Development of an intelligent e-healthcare system for the domestic care industry
| DOI | https://doi.org/10.1108/IMDS-08-2016-0342 |
| Pages | 1426-1445 |
| Date | 14 August 2017 |
| Published date | 14 August 2017 |
| Author | Bennie Wong,G.T.S. Ho,Eric Tsui |
| Subject Matter | Information & knowledge management,Information systems,Data management systems,Knowledge management,Knowledge sharing,Management science & operations,Supply chain management,Supply chain information systems,Logistics,Quality management/systems |
Development of an intelligent
e-healthcare system for the
domestic care industry
Bennie Wong, G.T.S. Ho and Eric Tsui
Department of Industrial and Systems Engineering,
The Hong Kong Polytechnic University, Hong Kong, Hong Kong
Abstract
Purpose –In view of the elderly caregiving service being in high demand nowadays, the purpose of this
paper is to develop an intelligent e-healthcare system for the domestic care industry by using the Internet of
Things (IoTs) and Fuzzy Association Rule Mining (FARM) approach.
Design/methodology/approach –The IoTs connected with the e-healthcare system collect real-time vital
sign monitoring data for the e-healthcare system. The FARM approach helps to identify the hidden
relationshipsbetween the datarecords in the e-healthcaresystem to support the elderlycare management tasks.
Findings –To evaluate the proposed system and approach, a case study was carried out to identify the
association between the specific collected demographic data, behavior data and the health measurements data
in the e-healthcare system. It is found that the discovered rules are useful for the care management tasks in
the elderly healthcare service.
Originality/value –Knowledge discovery in databases uses various data mining techniques and rule-based
artificial intelligence algorithms. This paper demonstrates complete processes on how an e-healthcare system
connected with IoTs can support the elderly care services via a data collection phase, data analysis phase and
data reporting phase by using the FARM to evaluate the fuzzy sets of the data attributes. The caregivers can
use the discovered rules for proactive decision support of healthcare services and to improve the overall
service quality by enhancing the elderly healthcare service responsiveness.
Keywords Internet of Things, e-Healthcare system, Elderly care service, Fuzzy Association Rule Mining
Paper type Research paper
1. Introduction
Population ageing is a common phenomenon around the world. Taking Hong Kong as an
example, the proportion of the population aged 65 and over is projected to rise markedly
from 15 percent in 2014 to 33 percent in 2064 (Financial Secretary’s Office HKSAR, 2013).
As a result, due to the increasing needs and to cope with the associated challenges,
e-healthcare systems for supporting elderly healthcare such as real-time non-invasively
biomedical monitoring without affecting the normal life of a person have been introduced.
Although there are large amounts of information about individual health records in the
e-healthcare system, effective analysis tools are lacking for discovering the hidden
relationships in the available data. In addition, it is found that in recent years, much research
work has been done to support disease prediction or diagnosis, disease correlation, disease
risk analysis and drug reaction detection, but rarely in the elderly healthcare service.
While the World Health Organization continues to perform studies and surveys for gap
analysis of ageing and health in different countries (Paul et al., 2012), research has
emphasized that artificial intelligence is useful to discover hidden relationships and trends
in healthcare system (Abin et al., 2015). Among them, Association Rule Mining (ARM) is one
of the commonest methods that is used to find interesting relationships in the large
databases in the healthcare system. Deriving association rules in the healthcare sector has
become even more common recently ( Jan et al., 2014) due to its simplicity and effectiveness
Industrial Management & Data
Systems
Vol. 117 No. 7, 2017
pp. 1426-1445
© Emerald PublishingLimited
0263-5577
DOI 10.1108/IMDS-08-2016-0342
Received 25 August 2016
Revised 18 January 2017
21 March 2017
29 March 2017
30 March 2017
Accepted 5 April 2017
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/0263-5577.htm
The authors would like to thank the Department of Industrial and Systems Engineering of the Hong
Kong Polytechnic University for supporting this research.
1426
IMDS
117,7
in reflecting human interpretation of the defined categories. In order to handle linguistic
representation of the healthcare data more effectively, Fuzzy Association Rule Mining
(FARM) was deployed in the e-healthcare system for elderly care services developed in this
research in order to have a more realistic and practical classification of the data attributes in
their relationships. By converting the crisp values of the healthcare data with membership
functions into different data clusters, depending on their closeness to the predefined
member categories, it addressed the problem of dealing with data uncertainties that fall into
sharp value boundaries (Bilal et al., 2013). Therefore, it is more efficient and effective to
extract potentially interesting rules which are useful in providing health status predications
and to drawing special attention to the targeted elderly people under the healthcare
monitoring. The collected data are expressed in linguistic terms which makes them more
natural and understandable.
Owing to the need for quick responses in managing healthcare tasks, this paper
describes different data processing phases of the e-healthcare system from data collection,
analysis and reporting processes. Apart from the potentially useful knowledge discovered in
the data analysis phase for supporting proactive healthcare service, the e-healthcare system
provides standard user interfaces and predefined procedures to safeguard the operations of
the stakeholders, who may have different level of expertise in the caregiving service
provision. This ensures the operation procedures are following the formalized work flows.
The development not only can help save the lives of the elderly but can deliver a more
reliable healthcare service and lead to higher customer satisfaction.
To analyze the available data in the e-healthcare system, different categories of
attributes are examined and selected. Referring to the previous studies related to disease
diagnosis and the finding of relationships between the demographic factors and health-
related behavior (Kuwahara et al., 2004) “Age”and “Education”characteristics are chosen
for the evaluation. Similarly, from the studies on the behavior risk factors to adverse health
outcomes (Azari, 2006), “Smoking”and “Drinking”behavior factors are chosen for the
analysis. In addition, since health condition monitoring for the elderly (Ayman et al., 2014)
commonly uses the “Body And Mass Index (BMI)”,“Norton scale (Norton)”(Marta et al.,
2012) and “Modified Early Warning Signal (MEWS)”(Cei et al., 2009) as the measurement
parameters, they are also the target data attributes in the data analysis for determining their
association relationships.
The remainder of this paper is organized as follows. The selection of the health data sets,
fuzzy theory and ARM techniques’related studies are reviewed in Section 2. Section 3
introduces the proposed FARM approach and Section 4 demonstrates the approach with an
example. Section 5 discusses an experiment for extracting the rules in sample health data.
Section 6 gives the conclusions and recommendations for future work.
2. Related studies
Today, the demand for elderly healthcare service is intense. The healthcare industry has
evolved to deployintelligent e-healthcaresystems to support the servicedelivery as well as to
enhance customer satisfaction. Recent research has explored the possibilities of integrating
the Internet of Things(IoTs) in the e-healthcare system to monitorthe patient’s health status
(Swiatek and Rucinski, 2013). Through the remote sensors of the IoTs infrastructure
(Vishakha and Sanjeev, 2015), the health monitoring status can be collected in real time by a
wired or wireless transmission network to the central application server (Yang et al., 2014).
Certainly,artificial intelligenceis widely adopted in the healthcare industryin order to provide
health and diseases analysis, prediction or detection. Among the artificial intelligence
methodologiesthat are used in the data mining processes in the medicalarea, FARM is one of
the popular techniques being utilized (Sunita and Vyas, 2010). WhileARM is used to identify
the relationshipsof the crisp data values, FARM (Honget al., 2003) has been propo sed to solve
1427
Intelligent
e-healthcare
system
To continue reading
Request your trial