How ‘Dr.Alexa’ will Reshape Healthcare
At many levels, it is almost as if Voice is the next logical step in making our lives easier. Think about it — lesser screen time, smoother communication and easier access to information.
Every tech giant, be it Amazon, Google, Microsoft or Alibaba, Tencent and Xiaomi, is engaged in a competitive war to win the highest share of users.
Talking is the new typing, and why not? Talking comes to us naturally, reduces the friction of a human to machine interaction and encourages an almost ‘communal’ form of correspondence.
Cloud Computing and AI driven NLP algorithms parse common language, convert them to text, derive meaning and take appropriate action. Many of which are astonishingly accurate. Baidu, for instance, claimed a ‘language understanding’ accuracy of nearly 96%, according to a 2016 McKinsey report. Shortly after, in 2017, Google claimed that not only was speech recognition for its product at 95%, it could even return meaningful results.
Smart speakers are embedded across most of our everyday devices — smartphones, cars, homes and even work places. We use them to search for information, ask questions, listen to music and even look for weather information.
Time is ripe to see the impact of more sophisticated applications — What would the repercussions be if we combined the natural usage pattern of smart speakers with an industry need?
In terms of the healthcare sector, it’s almost as if common usage has bled into the industry already because the use cases are plenty.
As everyday users of this technology ourselves, many results are, for us, almost expected. For instance, with respect to “health-related queries via voice assistance”, a recent study revealed the below categories of search against % users of smart speakers in the US:
These trends although exciting, are merely replacing web-search with voice-search. The real innovation lies in being able to use voice-based smart-interaction to smoothen systems and processes within complex set-ups.
The IDHA (Innovation and Digital Health Accelerator) at Boston Children’s Hospital piloted Voice Assistants in the hospital and stands out as a particularly terrific example.
The tests began sometime in 2016. Here,
1. lab results were spoken to technicians in mock outpatient settings,
2. nurses were given instructions for collecting specific specimen in the ICU,
3. physicians in the OR voice documented endoscopy images in real-time
4. parents learnt through voice instructions to clean central intravenous catheter.
On the basis of the reactions, success and outcomes, the pilots were deployed for usage in real-world scenarios. Two of the most interesting ones were:
1. The ICU: Where largely the focus was to increase sterility and maintain an infection-free environment while speeding up processes. Using voice tech to ask for administrative information like “Which nurse is in-charge at 3 this afternoon?” or “How many wards are free on the 5th Floor?” allowed to save critical time which would usually be spent in manual work to look for information. It also boosted sterility by virtue of being “hands-free”.
2. Organ Transplant: Piloting a voice-guided checklist process to streamline the pre-operative procedure helped speed up the process, allow for hands-free navigation and help reduce human error in a usually busy department.
In such situations, while the accuracy of understanding the voice plays an important role, a more important feature is the content of the search results itself. A lot of time needs to be invested in building, organizing and vetting the final content on which the voice feature can be enabled. The feature truly is only as good as its content. Further there is a need to protect patient privacy and this software needs to be compliant with local patient privacy policies.
Many times, voice alone cannot be a strong enough guide in such fields — visuals become imperative. Hence for practical usage the ability to switch between the two and the hardware to support that is also a point of consideration.
The overall application of the technology has definitely improved, however in order for us to truly apply and use it in a disruptive way there’s a long way to go. While voice search can be an ally in decision making in a lot of cases, sensitive fields like healthcare require it to be more dependable, secure, trustworthy and as far as possible error-free.
One doctor even raised a concern about trust — In a doctor-patient interaction, what if the smart speaker in the room throws up an answer different from the doctor? Would it create friction and possible trust issues between the patient and doctor? Perhaps then it is important that smart speakers are not a direct part of such interactions, and only help in the background.
Supplementing this however, are some use-cases where Alexa skills have been built to make patient’s experience smoother from a logistics perspective. For e.g.: New York’s Northwell Health lets patients use Alexa to find the shortest wait-time for emergency rooms within a particular city or zip-code. Boston’s Commonwealth Care Alliance allows patients to set up appointments and reminders using voice. While, Boston’s Beth Israel Deaconess Medical Center allowed patients to enquire about and then order meals as per dietary requirements.
In the upcoming future, a lot of work is being done on complex voice analysis. Can we hope to diagnose depression by analyzing the cadence, tonality and speed of speech? Perhaps voice biometrics could help solve for current security concerns? Could we get treated right at the comfort of our homes through digital voice and visual communication?
The answers to these questions will unfold sooner than we think. Indeed, the development has grown leaps and bounds and while challenges remain, these will soon be overcome as we move towards smoother digital communication for faster, more accurate healthcare.
This story is published in The Startup, Medium’s largest entrepreneurship publication followed by +406,714 people.
Subscribe to receive our top stories here.
