|Bartosz Trzaskowski, W. Wiktor Jędrzejczak, Edyta Piłka, Krzysztof Kochanek, Henryk Skarżyński|
Introduction. Auditory brainstem responses (ABRs) are one of
the most commonly used methods for objective determination
of hearing threshold. However, the part related to visual scoring
of ABRs influences the risk of discrepancies in the results of the
assessments performed by different testers. Development of a
system capable of automatic response detection could add to
standardization and objectification of the results of ABR tests.
Aim. The aim of this study was to develop a system for automatic
detection of auditory brainstem responses and to evaluate
Material and methods. The system was based on signal
processing and statistical methods. Its performance was tested
using a large set of simulated data, as well as a set of real traces
recorded in a group of patients with normal hearing and with
Results. Differences in scores assigned by the system and human
experts were not significantly higher than differences between
experts typically observed in clinical practice.
Conclusions. The results suggest, that presented algorithm
could be used with success in a system supporting ABR operator.
keywords: słuchowe potencjały wywołane pnia mózgu, automatyczna detekcja odpowiedzi, ABR auditory brainstem response, automatic responsedetection, ABR
pages: from 137 to 147
|estimated time of download (294 kB)|