Borrow it
Coverart for item
The Resource Computational algorithms for predictive health assessment, by Zahra Hajihashemi

Computational algorithms for predictive health assessment, by Zahra Hajihashemi

Creator
Author
Contributor
Summary
Rapid aging of the population in the US requires increased attention from health care providers and from the entire society as a whole. While the elderly population (aged over 65) will increase by 8% until 2050 in the developed countries, the working-age population (age between 15 and 64 years) will decrease and its ratio to the elderly population will decline from 4.3 to 2.3 [1]. While older adults prefer to live independently, many of the health conditions associated with old age, such as frailty, dementia, and risk of falling require increased attention and monitoring. However, independent living may lead to infrequent health assessments due to lack of continuous monitoring or fear of being institutionalized. Late health assessments may miss unreported complications, which in turn lead to poor long-term prognosis and quality of life [2]. A possible solution to prevent unreported health problems in independently living older adults is through automatic health monitoring systems. The aim of this dissertation is to use sensor network technology to detect changes in health status of elderly living alone, alert health care providers, and augment traditional health care. We review the Aging In Place (AIP) research and the sensor technology developed at the University of Missouri (MU) to support AIP. AIP represents ones ability to live in his/her own home safely and independently regardless of age, income, or medical condition. In this dissertation, we address two topics. First, we discuss the problem of measuring the temporal similarity of two multi-dimensional time series. The second topic of this work is predicting health patterns using time series similarities. For measuring the similarity of multidimensional time series, we focus on health care applications. We review already discovered similarity functions for time series. Inspired by the bioinformatics approach, we propose a new method to measure the similarity of two multi-dimensional time series. We introduce the idea of modified Smith-Waterman framework, Temporal Smith-Waterman (TSW), for Early Illness Recognition (EIR) that uses temporal sequences. We analyze the properties of TSW, introduce a faster way to compute it based on genetic programming, and propose a new method for searching large time series. Our approach doesnt require series data conversion to continuous format. Instead, we arrange all different sensor hits together with their time stamps into a one-dimensional sequence. Our method overcomes difficulties related to data uncertainty and aggregation that often arise when processing sensor data. For health pattern recognition, we describe a new framework that uses TSW for abnormal pattern recognition and its application to eldercare and early illness recognition. The new framework for predicting abnormal health patterns uses non-wearable sensors and sensor pattern similarity. A sensor pattern is classified as 2abnormal3 if it is much less similar to the previous ones. In this work, we consider 2abnormal3 days as days with documented nurse assessment in the electronic health record. To further refine the nature of the health assessment used in our EIR framework, we proposed a novel semi-supervised for clustering the nursing notes. This method represents each cluster of notes by a language model. We use the bi-grams model of terms to represent term to term dependencies and improve the expressivity of the proposed clustering method. We estimate the cluster model by a maximum a posterior probability (MAP) to address the challenge of highly imbalance nursing visit notes dataset. In this work, we also propose three methods for identification of deviations in patterns of activities of daily livings (ADL) of older adults and use them to generate alerts for the healthcare providers. ADLs such as bathroom visits can be monitored by automated in-home sensor systems. Our proposed methods find periodicity in sensor time series data using clustering, item set mining, and statistical approaches. A retrospective multiple case study (N=3) design was used to quantify bathroom visits as parts of the older adults daily routine, over a 10-day period. The performances of the proposed methods are evaluated using data collected from TigerPlace, an independent living community situated in Columbia, Missouri
Language
eng
Work
Publication
Extent
1 online resource (xxi, 214 pages)
Note
  • "A Dissertation presented to the Faculty of the Graduate School at the University of Missouri-Columbia Department of Computer Science College of Engineering In Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy."
  • Dissertation supervisor: Dr. Mihail Popescu
  • Includes vita
Instance
Label
Computational algorithms for predictive health assessment
Title
Computational algorithms for predictive health assessment
Statement of responsibility
by Zahra Hajihashemi
Creator
Contributor
Author
Thesis advisor
Subject
Genre
Language
eng
Summary
Rapid aging of the population in the US requires increased attention from health care providers and from the entire society as a whole. While the elderly population (aged over 65) will increase by 8% until 2050 in the developed countries, the working-age population (age between 15 and 64 years) will decrease and its ratio to the elderly population will decline from 4.3 to 2.3 [1]. While older adults prefer to live independently, many of the health conditions associated with old age, such as frailty, dementia, and risk of falling require increased attention and monitoring. However, independent living may lead to infrequent health assessments due to lack of continuous monitoring or fear of being institutionalized. Late health assessments may miss unreported complications, which in turn lead to poor long-term prognosis and quality of life [2]. A possible solution to prevent unreported health problems in independently living older adults is through automatic health monitoring systems. The aim of this dissertation is to use sensor network technology to detect changes in health status of elderly living alone, alert health care providers, and augment traditional health care. We review the Aging In Place (AIP) research and the sensor technology developed at the University of Missouri (MU) to support AIP. AIP represents ones ability to live in his/her own home safely and independently regardless of age, income, or medical condition. In this dissertation, we address two topics. First, we discuss the problem of measuring the temporal similarity of two multi-dimensional time series. The second topic of this work is predicting health patterns using time series similarities. For measuring the similarity of multidimensional time series, we focus on health care applications. We review already discovered similarity functions for time series. Inspired by the bioinformatics approach, we propose a new method to measure the similarity of two multi-dimensional time series. We introduce the idea of modified Smith-Waterman framework, Temporal Smith-Waterman (TSW), for Early Illness Recognition (EIR) that uses temporal sequences. We analyze the properties of TSW, introduce a faster way to compute it based on genetic programming, and propose a new method for searching large time series. Our approach doesnt require series data conversion to continuous format. Instead, we arrange all different sensor hits together with their time stamps into a one-dimensional sequence. Our method overcomes difficulties related to data uncertainty and aggregation that often arise when processing sensor data. For health pattern recognition, we describe a new framework that uses TSW for abnormal pattern recognition and its application to eldercare and early illness recognition. The new framework for predicting abnormal health patterns uses non-wearable sensors and sensor pattern similarity. A sensor pattern is classified as 2abnormal3 if it is much less similar to the previous ones. In this work, we consider 2abnormal3 days as days with documented nurse assessment in the electronic health record. To further refine the nature of the health assessment used in our EIR framework, we proposed a novel semi-supervised for clustering the nursing notes. This method represents each cluster of notes by a language model. We use the bi-grams model of terms to represent term to term dependencies and improve the expressivity of the proposed clustering method. We estimate the cluster model by a maximum a posterior probability (MAP) to address the challenge of highly imbalance nursing visit notes dataset. In this work, we also propose three methods for identification of deviations in patterns of activities of daily livings (ADL) of older adults and use them to generate alerts for the healthcare providers. ADLs such as bathroom visits can be monitored by automated in-home sensor systems. Our proposed methods find periodicity in sensor time series data using clustering, item set mining, and statistical approaches. A retrospective multiple case study (N=3) design was used to quantify bathroom visits as parts of the older adults daily routine, over a 10-day period. The performances of the proposed methods are evaluated using data collected from TigerPlace, an independent living community situated in Columbia, Missouri
Cataloging source
MUU
http://library.link/vocab/creatorName
Hajihashemi, Zahra
Degree
Ph. D.
Dissertation note
Thesis
Dissertation year
2015.
Government publication
government publication of a state province territory dependency etc
Granting institution
University of Missouri--Columbia
Illustrations
illustrations
Index
no index present
Literary form
non fiction
Nature of contents
  • dictionaries
  • bibliography
  • theses
http://library.link/vocab/relatedWorkOrContributorDate
1962-
http://library.link/vocab/relatedWorkOrContributorName
Popescu, Mihail
http://library.link/vocab/subjectName
  • Sensor networks
  • Older people
  • Living alone
  • Medical care
  • Missouri
Label
Computational algorithms for predictive health assessment, by Zahra Hajihashemi
Instantiates
Publication
Note
  • "A Dissertation presented to the Faculty of the Graduate School at the University of Missouri-Columbia Department of Computer Science College of Engineering In Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy."
  • Dissertation supervisor: Dr. Mihail Popescu
  • Includes vita
Bibliography note
Includes bibliographical references (pages 195-213)
Carrier category
online resource
Carrier category code
cr
Carrier MARC source
rdacarrier
Content category
text
Content type code
txt
Content type MARC source
rdacontent
Control code
962489396
Extent
1 online resource (xxi, 214 pages)
Form of item
online
Media category
computer
Media MARC source
rdamedia
Media type code
c
Other physical details
illustrations (some color)
Specific material designation
remote
System control number
(OCoLC)962489396
Label
Computational algorithms for predictive health assessment, by Zahra Hajihashemi
Publication
Note
  • "A Dissertation presented to the Faculty of the Graduate School at the University of Missouri-Columbia Department of Computer Science College of Engineering In Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy."
  • Dissertation supervisor: Dr. Mihail Popescu
  • Includes vita
Bibliography note
Includes bibliographical references (pages 195-213)
Carrier category
online resource
Carrier category code
cr
Carrier MARC source
rdacarrier
Content category
text
Content type code
txt
Content type MARC source
rdacontent
Control code
962489396
Extent
1 online resource (xxi, 214 pages)
Form of item
online
Media category
computer
Media MARC source
rdamedia
Media type code
c
Other physical details
illustrations (some color)
Specific material designation
remote
System control number
(OCoLC)962489396

