Growing up, like many of us, I loved Star Trek.
The Original Series (re-runs), The Next Generation, Deep Space Nine, Voyager… all before DVRs were prevalent, so you can appreciate the dedication there. I liked the characters and the stories, but mostly I liked the promise of that version of the future. It was a vision of a time where people and technology seemed to work in harmony with one another. In many ways that vision has shaped much of the technology that we enjoy today. Everything from our phones, the hypospray, optical tweezers (tractor beam), the universal translator, to some of the more far fetched things like transparent aluminum and (conceptually) warp drive, can be seen on pretty much any episode of Star Trek.
There is one technology that was so common in the Star Trek universe that few people gave a second thought as they watched the show. Everybody talks to the computer. Even better, the computer responds in a meaningful manner. I’m sure this was just a Hollywood trick to skirt paying for CGI, but the concept ingrained itself in my mind. The characters would have random questions about a plot point that they would voice aloud, and the computer would helpfully provide a solution (most of the time). Nowhere was this more poignant than on Voyager when Harry Kim burst into sickbay and exclaimed “Computer, initiate Emergency Medical Hologram”. “Please state the nature of the medical emergency”.
With the advent of virtual digital assistants like Siri, Cortana, Google Now, Alexa and others, we can finally talk to our computers just like our favorite Star Trek characters did. These assistants can already tell us what the weather will be, read us the news, tell us when our package is going to arrive and how bad our commute is going to be. When we combine that capability with an increasing focus on health related applications and data collection, our own Emergency Medical Hologram (EMH) may not be the stuff of science fiction for very much longer.
Amazon’s Echo may be, at least for the moment, the most interesting of them all. While speaking to my phone or my computer is compelling, being able to call out a question while I’m cooking, getting ready for work, or working out is by far a more compelling use case. Amazon’s Echo device enables this in a way that no other device currently does. By providing an easy way for developers to extend the capabilities of Echo, Amazon opened the door to a plethora of opportunities. Some of them are mundane, like trivia applications, while others, like Boston Children’s Hospital KidsMD app have the potential to completely change the landscape.
What role should Virtual Digital Assistants play in healthcare?
Examining all of these devices/services from a Health and Wellness perspective, the possibilities are tremendous. Imagine an operating room where the surgeon has already scrubbed in and they need to refresh their memory of a particular detail of the patient's most recent blood test. They could simply voice their request and Alexa could reply with the requested information. That same kind of interaction could be applied to an Emergency Room, ICU, labs, etc. Information that was once buried on a server, or in a warren of floor to ceiling file drawers, would be there for the asking. These digital assistants could learn how to quickly pull up a patient's recent lab results, charts, medical history, etc. Additionally, it could provide a quick means of adding notes to a patient's chart. These interactions could enable medical staff to react much more quickly, and have more time to devote to their patients.
From the patient's perspective, there are an equal number of applications for this technology. While the patient is still in the hospital, they could use a voice activated digital assistant to control the room, bed, tv, etc using existing home automation technologies. They could also call the nurse without having to lift a finger. After they have been discharged, they could continue to use these services to interact with their physicians and nurses from home.
At Medullan, we are working with CareCentrix to improve home care coordination between doctors, nurses, and patients. As part of that effort, we are involving the Amazon Echo as a communication touchpoint. Simple information like daily food intake or blood pressure readings can be sent to a physician so that they can help monitor the patient's after-care. The doctor could also set up reminders for the patient to take their medication, and the patient could inquire about any interactions with their medications, or how to take it.
As the underlying AI’s continue to evolve, these devices could even begin to be used to assist with diagnosis. In April of this year, IBM’s Watson was able to correct a mis-diagnosis of a 60-year-old woman’s rare form of leukemia. In roughly 10 minutes, it compared the woman’s records against a massive database of 20 million cancer research papers and records and was able to provide an accurate diagnosis. It will be a while before this capability is common enough that it can be leveraged for the larger health-care industry, but combined with the existing digital assistant capabilities we have today, we find ourselves one step closer to a viable Emergency Medical Hologram. This kind of a tool, and it is a tool, would become an invaluable part of a physician’s daily life. It would not, and should not, take the place of a hand’s on diagnosis, but it could easily be utilized to narrow the possibilities and speed up the process.
So, what’s the downside here?
The biggest hurdle that will need to be overcome for these devices to become ubiquitous in hospitals, doctor's offices and other care facilities is privacy. To enable some of the interactions described above, these devices would require access to an unprecedented database of patient and medical information. In order for the AI’s and virtual digital assistants to be effective, they would need the ability to access and cross reference the records of any patient. While it is easy enough to safeguard how this information is used by the machine itself, restricting who has access to that information is difficult. When a device like Amazon’s Echo is sitting in a room, it will answer the question regardless of who is asking. There would be nothing to stop a patient from asking for information that should be available only to a doctor. Conversely, safeguards would need to be put into place so that a doctor could only ask about patients that are currently in their care. CITI and others have tried to overcome this by using a pin code. However, that only works if there is nobody else in the room to overhear. A better solution would be to create a voiceprint to identify the speaker, but traditionally, those kinds of systems are reasonably easy to circumvent.
Looking to Star Trek for inspiration again, these devices could network with sensors in a room to positively identify exactly who is in the room and gauge their responses accordingly, but that is undoubtedly still a ways off. It is difficult to prescribe a solution for this problem that meets the current HIPAA guidelines, but it is going to be critical to get it right.
Another aspect that needs to be considered as the capabilities of the virtual digital assistants extend into the healthcare space is whether they need to be classified as medical devices. If that were to occur, then every unique application or utilization would likely need to undergo approval by the FDA. A process which can be both costly and time-consuming. Large corporations can bear this burden while they wait for the approval to come in, but it would be a prohibitive barrier to entry for smaller companies trying to innovate.
Finally, policies would need to be put into place to help ensure that these devices and services don’t become a crutch. While the depth of capability that they represent can certainly be beneficial, it should not be seen as a replacement for the knowledge and experience of the medical staff. Alexa, Cortana, Siri, Google Now, Watson, Voyager’s EMH… they are all tools. Very capable tools, but tools none-the-less.
The future will be talking back...
All of these issues present opportunities for companies of all sizes, from startups to the giants that are already playing in this space. According to a survey conducted by RockHealth, there was over $4.3 Billion invested by venture capital firms toward medical applications in 2015. If a portion of those funds was devoted toward addressing these and other issues, then it is very possible that we could see these kinds of services achieve their potential while ensuring that no individual's personal information is compromised.
Perhaps the Emergency Medical Hologram said it best:
“In the beginning, there is darkness, the emptiness of a matrix waiting for the light. Then, a single photon flares into existence, then another. I awaken into this world fully programed, yet completely innocent, unaware of the hardships I'll endure, or the great potential I will one day fulfill.”
To read more about the impact that digital applications can have on care, read a blog post from our VP of Platform Solutions, Ryan Rossier, entitled: "Why the Digital Therapeutics market is poised for epic takeoff".