The Future for Healthcare is Sensors and Artificial Intelligence
Every year, in the first part of January, the Consumer Electronics Show (CES) is held in Las Vegas. It has been going on for more than 57 years and this year attracted 138,000+ attendees. It is the showplace for unveiling the latest in consumer technology and what the future looks like for computer hardware, electronics, and software. This is the third year I have visited the show, and each year, I am absolutely astounded by what technology can do now and what it has the power to do when it comes to making us healthier and enjoy longer lives in the future.
As you might guess, I concentrate on the technology that deals with healthcare and there was a lot to see. There were things that kept us safe, continually monitored our well-being, and entertained us. Some products made it easier to age in place and made it possible to perform healthcare procedures at home that historically could only be done at the hospital. The new things that were showcased dealt with all facets of healthcare, but there were two themes that seemed to stand out as I toured the cavernous exhibit halls and listened to the keynotes and panels . . . the advancements in sensors and the prevalence of artificial intelligence (AI) in virtually every technology.
Sensors have been around for many years, but their size, their chameleon-like ability to blend into our lives, their versatility, and their accuracy have all taken huge steps forward. For instance, doctors have been using wearable EKG monitors to study a person’s heart health at home, usually for a couple of days. It required some strap-on devices and restricted some activities. At CES, there was a wearable sports bra-type clothing that gave continuous and accurate EKG readings. There is a ring that fits on your finger that will monitor heart rate, blood oxygen, and other vital signs 24/7, as well as sleep, activity, and stress. It will also track your movements through accelerometers that can alert others if you have fallen. These capabilities themselves aren’t that new but the comfort of clothing or a ring are. The sensors are less obtrusive and, therefore, more likely to be worn.
There was another point about the data that is collected by sensors that was discussed in one of the panels. It was pointed out that when we go to the doctor, our vitals are captured at that point in time. I think all of us have experienced higher blood pressure when it’s measured at the doctor’s office. Everything from bloodwork to EKGs is for a specific period of time. Wearable sensors can provide continuous data that can be averaged and/or linked to what activity you were involved with at the time, a much more accurate measurement of your body’s true status over time. It also allows the creation of an accurate baseline which, when compared to a future period of data, can identify important deviations that can alert us to developing problems. It is this type of technology that can truly have an impact on our lives.
It seemed like the words artificial intelligence were everywhere, on every type of technology. These two words seem to cause a great deal of joy to the attendees and a great deal of consternation to the rest of us. This is a new tool and can certainly be mystifying. Many are worried that it will take over the world. Our lawmakers are all grappling with what regulations will be required.
AI is difficult to understand; I try to understand how AI works in healthcare by thinking of a doctor who could:
Instantly access all of your health records – medicines you take, maladies you have, the health of your ancestors, your DNA, etc.
Update them with the vitals that were taken today.
Understand your symptoms and complaints that brought you into the office.
Take this information and compare it to all the other people in the world, filtering it down to people who fit you as closely as possible in terms of your health and the problems you are experiencing now.
Ascertain what further tests would give more relevant information.
Take the results of this added information and see what form of treatment has worked the best for others in the world who have had similar health histories and symptoms.
Recommend your treatment after considering any personal trends in your data and those things that are unique to you.
This is an overly simplified description of an “AI doctor,” but it shows how dependent AI is on accurate data, how it learns by observing what has worked and what hasn’t worked, how it makes the best use of the long-term collection of personal data, and how fast it can work. AI technologists are working to create databases that can support this type of learning and reasoning.
This type of “AI doctor” is in the future, but technologists are finding that the intelligence part of AI is tough to come by in healthcare. One of the panelists at CES said, “we are drowning in data but starving in insight”. Using AI to improve the insight and the diagnosis is great; accepting this diagnosis before an acceptable level of accuracy is obtained is not so great. It’s been said that AI is becoming the new doctor; we just need to make sure it takes the Hippocratic oath seriously. This is a great challenge for AI in healthcare. We are many steps away from AI being the last word of diagnosis and treatment, but using it in the near term to augment the tools available to our doctors will be a huge step forward.
One area where the increased use of personal data and AI could quickly benefit value-based care is health care. I have always championed the benefits of moving away from the expensive fee-for-service model to value-based care. The gathering of consistent ongoing health data on an individual will allow better tracking of outcomes and more accurately rewarding value-based care.
Think how much quicker we would have known about the weight loss benefits of the diabetes drugs if we had had the data collection mechanism to gather the information from the patients and the AI logic that would see the pattern of weight loss in the patients.
Great strides are being made in developing future technologies. There has been much talk about using sensors and AI in all facets of healthcare, some of it in a negative context. However, with the right controls and oversight, using them to benefit the patient and to be more efficient will serve us well in the future.
Best, Thair