The team used a convolutional neural network (CNN) to find a "biologically relevant" pattern in US health survey data and to find the relationship between the data and the health of a person and his age estimate.
This AI not only takes into account the number of steps taken but also how often someone actively moves after silence. This can indicate user habits such as sleeping habits or exercising.
After that, AI simply process the data that he got from the user's phone. Currently, you can already try it using the Gero Lifespan iPhone app. The app uses data from Apple Health, Fitbit and Rescuetime to estimate your age.
However, this AI can not estimate the overall health of a person, considering he is not getting information such as food consumed, genetics and several other important factors.
For the better, doctors will use this mobile application and clinical analysis to estimate the patient's age more accurately.