Minimal repetition BAER signal enhancement by filtering

Brain stem audiometry results in short latency signals severely contaminated by noise from a variety of sources. To obtain a signal of diagnostic quality, it is customary to perform an ensemble average using responses from up to 2000 repetitive excitations, which implies that the testing process can be both cumbersome and uncomfortable for the patient. The goal has always been to reduce the number of stimulations and to shorten the test period thus creating the need to obtain diagnostic features using ensembles comprising responses from fewer stimulus repetitions. This paper is based on data obtained from real tests on a patient, and describes a filtering method which has been applied to enhance the diagnostic features of the evoked potentials. The basis filter has been derived from the characteristics of the clinical quality signal from brain stem auditory evoked response (BAER) testing. The approach here is relatively simple when compared to complex methods described elsewhere for enhancing signals from evoked responses. It is remarkable that diagnostic features were obtained by processing an ensemble containing as few as 10 realisations of auditory evoked potentials.