Prediction of Assistive Technology Adoption for People with Dementia

Research output: Chapter in Book/Report/Conference proceedingChapter

1 Citation (Scopus)

Abstract

Assistive technology can enhance the level of independence of people with dementia thereby increasing the possibility of remaining in their own homes. It is important that suitable technologies are selected for people with dementia, due to their reluctant to change. In our work, a predictive model has been developed for technology adoption of a Mobile Phone‐based Video Streaming solution developed for people with dementia, taking account of individual characteristics. Relevant features for technology adoption were identified and highlighted. A decision tree was then trained based on these features using Quinlan’s C4.5 algorithm. For the evaluation, repeated cross-validation was performed. Results are promising and comparable with those achieved using a logistic regression model. Statistical tests show no significant difference between the performance of a decision tree model and a logistic regression model (p=0.894). Also, the decision tree demonstrates graphically the decision making process with transparency, which is a desirable feature within healthcare based applications. In addition, the decision tree provides ease of use and interpretation and hence is easier for healthcare professionals to understand and to use both appropriately and confidently.
LanguageEnglish
Title of host publicationHealth Information Science, Lecture Notes in Computer Science
Place of PublicationLondon
Pages160-171
Volume7798
DOIs
Publication statusPublished - 2013

Fingerprint

Decision trees
Logistics
Statistical tests
Video streaming
Mobile phones
Transparency
Decision making

Cite this

Zhang, Shuai ; McClean, Sally I ; Nugent, CD ; O'Neill, Sonja ; Donnelly, Mark ; Galway, Leo ; Scotney, Bryan ; Cleland, Ian. / Prediction of Assistive Technology Adoption for People with Dementia. Health Information Science, Lecture Notes in Computer Science. Vol. 7798 London, 2013. pp. 160-171
@inbook{b826857ff7614c01acfd8a3cc956433b,
title = "Prediction of Assistive Technology Adoption for People with Dementia",
abstract = "Assistive technology can enhance the level of independence of people with dementia thereby increasing the possibility of remaining in their own homes. It is important that suitable technologies are selected for people with dementia, due to their reluctant to change. In our work, a predictive model has been developed for technology adoption of a Mobile Phone‐based Video Streaming solution developed for people with dementia, taking account of individual characteristics. Relevant features for technology adoption were identified and highlighted. A decision tree was then trained based on these features using Quinlan’s C4.5 algorithm. For the evaluation, repeated cross-validation was performed. Results are promising and comparable with those achieved using a logistic regression model. Statistical tests show no significant difference between the performance of a decision tree model and a logistic regression model (p=0.894). Also, the decision tree demonstrates graphically the decision making process with transparency, which is a desirable feature within healthcare based applications. In addition, the decision tree provides ease of use and interpretation and hence is easier for healthcare professionals to understand and to use both appropriately and confidently.",
author = "Shuai Zhang and McClean, {Sally I} and CD Nugent and Sonja O'Neill and Mark Donnelly and Leo Galway and Bryan Scotney and Ian Cleland",
year = "2013",
doi = "10.1007/978-3-642-37899-7_14",
language = "English",
isbn = "978-3-642-37898-0",
volume = "7798",
pages = "160--171",
booktitle = "Health Information Science, Lecture Notes in Computer Science",

}

Prediction of Assistive Technology Adoption for People with Dementia. / Zhang, Shuai; McClean, Sally I; Nugent, CD; O'Neill, Sonja; Donnelly, Mark; Galway, Leo; Scotney, Bryan; Cleland, Ian.

Health Information Science, Lecture Notes in Computer Science. Vol. 7798 London, 2013. p. 160-171.

Research output: Chapter in Book/Report/Conference proceedingChapter

TY - CHAP

T1 - Prediction of Assistive Technology Adoption for People with Dementia

AU - Zhang, Shuai

AU - McClean, Sally I

AU - Nugent, CD

AU - O'Neill, Sonja

AU - Donnelly, Mark

AU - Galway, Leo

AU - Scotney, Bryan

AU - Cleland, Ian

PY - 2013

Y1 - 2013

N2 - Assistive technology can enhance the level of independence of people with dementia thereby increasing the possibility of remaining in their own homes. It is important that suitable technologies are selected for people with dementia, due to their reluctant to change. In our work, a predictive model has been developed for technology adoption of a Mobile Phone‐based Video Streaming solution developed for people with dementia, taking account of individual characteristics. Relevant features for technology adoption were identified and highlighted. A decision tree was then trained based on these features using Quinlan’s C4.5 algorithm. For the evaluation, repeated cross-validation was performed. Results are promising and comparable with those achieved using a logistic regression model. Statistical tests show no significant difference between the performance of a decision tree model and a logistic regression model (p=0.894). Also, the decision tree demonstrates graphically the decision making process with transparency, which is a desirable feature within healthcare based applications. In addition, the decision tree provides ease of use and interpretation and hence is easier for healthcare professionals to understand and to use both appropriately and confidently.

AB - Assistive technology can enhance the level of independence of people with dementia thereby increasing the possibility of remaining in their own homes. It is important that suitable technologies are selected for people with dementia, due to their reluctant to change. In our work, a predictive model has been developed for technology adoption of a Mobile Phone‐based Video Streaming solution developed for people with dementia, taking account of individual characteristics. Relevant features for technology adoption were identified and highlighted. A decision tree was then trained based on these features using Quinlan’s C4.5 algorithm. For the evaluation, repeated cross-validation was performed. Results are promising and comparable with those achieved using a logistic regression model. Statistical tests show no significant difference between the performance of a decision tree model and a logistic regression model (p=0.894). Also, the decision tree demonstrates graphically the decision making process with transparency, which is a desirable feature within healthcare based applications. In addition, the decision tree provides ease of use and interpretation and hence is easier for healthcare professionals to understand and to use both appropriately and confidently.

U2 - 10.1007/978-3-642-37899-7_14

DO - 10.1007/978-3-642-37899-7_14

M3 - Chapter

SN - 978-3-642-37898-0

VL - 7798

SP - 160

EP - 171

BT - Health Information Science, Lecture Notes in Computer Science

CY - London

ER -

Zhang S, McClean SI, Nugent CD, O'Neill S, Donnelly M, Galway L et al. Prediction of Assistive Technology Adoption for People with Dementia. In Health Information Science, Lecture Notes in Computer Science. Vol. 7798. London. 2013. p. 160-171 https://doi.org/10.1007/978-3-642-37899-7_14