sis 15(5): e6

Research Article

Behavioural health analytics using mobile phones

Download447 downloads
  • @ARTICLE{10.4108/sis.2.5.e6,
        author={P. Wlodarczak and P. Soar and M. Ally},
        title={Behavioural health analytics using mobile phones},
        journal={EAI Endorsed Transactions on Scalable Information Systems},
        volume={2},
        number={5},
        publisher={ICST},
        journal_a={SIS},
        year={2015},
        month={7},
        keywords={Reality mining, Big Data, machine learning, eHealth, predictive analytics, behavioural health analytics, mobile sensing},
        doi={10.4108/sis.2.5.e6}
    }
    
  • P. Wlodarczak
    P. Soar
    M. Ally
    Year: 2015
    Behavioural health analytics using mobile phones
    SIS
    ICST
    DOI: 10.4108/sis.2.5.e6
P. Wlodarczak1,*, P. Soar1, M. Ally1
  • 1: University of Southern Queensland, Toowoomba, Queensland, Australia
*Contact email: wlodarczak@gmail.com

Abstract

Big Data analytics in healthcare has become a very active area of research since it promises to reduce costs and to improve health care quality. Behavioural analytics analyses a patients behavioural patterns with the goal of early detection if a patient becomes symptomatic and triggering treatment even before a disease outbreak happens. Behavioural analytics allows a more precise and personalised treatment and can even monitor whole populations for events such as epidemic outbreaks. With the prevalence of mobile phones, they have been used to monitor the health of patients by analysing their behavioural and movement patterns. Cell phones are always on devices and are usually close to their users. As such they can be used as social sensors to create "automated diaries" of their users. Specialised apps passively collect and analyse user data to detect if a patient shows some deviant behaviour indicating he has become symptomatic. These apps first learn a patients normal daily patterns and alert a health care centre if it detects a deviant behaviour. The health care centre can then call the patient and check on his well-being. These apps use machine learning techniques to for reality mining and predictive analysis. This paper describes some of these techniques that have been adopted recently in eHealth apps.