This AI engine predicts your lifespan through activity tracking apps
Researchers from the Moscow Institute of Physics and Technology and biotech company GERO have developed an AI algorithm that uses data from activity tracking and fitness apps in smartphones and wearables to estimate your lifespan. In a breakthrough demonstration, the researchers showcased how the combination of wearable sensors and AI technologies can be used for continuous health risk monitoring.
Neural network was trained to find "biologically relevant" motion patterns
The researchers analyzed physical activity and clinical records of several participants through a large-scale 2003-2006 US National Health and Nutrition Examination Survey. They trained the neural network to find "biologically relevant" motion patterns, i.e., step counts, sleeping patterns, gym visits, and frequency of switching between active and inactive periods. This way the AI learned to predict biological age and mortality risks of participants.
Gero Lifespan app analyzes Fitbit data to estimate your lifespan
The researchers have already developed a free beta-version of an iOS app called Gero Lifespan that estimates users' lifespan using the built-in smartphone accelerometer and data from apps like Apple Health and Fitbit. Such health risk monitoring and production of digital biomarkers of aging and frailty can provide real-time feedback to life and health insurance companies, and healthcare and wellness providers, researchers said.
The app can help you in pension planning, insurance mitigation
However, they acknowledged that the app isn't completely precise because it doesn't take into account diet, genetics, and other crucial factors. Asking doctors to use both the app and clinical analysis for a proper lifespan estimate, researchers said, "Combination of aging theory with modern machine learning tools will produce better health risks models to mitigate longevity risks in insurance and help in pension planning."
