An Automatic Recognition for the Auditory Brainstem Response Waveform

Abstract

The Auditory Brainstem Response (ABR) is Brainstem Auditory Evoked potentials and often used in the neurophysiology. The waveform of ABR is usually recorded right after stimulation applied, as a response characteristic with a five peaks. These each peaks from the recording electrodes is identified by (a) neural transmission times and (b) amplitude in measured potentials. These sequential of few msec peaks with the amplitude are all correlated each other to form a unique-pattern and that can be observed as a health-monitoring indicator. In this paper, an automatic recognition pattern for ABR waveform is proposed. Firstly, diverse ABR applications and recent techniques reviewed. Than, knowledge based information obtained from these recent techniques to develop a similar methodology, secondly to model the complete set of peaks in the ABR waveform. Several curve fitted functions tested to narrow down the suitable function to be used for the ABR model. The outcome is the parameter of this mathematical modelling of ABR pattern, and put forward the use for an automatic health diagnostic tool as a machine learning application.