Acoustic gait analysis for elderly care v1

  • Research type

    Research Study

  • Full title

    AI-based acoustic analysis of gait patterns for non-invasive elderly care

  • IRAS ID

    300124

  • Contact name

    Christos Efstratiou

  • Contact email

    c.efstratiou@kent.ac.uk

  • Sponsor organisation

    University of Kent

  • Duration of Study in the UK

    0 years, 9 months, 0 days

  • Research summary

    The primary focus of this study is to explore the use of acoustic sensing to capture the gait patterns of older people and people living with dementia (PLWD). Through the use of machine learning (ML), we aim to extract unique features from the gait patterns, and explore the feasibility of predicting the risk of fall accidents, through the analysis of gait patterns. The objective is defined to address an unmet need among PLWD, that does not rely on any wearable devices that are generally considered not fitting for people with dementia.
    This study will involve the use of datasets collected by assistive technologies developed by MiiCare Ltd, which provides assistive living healthcare solutions for elderly users. The technology developed by MiiCare is able to collect acoustic signals passively from within the living environment of elderly users. The study aims to utilise datasets that MiiCare technology can collect through the deployment of their technology within elderly care facilities. We aim to utilise datasets corresponding to two cohorts of older people, one cohort of people with dementia, and another of non dementia people. The study will allow us to collect acoustic signals of participants within their living environment, and help us develop the necessary ML algorithms to identify gait pattern characteristics of each individual, and investigate correlations of such patterns with the risk of accidents as reported by the participants’ recent history.
    The successful completion of this project will enable the development of non invasive technology that can detect and predict potential increase of falling risk of older people and PLWD, without the need of any invasive wearable technology.

  • REC name

    Wales REC 6

  • REC reference

    21/WA/0330

  • Date of REC Opinion

    22 Oct 2021

  • REC opinion

    Favourable Opinion