Evaluating the performance of an electronic hearing protection device with selective noise attenuation

Abstract

Objective: This study evaluated the performance of the recently developed IIUM e-HPD prototype in noise attenuation and speech intelligibility scores. Method: This study used a quasi-experimental study design. Firstly, the applications of digital signal algorithms and their effect on speech intelligibility for in-ear type e-HPD were discovered through a systematic review. First, six databases were searched using the following key concept of terms: “hearing protection device” and “algorithm”, and “speech intelligibility”. Next, a Modified Acoustic Test Fixture (MATF) test was conducted to measure and calculate the insertion loss (dB) at different ear conditions (Open ear, Occluded ear, Record and playback algorithm and NLMS filter) at 500 and 2000 Hz at 80 and 90 dB SPL. Meanwhile, the Modified Hearing in Noise Test (MHINT) were conducted to measure the speech recognition threshold in noise (dB SNR) using The Malay Hearing in Noise Test speech material. The study compared the insertion loss (dB) between the ear conditions at at 500 and 2000 Hz at 80 and 90 dB SPL . The speech recognition threshold in noise was compared between the all four experimental conditions. Results: The systematic review revealed nine references testing the speech intelligibility of the in-ear type of e-HPD. These findings were identified under three themes: Population, algorithm and speech intelligibility tests. The insertion loss calculated showed the highest insertion loss achieved by the occluded ear at 500 and 2000 Hz. Negative insertion loss observed at 2000 Hz for the open ear, the occluded ear and the Normalised Least-Mean-Square (NLMS) filter with the NLMS filter were the lowest insertion loss calculated. RM-ANOVA was performed on the insertion loss (dB) from MATF test results it showed a significant difference (p<.05) between the main effect and pairwise comparison of the ear conditions, noise level, and frequencies of interest. Polynomial contrast showed significant linear, quadratic and cubic components (P>0.10) except at the interaction between condition, noise level and frequencies of interest. A one way RM-ANOVA of the MHINT on the ear conditions indicated a significant main (p<.05) while the pairwise comparison of the ear conditions showed a significant difference between the open ear and NLMS filter only. Conclusion: In summary, the developed e-HPD prototype was able to attenuate noise although not as comparable to an occluded ear while preserving the speech intelligility in noise compared with open ear.