Predict Early Cognitive Changes in a Healthy Elderly Population by Passively Recording Sensor Data From Mobile Devices

The study will recruit 100 participants aged 65+ residing in assisted living facilities. Eligible participants will undergo an initial cognitive assessment to document their baseline cognitive status using cognitive tests conducted either in their assisted living facility or at Soroka Medical Center. An app designed for research data collection (Aware-Light) will be installed on participants' mobile devices, collecting data only after participants provide consent. Over three months, the app will automatically gather sensor data, including movement data from the accelerometer, orientation data from the gravity sensor, rotation data from the gyroscope, proximity data for measuring distance from nearby objects, ambient light levels from the light sensor, and typing dynamics such as typing frequency and duration without recording actual text. The app will also track app usage, including activation, pauses, closures, background activity, notifications, and crashes, as well as air pressure through the barometer, phone capabilities through the telephony detector, and screen states such as on, locked, or off. It will log communication events such as calls and messages without recording personal details, monitor vehicle speed through the linear accelerometer, track rotation angles through the rotation detector, and identify significant movements like walking, cycling, or traveling. No typed data, such as passwords, will be collected. After one month, the research team will contact participants to ensure the app functions correctly. After three months, participants will undergo a second cognitive assessment, and the app will be uninstalled.

Sex

All

Minimum age

65 Years

Inclusion Criteria:

* Age 65 and over
* Clalit HMO members.
* Without a known diagnosis of dementia
* Uses Android mobile phones and writes text messages

Exclusion Criteria:

* A known diagnosis of dementia
* Uses IOS mobile.
* Doesn't write text messages