Análise de potenciais evocados auditivos de tronco encefálico através de atributos instantâneos de séries temporais

Abstract

The extraction of signal components from Auditory Brainstem Response (ABR), referred to as Jewett waves, is an important step in studies regarding the investigation of the behavior of the auditory system. However, this evaluation is complex and depends on the experience of the specialist. In this context, the objectives of this research are to compare ABR analyses provided by different examiners, and also to verify the possibility of discrimination of Jewett waves through the analysis of instantaneous attributes (IA) of time-series, estimated by the Hilbert Transform (HT). The ABR data were collected from healthy subjects, following the criterion of auditory normality, and analyzed by four examiners. A statistical study aiming to verify the coherence among these analyses was performed. The HT was then employed for calculating the IA from ABR signals. The compatibility among the results provided by the examiners was verified. The analysis of IA suggests that they can be used as discriminative features for the different waves that compose the ABR. These results are relevant for the development of tools for automatic ABR analysis. Furthermore, it contributes to investigations that aim to discriminate ABR related to distinct disorders.