The Future of Connected Health

At the same time that the healthcare space is shifting its focus from billing for procedures to billing for patient outcome, doctors are required to see more patients per hour and spend more time entering data into computers than interacting with patients. Doctors no longer have the luxury of taking the time to get to know their patients better. Doctors need data points on patients, especially in settings outside of hospitals—including information about patients’ diets, how much they exercise, and if they are taking their medications properly—to help them successfully understand and diagnose illnesses. An example of an early foray into gathering those data points via telemedicine is the nurses in the Department of Veterans Affairs, who call veterans at home once a week and ask them a list of health-related questions. With the development of wireless networks, telemedicine expanded to video conferencing, remote patient monitoring, the sharing of patient information via mobile devices, or a telemetry technician monitoring the data sent from sensors on a hospitalized patient over the internal network.

Recently, the growth of the Internet of Things (IoT) and the decreasing cost of central data storage and analytics have begun to enable more effective patient monitoring, data collection, and improvements in care. Of the various classes of remote monitoring IoT devices, wearables have proved particularly effective in patient engagement and care cost reduction. According to Gartner, wearable devices will be crucial in driving the next phase of connected health. The direction and velocity of the wearable space will depend on factors such as changes in government regulation, infrastructure, reimbursements by insurance companies, and advancements in the hardware and software designs of wearables.

Wearables will be driving the future of connected health. (Image courtesy of Pixabay.)

Recent regulatory decisions by the U.S. federal government are likely to have a lasting impact on the development of the connected health space. The U.S. Food and Drug Administration (FDA) has deregulated consumer wellness devices by waiving approval for low-risk devices that do not make specific medical claims and by refraining from establishing standard security practices. As a result, the number of low-risk devices has exploded, despite strong caution from safety advocates. The U.S. Federal Communications Commission (FCC) has formulated strategies to promote 5G, such as making more spectrum available to the marketplace, updating regulations, and limiting the ability of local city governments to impede the installation of 5G infrastructure.

Although wearables provide apparent benefits,network data traffic capacity can be a limiting factor for remote patient monitoring. Slow speeds and unreliable connections will prevent doctors from getting the patient data to make real-time decisions, which can be detrimental in the case of emergencies.

The fifth generation wireless system (5G), which is much faster than 4G, will reduce the time needed to transfer large files of images or video from tens of minutes to tens of seconds, so that doctors can make real-time diagnoses as well as monitor more patients simultaneously. In addition, by reducing network congestion, a massive number of IoT devices can stay on the network, process the data they have collected using artificial intelligence, and make real-time recommendations to their users.

Moreover, 5G will enable remotely controlled surgeries, where doctors make precise and immediate decisions based on the data that was just collected and transmitted. Lastly, the 5G network’s low latency will allow patients and doctors who speak different languages to interact as translators interpret for them in real time. The future is already here. Using a 5G network powered by China Mobile and Huawei and in an operation that lasted three hours, a surgeon in China successfully implanted a deep brain stimulation device into a patient with Parkinson’s disease who was located nearly two thousand miles away and was unable to travel.

Fast data transfer by 5G is not enough, however; doctors also need help in making sense of vast amounts of data. Since most of the data collected by the IoT devices are repetitive and irrelevant, AI is used to identify and analyze relevant information to make recommendations. A common version of telemetry, such as what is used at the Scripps Hospital in San Diego, has a human technician working shifts to continuously monitor 10 to 20 computers, each of which tracks the monitor on a patient. AI can help the technician monitor more patients without delay, make faster and more accurate diagnoses, and push treatment recommendations to the technician. In addition, AI can help assess the cost of care and mine patient medical records for at-risk patients.

AI will also enhance the usability of wearables; its capability in face, voice, movement and gesture recognition will help wearables meet the market demand for no-touch user interfaces, which are a must-have feature due to the popularity of virtual personal assistants like Siri and Alexa. The integration of information from different health databases and datasets collected by IoT devices will be made possible by AI for meaningful analysis.Because running AI software requires a lot of processing power, AI is usually done in the central cloud. Data will be collected on the wearable device and sent to the cloud; after analysis and review by the doctor, the action items will be sent back to the device. However, transferring data between the IoT device and the cloud is expensive, not to mention time consuming (hence another big timelag). Therefore, there is a shift toward performing the data processing at the IoT device (edge computing) to reduce the cost and timelag.

Edge computing will place specific demands on IoT device hardware. IoT wearables need to function continuously for a long time, so they need processor chips that strike a good balance between processing power and energy efficiency. Also, because IoT devices will be deployed on a large scale for many different applications, sensors that are low-cost, small, energy efficient, noninvasive, sensitive and durable will make IoT devices more user friendly.

Lowcost and small. New manufacturing platforms such as Integra Devices’ Amalga can produce small sensors with lower design, development and production overhead. 

Energy efficient. Reducing a sensor’s continuous power consumption to ultra low levels or making the sensor dormant until it is activated by an event can lower overall energy use. In addition, the net battery energy consumption can be decreased by harvesting motion, pressure, light or body heat energy. 

Lower invasiveness. Using sweat instead of blood to measure metabolites, such as in the case of continuous glucose monitoring in patients with Type 1 diabetes, will increase the ease of use. Also, using microsensors to visualize the intestinal tract or track the healing of internal injuries will minimize the level of intrusiveness that patients experience.

Sensitivity. Flexible sheet-like sensors that adhere closely to a patient’s skin will improve the accuracy of data collection. Hair-like passive sensors require no power supply and can measure temperature, pressure and position without contact.

Durability. Because both battery replacement and maintenance can be cumbersome to patients, sensors that can self-calibrate or self-repair will be a cost-effective option.

Accuracy. In the future, sensors that integrate data from multiple channels will provide higher accuracy than single-channel sensors can.

Like consumer electronics, health-related wearables may also become multifunctional. For example, Starkey's Livio AI hearing aid can translate 27 languages in real time and double as a hands-free tracker of a user’s physical and mental health. iHEAR Medical is developing hearing aids that will not only help a patient hear but also listen and respond to social media and text messages. Another example of this innovation will be enlisting virtual personal assistants like Alexa and Siri to monitor patients, collect environmental data, and send reminders. While the outlook of wearables is promising in the connected health space, several challenges remain, such as cleaning diverse datasets for analysis and decision integration, keeping patient data safe and private, and responding to future government regulation on data management.