The global pandemic has driven healthcare practitioners and fellow staff to implement more effective, remote-friendly approaches to diagnosing and interacting with patients. Despite their best efforts, however, a Zoom or Facetime call can’t fully replace the diagnostic and consultation processes entirely. A real challenge is checking the vital signs of patients including their pulse.
Gratefully, there is a solution, and it’s thanks to technical innovation that it even came to be. Spearheaded by a dedicated team at the University of Washington and presented at the December 2020 Neural Information Processing Systems (NIPS) conference, it’s set to give your telehealth services a much-needed shot in the arm. Let’s find out how!
The Power – and Practicality – of Machine Learning
Automation is nothing new, but how it can be used today is leaps ahead of what was achievable even a few short years ago. The University of Washington team’s solution utilizes a state-of-the-art machine learning algorithm, which can use any ordinary smartphone or tablet camera to determine real-time vitals with greater accuracy. What’s more, the technology is foolproof in the sense that variations in lighting, skin tone, and camera quality aren’t going to easily interfere with the readings taken. With continued measurements, the algorithm better understands a patient’s respiration rate, blood flow in the facial region, pulse, and more – and so does the practitioner on the other end.
A Personalized Experience
In its initial phase, the machine learning capabilities engineered by the team were limited in scope, as it didn’t function correctly in some filming conditions. By custom-tailoring the software so it develops a deeper understanding of each individual, combined with performance optimizations in the back end, it became more effective and accurate over time.
Testing and Refinement
Improvements are continuing to be made especially when analyzing vital signs of those with darker complexions, as the system still struggles in this area from time to time despite being a drastic improvement over the previous iteration. However, with the dedication of the team and the technology in the spotlight of telehealth innovation, it’s only a matter of time before these issues are ironed out. With the machine learning software largely effective and operational on almost any modern device with a decent selfie camera, healthcare facilities including clinics are working closely with the research team to test it out and determine whether it’s a viable solution in real-world telehealth scenarios.
We at Qualicare Waterloo can’t wait to see how this plays out! If the promising results so far are any indication, it may give remote diagnostic processes a boost to improve efficiency and accuracy. In the meantime, for more details on how we can assist you remotely today, contact our caregiving team.
Comments are closed.