Dissertação de Mestrado
Técnicas de detecção objetiva de resposta evocada auditiva em regimepermanente obtida por tons e ruídos modulados em amplitude
Fecha
2014-09-16Autor
Jeferson Jhone da Silva
Institución
Resumen
The auditory steady state responses (ASSR) have been considered an important audiological assessment tool mainly for non-collaborative individuals. Thus, many studies reported ASSR as promising for threshold investigation of multiple tones simultaneously. Recent studies reported its use in newborn hearing screening. Since the frequency specificity is not the main factor for this application, but rather the detection speed and the sensitivity to auditory losses, this work aims at studying the performance of ASSR using amplitude-modulated noise, in order to reduce exam duration and increase the detection rates. Therefore, different objective response detection (ORD) techniques were compared: Magnitude-Squared Coherence (MSC), Component Synchrony Measure (CSM) and Rice Detector (RD). The ASSR were obtained by the application of the following stimuli: broadband noise (BBN), low-band noise (LBN), high-band noise (HBN) and simultaneously low- and high- band noise (DBN, double band noise), all of them amplitude-modulated. Moreover, a stimulus composed of the sum of amplitude-modulated tones (AM) was also implemented for comparison with the noise-based ones, all of them presented by insert earphone at the intensities of 25, 45 and 55dBSPL (SPL Sound Pressure Level). EEG signals during auditory stimulation were obtained from 19 volunteers (38 ears) with normal hearing, verified by tonal audiometry. The results of detection time and percentage for the different techniques and different stimuli were statistically compared. Among the techniques, MSC and CSM presented similar results for detection time and rates for all intensities. RD showed significantly lower detection rates than other techniques and detection time statistically higher than the best technique for each intensity. Among studied types of stimuli, the noise-based ones showed better performance for ASSR detection, with BBN, HBN and DBN-high presenting the highest percentages and lowest detection times.