A fast algorithm for the measurement of stimulus frequency otoacoustic emission suppression tuning curves

Xing Dongjia,Gong Qin
DOI: https://doi.org/10.1016/j.apacoust.2017.06.017
IF: 3.614
2018-01-01
Applied Acoustics
Abstract:Frequency selectivity is an important indicator of auditory function in the human ear. Stimulus frequency otoacoustic emission (SFOAE) suppression tuning curves (STCs) have great potential in the objective analysis of human auditory frequency selectivity but are costly to measure by the traditional algorithm using pure-tone adaptive method. To improve the measurement efficiency of the SFOAE STCs, a faster algorithm based on a suppressor with gradiently changing intensity and interpolation is proposed in this paper. Twelve subjects participated in this study by measuring their SFOAE STCs through the traditional and fast algorithm. The results shows that the average correlation coefficients of the SFOAE STCs measured using the fast and the traditional algorithms at different probe frequencies is 0.94. And the measurement speed of the fast algorithm was approximately 2.10 times higher than the traditional one. Also, the fast algorithm is more efficient than the algorithms of Keefe et al. (2008) and Charaziak et al. (2013). Experimental evidence is provided that the proposed fast algorithm greatly improves the measurement speed and reduces the processing time, exhibiting good accuracy and reliability.
What problem does this paper attempt to address?