Wearable multimodal sensors for the detection of behavioral and psychological symptoms of dementia using personalized machine learning models

Abstract

Introduction Behavioral and psychological symptoms of dementia (BPSD) signal distress or unmet needs and present a risk to people with dementia and their caregivers. Variability in the expression of these symptoms is a barrier to the performance of digital biomarkers. The aim of this study was to use wearable multimodal sensors to develop personalized machine learning models capable of detecting individual patterns of BPSD. Methods Older adults with dementia and BPSD (n = 17) on a dementia care unit wore a wristband during waking hours for up to 8 weeks. The wristband captured motion (accelerometer) and physiological indicators (blood volume pulse, electrodermal activity, and skin temperature). Agitation or aggression events were tracked, and research staff reviewed videos to precisely annotate the sensor data. Personalized machine learning models were developed using 1‐minute intervals and classifying the presence of behavioral symptoms, and behavioral symptoms by type (motor agitation, verbal aggression, or physical aggression). Results Behavioral events were rare, representing 3.4% of the total data. Personalized models classified behavioral symptoms with a median area under the receiver operating curve (AUC) of 0.87 (range 0.64–0.95). The relative importance of the different sensor features to the predictive models varied both by individual and behavior type. Discussion Patterns of sensor data associated with BPSD are highly individualized, and future studies of the digital phenotyping of these behaviors would benefit from personalization.

Noninvasive monitoring technologies to identify discomfort and distressing symptoms in persons with limited communication at the end of life: a scoping review

“…Several systems combined ECG, respiration monitors, EDA, sEMG, skin temperature, and breath temperature to monitor the activation of the sympathetic nervous system, which was indicative of pain [ 119 ], mental stress [ 120 ], anxiety [ 121 ], agitation [ 122 ], or significant moments [ 123 ] in different studies. In most studies, the sensors were implemented separately, but several studies integrated the sensors into a chest belt [ 122 ], a wristband [ 124 ], or a system worn on the fingertip [ 123 ]. New algorithms for analyzing the data were proposed [ 125 ].…”

“…Compared with self-report, the systems had an acceptable accuracy in classifying pain into three severity categories in healthy adults in the laboratory [ 119 ], had good clinimetrics in detecting moderate to severe pain among patients in the ICU, but the results were not optimal when the pain was milder [ 128 ]. They could detect significant moments [ 123 ], agitation and aggression in people with dementia [ 124 , 129 ], and correlated with mental fatigue [ 130 ] and stress in healthy adults [ 120 , 126 ]. One system could indicate anxiety with high precision but low specificity compared to self-report and observation in healthy adults [ 121 ].…”

Noninvasive monitoring technologies to identify discomfort and distressing symptoms in persons with limited communication at the end of life: A scoping review

“…( 119) had good clinimetrics in detecting moderate to severe pain, but the results were not optimal when the pain was milder. (128) They could detect signi cant moments, (123) agitation and aggression in people with dementia, (124,129) and correlated with mental fatigue (130) and stress in healthy adults. (120, 126) One system could indicate anxiety with high precision but low speci city compared to self-report and observation.…”

Applications of artificial intelligence in dementia research

“…72 With advances in wearable technologies, plenty of data collected from wearable sensors can also be applied in machine learning models and thus improve the performance. 73 Computerized Cognitive Screening Dementia screening aims to identify those in the prodromal phase of dementia by using neuropsychological tests. 74 The tests include the Abbreviated Mental Test, the Montreal Cognitive Assessment (MoCA), the Mini-Mental State Examination (MMSE), and others.…”

“…80 The flexibility of wearable platforms has also provided a variety of data to detect cognitive status. 73…”