8th International Conference on Pervasive Computing Technologies for Healthcare

Research Article

Evaluating Auditory Contexts and Their Impacts on Hearing Aid Outcomes with Mobile Phones

  • @INPROCEEDINGS{10.4108/icst.pervasivehealth.2014.254952,
        author={Syed Shabih Hasan and Octav Chipara and Yu-Hsiang Wu and Nazan Aksan},
        title={Evaluating Auditory Contexts and Their Impacts on Hearing Aid Outcomes with Mobile Phones},
        proceedings={8th International Conference on Pervasive Computing Technologies for Healthcare},
        publisher={ICST},
        proceedings_a={PERVASIVEHEALTH},
        year={2014},
        month={7},
        keywords={hearing aids computer based ema context-aware sensing},
        doi={10.4108/icst.pervasivehealth.2014.254952}
    }
    
  • Syed Shabih Hasan
    Octav Chipara
    Yu-Hsiang Wu
    Nazan Aksan
    Year: 2014
    Evaluating Auditory Contexts and Their Impacts on Hearing Aid Outcomes with Mobile Phones
    PERVASIVEHEALTH
    ACM
    DOI: 10.4108/icst.pervasivehealth.2014.254952
Syed Shabih Hasan1, Octav Chipara1,*, Yu-Hsiang Wu1, Nazan Aksan1
  • 1: University of Iowa
*Contact email: octav-chipara@uiowa.edu

Abstract

This paper evaluates the relationship between auditory contexts, hearing aid features, and hearing outcomes based on real-world measurements. We use a mobile phone application to concurrently evaluate the auditory contexts and hearing aid outcomes using Ecological Momentary Assessments. The collected dataset includes 3437 surveys collected from nineteen patients over ten months. Our analysis indicates that the most frequent listening activities were conversations (32.7% of the time) and listening to media (30.7% of the time), commonly occurring at home, in predominantly quiet environments. Subjects do not attribute equal importance to hearing well in all auditory contexts: it is more important to hear well in contexts that involve social interactions. We show that hearing aid outcomes measures are moderately correlated. By leveraging on these correlations, we propose a method of combining measurements of hearing aid outcomes into a single score to reduce measurement error. Finally, we show that it is possible to discriminate between poor and good hearing aid outcomes with an accuracy of 78% solely based on auditory contexts and hearing aid features. This shows the central role that auditory contexts play in understanding hearing aid outcomes in situ.