Multidimensional digital biomarker of cognitive health: unobtrusive and continuous monitoring of cognitive changes using smartphones DOI Open Access
Maciej Kos

Опубликована: Янв. 1, 2024

BACKGROUND. Cognitive functionality is a critical determinant of quality life. Acquired cognitive impairment associated with aging, neurocognitive disorders, like Alzheimer's disease, and traumatic brain injury, pose major challenges to healthcare systems throughout the world. Depending on etiology, may start in early middle adulthood, between 3.1% 21% global population suffering from depending age location. The number dementia cases alone expected reach 115 million by 2050. Yet, impairments are typically detected late decline. Positive diagnoses often so process that there not much can be done. Hence, detection intervention necessary intervene as well develop effective treatments. To facilitate successful methods needed for monitoring cognition detect signs mild dementia. Furthermore, development efficacious therapeutics necessitates subtle changes functions over time. However, effectiveness existing neuropsychological assessments diminished their sporadicity, difficulty accounting context-dependent nature individuals' health (e.g., having "good" or "bad" day), reliance frequently inaccurate individual caregiver reports. Thus, new approaches objective ecologically valid assessment impaired disorders. APPROACH. Current mHealth AI enable continuous context inference smartphone use location data. Studies involving younger older adults suggest characteristics an individual's mobile app typing speed correlate working memory, attention, psychomotor function. Individuals' reports Instrumental Activities Daily Living (IADLs) significantly enhanced continuous, inferences estimates engagement digital equivalents subset IADLs shopping banking, instant messaging) passively collected smartphone-based data, contributing earlier more accurate Therefore, address tests, I designed approach augment smartphone-derived changes, interactions smartphones require variety skills. used data inform multidimensional biomarker (henceforth, biomarker). proposed - developed using combination AI/ML mechanistic modeling would enrich clinical tests derived computationally unobtrusively application characteristics, motor aspects speed), both diagnosis disorders treatment personalization amelioration. assess feasibility developing biomarker, conducted repeated-measures pilot study consisting collection installed participants' acquisition during two lab visits, four months apart. participants were 22 middle-aged levels varying no subjective decline impairment. FINDINGS. This reveals behavioral provide insights into effectively selected conventional life-space assessments. Smartphone-estimated correlated highly self-reported measures, particularly specific Life-Space Assessment Measure Functional Mobility. indicate significant relationship patterns usage abilities. Specifically, lower ability linked unpredictable apps categories (high entropy), extended time, switching often. These correspond poorer performance range functions, including executive perceptual reasoning, visual learning, evidenced longer average daily total interaction times. Additionally, variability duration memory learning tasks. Participants inhibition capabilities showed higher entropy durations, suggesting high reflect capacity. Moreover, proportion time spent IADLs-related (shopping, food ordering, maps navigation, taxi/ride-share) abilities terms sustained functions. result suggests spending proportionally these conduct activities independently. CONCLUSIONS. underscores potential analysis providing near real-time, nuanced view functioning detecting changes. Overall, findings propose current characterizing self-reports. also demonstrates engaging comprehensive combines intensive testing collection.--Author's abstract

Язык: Английский

SmartSense-D: A safety, feasibility, and acceptability pilot study of digital phenotyping in young people with major depressive disorder DOI Creative Commons
Anthony Camargo, Scott D. Tagliaferri, Simon D’Alfonso

и другие.

