Modelling Audiograms for People with Dementia Who Experience Hearing Loss Using Multiple Linear Regression Method

This paper presents a mathematical model of audiogram by using multiple regression polynomial. The mathematical model will be used to represent audiograms of individuals with dementia who experience hearing loss. The mathematical model will help in designing an optimum hearing aid for the people with dementia. Research studies state that there is association between hearing loss and dementia. In addition, researches show that individuals who experience hearing loss can develop dementia or it can worsen their case. World health organizations recognize dementia as public health priority. In research, hearing aid designs are always being tested based on matching standard set of audiograms and some designs are tested using mild to moderate hearing loss audiogram only. In this paper, the authors introduce a mathematical model to represent the measured audiograms for the cohort of the study. This model is developed based on measured data for seven subjects' experiencing hearing problems and suffering from dementia. Multiple linear regression is used to find the polynomial that represent the measured data based on minimizing the difference between the measured data and the obtained polynomial. SPSS is used to simulate the data and calculate the required regression polynomial coefficients for both ears hearing levels.