ABSTRACT
Auditory evoked potential amplitude is typically much smaller than that of the background noise. When obtaining ABRs, this noise is minimized by the following methods: (1) ensuring low electrode impedance (EI) (i.e., less than 5 kΩ); (2) reducing noise from muscle activities by making the subject comfortable or using sedation; (3) dierential amplication, such as using three electrode leads per recording channel; (4) ltering the output of the bioamplier; (5) signal averaging; and (6) artifact rejection (Katz 2002). For dierential amplication, three electrodes are usually used that are referred to as the noninverting (positive), the inverting (negative), and the common (ground) leads. e voltages recorded in the noninverting and inverting leads are relative to the common lead. e dierence (voltage) from the inverting common channel is subtracted from the dierence (voltage) of the noninverting common channel, and only the remainder is amplied. By dierential amplication, common mode noise, which appears similarly at both the noninverting and inverting electrodes, can be reduced. Filtering eliminates noise that is outside the frequency range of the desired response. For click-evoked ABR, the response is in the 100-3000 Hz frequency range. So, selectively eliminating electrical activity below 100 Hz and above 3000 Hz will reduce the background noise. About 2000-4000 repetitions are averaged within a 10-15 ms recording window in ABR measurement. In addition, to
minimize artifact, single data trials are rejected when they contain out-of-bounds potential values at single electrodes. For example, if the electroencephalography (EEG) noise reaches a certain amplitude, for example, 20-40 µV, then the entire sweep is ignored and is not added to the accumulating average. Low artifact rejection level causes a high rejection rate and requires additional test time.