Digital Health, Год журнала: 2025, Номер 11

Опубликована: Апрель 1, 2025

Background Digital assessment of behaviours, including physical activity, sleep, and social interactions could be associated with changes in mood other mental health symptoms. This study assessed the safety, feasibility, acceptability, potential predictive value passive active sensing young people major depressive disorder (MDD). Methods Over eight weeks, (smartphone sensing, actigraphy) (ecological momentary assessment; EMA) data were collected from 40 participants MDD (aged 16–25 years). We acceptability daily this population. Additionally, linear mixed models correlation analysis explored associations between measures. Results Of 48 participants, 83% (n = 40) completed full protocol. No adverse events reported. averaged 35.9 days (65.3%) EMAs 37.9 (69%) actigraphy data. Smartphone sensors recorded communication for 21.1 (38.4%), location 43.1 (78.4%), maximum unlock duration 43.4 (79%), media use 34.8 (63.3%), inter-key delay 32.8 (59.6%). Regarding 83.1% found application usable comfortable. Secondary measures showed significant correlations sleep phone sensors. There was a negative association positive ratings QIDS total scores (Beta coefficient [95% CI]: 2.66 [−3.98, −1.34]; p 0.002). Conclusion Passive methods safe, acceptable among MDD.

Язык: Английский

Процитировано

0

Multidimensional digital biomarker of cognitive health: unobtrusive and continuous monitoring of cognitive changes using smartphones DOI Open Access
Maciej Kos

Опубликована: Янв. 1, 2024

BACKGROUND. Cognitive functionality is a critical determinant of quality life. Acquired cognitive impairment associated with aging, neurocognitive disorders, like Alzheimer's disease, and traumatic brain injury, pose major challenges to healthcare systems throughout the world. Depending on etiology, may start in early middle adulthood, between 3.1% 21% global population suffering from depending age location. The number dementia cases alone expected reach 115 million by 2050. Yet, impairments are typically detected late decline. Positive diagnoses often so process that there not much can be done. Hence, detection intervention necessary intervene as well develop effective treatments. To facilitate successful methods needed for monitoring cognition detect signs mild dementia. Furthermore, development efficacious therapeutics necessitates subtle changes functions over time. However, effectiveness existing neuropsychological assessments diminished their sporadicity, difficulty accounting context-dependent nature individuals' health (e.g., having "good" or "bad" day), reliance frequently inaccurate individual caregiver reports. Thus, new approaches objective ecologically valid assessment impaired disorders. APPROACH. Current mHealth AI enable continuous context inference smartphone use location data. Studies involving younger older adults suggest characteristics an individual's mobile app typing speed correlate working memory, attention, psychomotor function. Individuals' reports Instrumental Activities Daily Living (IADLs) significantly enhanced continuous, inferences estimates engagement digital equivalents subset IADLs shopping banking, instant messaging) passively collected smartphone-based data, contributing earlier more accurate Therefore, address tests, I designed approach augment smartphone-derived changes, interactions smartphones require variety skills. used data inform multidimensional biomarker (henceforth, biomarker). proposed - developed using combination AI/ML mechanistic modeling would enrich clinical tests derived computationally unobtrusively application characteristics, motor aspects speed), both diagnosis disorders treatment personalization amelioration. assess feasibility developing biomarker, conducted repeated-measures pilot study consisting collection installed participants' acquisition during two lab visits, four months apart. participants were 22 middle-aged levels varying no subjective decline impairment. FINDINGS. This reveals behavioral provide insights into effectively selected conventional life-space assessments. Smartphone-estimated correlated highly self-reported measures, particularly specific Life-Space Assessment Measure Functional Mobility. indicate significant relationship patterns usage abilities. Specifically, lower ability linked unpredictable apps categories (high entropy), extended time, switching often. These correspond poorer performance range functions, including executive perceptual reasoning, visual learning, evidenced longer average daily total interaction times. Additionally, variability duration memory learning tasks. Participants inhibition capabilities showed higher entropy durations, suggesting high reflect capacity. Moreover, proportion time spent IADLs-related (shopping, food ordering, maps navigation, taxi/ride-share) abilities terms sustained functions. result suggests spending proportionally these conduct activities independently. CONCLUSIONS. underscores potential analysis providing near real-time, nuanced view functioning detecting changes. Overall, findings propose current characterizing self-reports. also demonstrates engaging comprehensive combines intensive testing collection.--Author's abstract

Язык: Английский

Процитировано

0