Are Activity Trackers ready for Deep Learning?

Are Activity Trackers ready for Deep Learning?

At the “REWORK Deep Learning in Healthcare Summit” Johanna Ernst from Oxford University presented her research on activity trackers to collect patient’s vitals as a basis for deep learning monitoring apps. The patient’s vital data correlates with his heart frequency, giving information on the success of a surgery. That is why, vitals can be used for deep learning monitoring apps.

She sees the main challenges in the application of trackers by different brands in:

• Patent Expiration
• R&D Cost
• Creation of a One Size fits all Model

Further challenges are finding a marker / a parameter (as the basis for the development of a deep learning algorithm) and actually getting the raw data from the provider of the activity tracker. Also, markers are not always accurate enough.

Although an activity tracker is very useful for developing monitoring Apps with Deep Learning in creating patient compliance, the use of an activity tracker has a negative psychological effect on the patient: Patients tend to reward themselves, after looking at their successes tracked by their device, thus influencing their vitals negatively.

So what do you think? When will fitness trackers be fit for the application of AI monitoring apps?

Comments are closed.