is a malignant tumor responsible for over 70% of skin cancer-related deaths. Physicians usually rely on visual inspection to identify suspected skin lesions. Although in many cases it’s difficult to make an accurate diagnosis.
AI helps to solve that problem. A software system using DCNNs (deep convolutional neural networks) can analyze a wide-view image acquired with a smartphone camera and identify the lesions that need further investigation.
By storing images at the nearest edge server and processing them locally, results are returned within minutes.
Radiology
Edge computing has given rise to AI applications that reduce the time it takes for an MRI scan. New research by Facebook and the NYU Grossman School of Medicine shows that rapid MRI images generated with AI contain diagnostic information comparable to images taken by a slower, conventional MRI scanner.
By removing roughly three-fourths of the raw data used to create a scan, the AI model is able to generate a fastMRI scan comparable to that created by the normal MRI process.
Because the fastMRI scans require four times less data, they have the potential to scan patients faster so they can spend less time in the MRI machine.
Edge and 5G Go Hand-in-Hand
It would be wrong to think about either 5G or edge computing in isolation. Edge computing is the only way 5G will meet its latency targets of less that 5 milliseconds. While most people think about 5G in terms of lower latency, they forget about the amount of data that edge devices produce.
Devices such as wearables, sensors and other IoT equipment produce large amounts of data that need to be managed and processed locally, with the results transported back to doctors, hospital emergency rooms and remote facilities in near real-time.
However, with the suboptimal routing infrastructure at many telco locations, this is not always possible. 5G with edge promises to solve that problem by significantly reducing latency between the endpoint and the data center.