東華大學圖書館 |

Multidimensional Digital Biomarker of Cognitive Health: Unobtrusive and Continuous Monitoring of Cognitive Changes Using Smartphones.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Multidimensional Digital Biomarker of Cognitive Health: Unobtrusive and Continuous Monitoring of Cognitive Changes Using Smartphones./
作者:	Kos, Maciej.
出版者:	Ann Arbor : ProQuest Dissertations & Theses, : 2024,
面頁冊數:	136 p.
附註:	Source: Dissertations Abstracts International, Volume: 85-09, Section: B.
Contained By:	Dissertations Abstracts International85-09B.
標題:	Computer science. -
電子資源:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30820189
ISBN:	9798381749977

Multidimensional Digital Biomarker of Cognitive Health: Unobtrusive and Continuous Monitoring of Cognitive Changes Using Smartphones.
Kos, Maciej.

Multidimensional Digital Biomarker of Cognitive Health: Unobtrusive and Continuous Monitoring of Cognitive Changes Using Smartphones. - Ann Arbor : ProQuest Dissertations & Theses, 2024 - 136 p.

Source: Dissertations Abstracts International, Volume: 85-09, Section: B.

Thesis (Ph.D.)--Northeastern University, 2024.

This item must not be sold to any third party vendors.

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

ISBN: 9798381749977Subjects--Topical Terms:

523869
Computer science.
Subjects--Index Terms:

Cognitive impairment

Multidimensional Digital Biomarker of Cognitive Health: Unobtrusive and Continuous Monitoring of Cognitive Changes Using Smartphones.
LDR:07304nmm a2200409 4500 001 2396573
005 20240611104333.5
006 m o d
007 cr#unu||||||||
008 251215s2024 ||||||||||||||||| ||eng d
020 $a 9798381749977
035 $a (MiAaPQ)AAI30820189
035 $a AAI30820189
040 $a MiAaPQ $c MiAaPQ
100 1 $a Kos, Maciej. $3 3766269
245 1 0 $a Multidimensional Digital Biomarker of Cognitive Health: Unobtrusive and Continuous Monitoring of Cognitive Changes Using Smartphones.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2024
300 $a 136 p.
500 $a Source: Dissertations Abstracts International, Volume: 85-09, Section: B.
500 $a Advisor: Pavel, Misha.
502 $a Thesis (Ph.D.)--Northeastern University, 2024.
506 $a This item must not be sold to any third party vendors.
520 $a Background. Cognitive functionality is a critical determinant of quality of 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, cognitive impairment may start in early and middle adulthood, with between 3.1% and 21% of the global population suffering from cognitive impairment depending on age and location. The number of dementia cases alone is expected to reach 115 million by 2050. Yet, cognitive impairments are typically detected late in the decline. Positive diagnoses are often so late in the process that there is not much that can be done. Hence, early detection of and intervention are necessary to intervene as well as to develop effective treatments. To facilitate successful aging, effective methods are needed for monitoring cognition to detect early signs of mild cognitive impairment and dementia. Furthermore, the development of efficacious therapeutics necessitates the monitoring of subtle changes in cognitive functions over time. However, the effectiveness of existing neuropsychological assessments is diminished by their sporadicity, difficulty in accounting for the context-dependent nature of individuals' health (e.g., having a "good" or a "bad" day), and reliance on frequently inaccurate individual and caregiver reports. Thus, new approaches are needed for objective and ecologically valid assessment of cognitive functions and for early detection of impaired cognition associated with neurocognitive disorders.Approach. Current mHealth and AI approaches enable continuous context inference from smartphone use and location data. Studies involving younger and older adults suggest that characteristics of an individual's mobile app use and typing speed correlate with working memory, attention, and psychomotor function. Individuals' reports of Instrumental Activities of Daily Living (IADLs) can be significantly enhanced with continuous, objective context inferences and estimates of engagement in digital equivalents of a subset of IADLs (e.g., mobile shopping and banking, instant messaging) from passively collected smartphone-based data, contributing to earlier and more accurate diagnoses of neurocognitive disorders. Therefore, to address the challenges associated with neuropsychological tests, I designed an approach to augment the assessments with smartphone-derived estimates of cognitive changes, as interactions with smartphones require a variety of cognitive skills. I used data collected by smartphones to inform the development of a multidimensional digital biomarker (henceforth, digital biomarker). The proposed digital biomarker - developed using a combination of AI/ML methods and mechanistic modeling - would enrich existing clinical tests with continuous and objective estimates of cognitive changes derived computationally and unobtrusively from mobile application use characteristics, location data, and motor aspects of smartphone interactions (e.g., typing speed), to inform both the early detection and diagnosis of impaired cognition associated with neurocognitive disorders and treatment personalization for their amelioration. To assess the feasibility of developing the biomarker, I conducted a repeated-measures pilot study consisting of continuous data collection using an app installed on participants' smartphones and neuropsychological data acquisition during two lab visits, four months apart. The study participants were 22 middle-aged participants with levels of cognitive impairment varying from no impairment to subjective cognitive decline to mild cognitive impairment.Findings. This pilot study reveals that behavioral data derived from smartphones may provide insights into cognitive impairment as effectively as selected conventional cognitive tests and life-space assessments. Smartphone-estimated life-space correlated highly with conventional self-reported life-space measures, particularly in specific levels of the Life-Space Assessment Measure of Functional Mobility. The data indicate a significant relationship between specific patterns of app usage and cognitive abilities. Specifically, lower cognitive ability is linked to more unpredictable use of apps and app categories (high entropy), extended usage time, and switching between apps more often. These patterns correspond to poorer performance in a range of cognitive functions, including attention, executive functions, perceptual reasoning, visual memory, and learning, as evidenced by longer average daily total app interaction times. Additionally, variability in app use duration and app switching are associated with lower performance in visual memory and learning tasks. Participants with diminished working memory and inhibition capabilities showed higher entropy in app use durations, suggesting that high entropy may reflect lower behavioral inhibition and working memory capacity. Moreover, the higher proportion of time spent using IADLs-related app categories (shopping, food ordering, maps and navigation, taxi/ride-share) correlated with poorer cognitive abilities in terms of memory, sustained attention, and executive functions. This result suggests that spending proportionally more time in these categories may correspond to poorer ability to conduct these activities independently.Conclusions. The study underscores the potential of app usage analysis in providing a near real-time, nuanced view of individual functioning and detecting subtle behavioral changes. Overall, the findings propose that smartphone-based behavioral data may augment current approaches for characterizing cognitive impairment and individuals' life-space self-reports. The study also demonstrates the feasibility of engaging participants in a comprehensive data collection process that combines intensive cognitive lab testing with continuous smartphone data collection.
590 $a School code: 0160.
650 4 $a Computer science. $3 523869
650 4 $a Behavioral sciences. $3 529833
650 4 $a Health sciences. $3 3168359
653 $a Cognitive impairment
653 $a Computational modeling
653 $a Data science
653 $a Health monitoring
653 $a Mobile health
653 $a Personal health informatics
690 $a 0984
690 $a 0602
690 $a 0566
710 2 $a Northeastern University. $b Health Informatics. $3 3558537
773 0 $t Dissertations Abstracts International $g 85-09B.
790 $a 0160
791 $a Ph.D.
792 $a 2024
793 $a English
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30820189