Premium
Monitoring physical parameters from wearable sensors for detection of cognitive decline in routine care
Author(s) -
Goerss Doreen,
Amaefule Chimezie O.,
Kowe Antonia,
Köhler Stefanie,
Teipel Stefan J.
Publication year - 2021
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1002/alz.049767
Subject(s) - wearable computer , dementia , computer science , wearable technology , cognition , human–computer interaction , artificial intelligence , psychology , medicine , disease , embedded system , pathology , neuroscience
Background Wearable sensors offer the chance to measure activity and vital signs continuously and unobtrusively in elderly people living in domesticity or stationary care. To derive meaningful features from raw data and transform sensor data into clinically relevant information it is necessary to develop efficient algorithms and validate the results, e.g. with synchronized observational data. Only then sensor data bear the potential to detect functional changes or specific behaviors that are linked to Alzheimer’s disease and allow the complementation of the diagnostic process with sensor data. Method International data point to a potential role of serious games applications to detect early signs of cognitive decline. Along this line a study of the DZNE uses a comprehensive memory paradigm examining age sensitive decline in a citizen science study performed with digital platforms. Complementary approaches include non‐intrusive assessment of everyday behaviour. We performed several studies to examine the feasibility and potential of mobile sensors in detection of challenging behaviors, spatial disorientation and gait changes to classify people with MCI or dementia and detect situations with need for help. Result In a study with dementia‐patients that were wearing mobile sensors continuously over several weeks, we could show that multimodal sensors can be applied long‐term in the setting of PwD living in stationary care. Accelerometry can be used to detect behavioral patterns in dementia patients. We extended the set of sensor data with vital signs (e.g. heart rate, electrodermal activity) and indoor tracking to detect spatial disorientation in an ongoing wayfinding study to classify MCI /dementia based on sensor data. Conclusion Mobile sensors will be omnipresent in the following years. With complex algorithms we were able to derive features that allow information about crucial personal conditions with respect to mental health. Digital markers including sensor data won’t be a stand‐alone diagnostic tool and need to be embedded in an innovative diagnostic pathway.