Intelligibility and Quality of Hearing Aid Processed Speech: An Objective Comparison of Prescriptive Formulae

Abstract

The fitting hearing aid is the process of fine-tuning the hearing aid parameters to maximize the audibility, intelligibility & quality of the speech. The present study aimed to verify the intelligibility and quality of hearing aid processed speech across companies and prescriptive formulae through objective measures such as Hearing Aid Speech Perception Index (HASPI), Hearing Aid Speech Quality Index (HASQI), Speech Intelligibility Index (SII) and Long Term Average Speech Spectrum (LTASS). The study also evaluated the objective measures of hearing aid processed speech for Kannada sentences and compared them with the International Speech Test Stimuli (ISTS) sentences. The stimulus (Kannada sentences and ISTS) was delivered through the loudspeaker & the recordings obtained from the manikin were stored and collected through the sound level meter. The hearing aids of five companies (Oticon, Phonak, Resound, Starkey & Danavox) programmed to four formulae: Company fit, NAL NL 1, NAL NL 2 & DSLv5 were used to obtain the recordings. The results revealed that the HASPI and HASQI yielded similar values for Kannada and ISTS sentences for all the formulae across all companies. The SII values for ISTS was comparatively higher than Kannada sentences across companies in each formula. When the companies are compared, hearing aids of Danavox company consistently scored significantly higher SII values for all prescriptive formulae, while among the formulae company fit emerged with higher SII values across companies. The LTASS revealed differences for low, mid & high frequency LTASS across companies and prescriptive formulae. It can be concluded that Danavox company yielded better speech intelligibility, while among the formulas company fit can be used for better speech intelligibility (SII). The reasons for each of these findings are highlighted and recommendations for best practices on the use of objective measures for verifying the hearing aid output in clinical setups are suggested.