Fully Synthetic Longitudinal Real-World Data From Hearing Aid Wearers for Public Health Policy Modeling

• Published in: Frontiers in neuroscience 2019-08, Vol.13, p.850-850
• Main Authors: Christensen, Jeppe H, Pontoppidan, Niels H, Rossing, Rikke, Anisetti, Marco, Bamiou, Doris-Eva, Spanoudakis, George, Murdin, Louisa, Bibas, Thanos, Kikidiks, Dimitris, Dimakopoulos, Nikos, Giotis, Giorgos, Ecomomou, Apostolos
• Format: Article
• Language: English
• Subjects: Acoustics; Aging; Big Data; Cognitive ability; Compression; Computer science; Data analysis; evidence-based; Health care policy; Hearing aids; Hearing loss; longitudinal data; Neuroscience; Noise; Noise reduction; Older people; Privacy; Public health; public health policy; real-world data; Rehabilitation; Signal processing; Smartphones; Variables; Wireless networks
• ISSN: 1662-4548; 1662-453X; 1662-453X
• DOI: 10.3389/fnins.2019.00850

Disabling HL is associated with early cognitive decline in adults (Olusanya et al., 2014), and when unaddressed, HL restricts social integration and reduces employment and educational opportunities, hampers emotional well-being and, thus, poses an economic challenge at both the individual and national level (Wilson et al., 2017). [...]more and more individuals suffer from HL, which is primarily due to increases in everyday noise exposure and an increase of the aging population (World Health Organization, 2017). The hearing aids log data about the user's sound environment (Pontoppidan et al., 2018), hearing aid use (i.e., on/off) and hearing aid settings on a minute-by-minute basis; a phone app developed for the study collects information about the user's physical location via GPS (Dritsakis et al., 2018). [...]EVOTION will provide an evidence base for formulating and evaluating the impacts of public health policy pertaining to prevention, early diagnosis, and treatment/rehabilitation for adults with hearing impairment. Here, n is the size of the source input (rows), ε = 0.1, d is the number of variables, and k = 4. [...]the covariance between source parameters are preserved by allowing the empirical PDFs to be conditioned in the Bayesian network. [...]we do not have access to low-level details of the signal-processing taking place in the hearing aids. [...]we do not include real-time data on how the hearing aids autonomously reacts to the sound environment (e.g., adjusting noise reduction or compression characteristics).