Can smartphones and smartwatches help detect early signs of neurological and psychiatric disorders? Researchers at the University of Geneva (UNIGE) monitored a group of participants wearing connected devices and used artificial intelligence to analyze data such as heart rate, physical activity, sleep, and air pollution. Their findings show that connected devices can accurately predict emotional and cognitive fluctuations, opening new avenues for early detection of changes in brain health. This research npj digital medicine.
Brain health, including both cognitive and emotional function, is one of the major public health challenges of the 21st century. According to the World Health Organization (WHO), more than one in three people worldwide will have a neurological disorder such as stroke, epilepsy, or Parkinson’s disease, and more than one in two people will experience a mental disorder such as depression, anxiety, or schizophrenia at some point in their lives. This number continues to increase as the population ages.
Even in healthy adults, brain health fluctuates over time and reflects interactions between multiple factors such as environmental influences and individual lifestyle habits. Therefore, it is essential to analyze daily or weekly changes in cognitive and emotional functions to enable proactive and individualized prevention strategies.
A team at the University of Geneva (UNIGE) set out to determine whether wearable and mobile technology could be used to continuously and non-invasively monitor brain health. To this end, 88 volunteers aged 45 to 77 were fitted with a specialized smartphone app and smartwatch. Over a period of 10 months, these devices collected “passive” data such as heart rate, physical activity, sleep patterns, weather conditions, and air pollution levels without any intervention or change in participants’ daily habits. A total of 21 indicators were analyzed.
Every three months, participants also provided “active” data by completing questionnaires about their emotional state and taking cognitive ability tests.
Data analyzed by AI
Once data collection was complete, passive data was analyzed using artificial intelligence developed as part of the project.
The aim was to determine whether AI could predict fluctuations in participants’ cognitive and emotional well-being based on these data. ”
Igor Matias, doctoral assistant at the Institute of Statistics and Information Sciences, UNIGE Geneva School of Economics and Management (GSEM) and lead author of the study
We then compared the AI-based predictions to survey and test results. “On average, the error rate was only 12.5%, opening new possibilities for using connected devices for early detection of abnormalities and changes in brain health,” the researchers added.
Emotional states are the easiest to predict
Emotional states were most accurately predicted by artificial intelligence, with error rates generally ranging from 5% to 10%. In contrast, predictions of cognitive state were less accurate, with error rates ranging from 10% to 20%. In other words, AI is better at predicting responses to emotional surveys than cognitive tests.
Regarding the relevance of passive indicators, air pollution, weather conditions, daily heart rate, and sleep variability emerged as the most informative factors for cognition. Regarding emotional state, the most influential factors were mainly weather, sleep variability, and heart rate during sleep.
The research was carried out under the supervision of Professor Katarzyna Wac from GSEM’s Institute of Statistics and Information Sciences and Professor Matthias Kliegel from the Cognitive Aging Laboratory at the Faculty of Psychoeducational Sciences, and is part of the faculty joint project Providemus alz. The next stage has already begun. It aims to collect the same types of data over a 24-month period, examining the individual characteristics of participants associated with the best- and worst-performing AI models in order to better understand their applicability to individual real-world scenarios.
sauce:
Reference magazines:
Matthias, I others. (2026). Digital biomarkers for brain health: Passive and continuous assessment from wearable sensors. npj digital medicine. DOI: 10.1038/s41746-026-02340-y. https://www.nature.com/articles/s41746-026-02340-y
