Otoacoustic emissions and improved pass/fail separation using wavelet based denoising

A new technique is presented for the purpose of improving the pass/fail separation during a transient evoked otoacoustic emission screening procedure. The technique combines signal decomposition in octave bands using the wavelet transform with denoising, separate reconstruction of components representing octave bands and time windowing. The cross-correlation coefficient between processed TEOAE signals was used as a pass/fail criterion and assessed in relation to the mean hearing threshold. The performance is presented in terms of receiver operating characteristics for a database of 5214 individuals. The results show that the specificity improves by 12% in comparison with conventional TEOAE analysis when the sensitivity is held constant at 90%.