Combination of derived non-linear and linear methods to increase the reproducibility of transiently evoked otoacoustic emissions: Combinacion de metodos derivados, lineales y no lineales, para incrementar la reproducibilidad de las emisiones otoacusticas evocadas por transientes

This paper proposes a segmented average method that combines the derived non-linear response (DNLR) and linear-averaging methods to increase the reproducibility of transiently evoked otoacoustic emissions (TEOAEs). In this method, the DNLR method is used at the 2.5-P-ms segment of each TEOAE signal, and the linear-averaging method is used from P to 20 ms. This exploits the reduced-artefact advantage of the DNLR method and the high-reproducibility advantage of the linear-averaging method. To evaluate the performance of the proposed method, 26 normal ears were tested, with P selected as 8 ms. Each TEOAE signal obtained was decomposed into three scales using the discrete wavelet transform. Then the similarities and differences between the signals obtained with this method and with the DNLR method were assessed. The results show that the average reproducibility of the TEOAE signals increases from 78.1% to 88.7% when applying the proposed method.SumarioEste articulo propone un metodo segmentado promedio que co...