Subject

Genre

Library Locations

  • Ellis LibraryBorrow it
    1020 Lowry Street, Columbia, MO, 65201, US
    38.944491 -92.326012
  • Engineering Library & Technology CommonsBorrow it
    W2001 Lafferre Hall, Columbia, MO, 65211, US
    38.946102 -92.330125
  • Fisher Delta Research CenterBorrow it
    2-64 Agricultural Bldg, Columbia, MO, 65201, US
    38.958397 -92.303491
  • Geological Sciences LibraryBorrow it
    201 Geological Sciences, Columbia, MO, 65211, US
    38.947375 -92.329062
  • J. Otto Lottes Health Sciences LibraryBorrow it
    1 Hospital Dr, Columbia, MO, 65201, US
    38.939544 -92.328377
  • Journalism LibraryBorrow it
    102 Reynolds Jrnlism Institute, Columbia, MO, 65211, US
    38.947290 -92.328025
  • Mathematical Sciences LibraryBorrow it
    104 Ellis Library, Columbia, MO, 65201, US
    38.944377 -92.326537
  • University ArchivesBorrow it
    Columbia, MO, 65201, US
  • University Archives McAlester AnnexBorrow it
    703 Lewis Hall, Columbia, MO, 65211, US
    38.934630 -92.342290
  • University of Missouri Libraries DepositoryBorrow it
    2908 Lemone Blvd, Columbia, MO, 65211, US
    38.919360 -92.291620
  • Zalk Veterinary Medical LibraryBorrow it
    Veterinary Medicine West, Columbia, MO, 65211, US
    38.941099 -92.317911

Library Links

Structured data from the Bibframe namespace is licensed under the Creative Commons Attribution 4.0 International License by University of Missouri Libraries. Additional terms may apply to data associated with third party namespaces.
Processing Feedback ...