Abstract
The auditory evoked slow vertex response (SVR), or so called auditory cortical response, has been used as a method of objective audiometry. The processing of SVR has been done mostly by conventional averaging so far, but there are some unsolved problems, such as the false positive response and the unreliability of response especially with infants under sleeping states. We have, under the circumstances, designed a series of computerized procedures using the PDP 11/60 computer in order to obtain reliable SVR.
Methods:
(1) Analysis of average power spectrum of EEG, with or without acoustic stimuli.
(2) A non-recursive digital bandpass filter using the fast Fourier transform has been designed in order to detect SVR.
(3) A recursive digital bandpass filter (Butterworth) using Z-transform has been designed in order to analyze the alpha-attenuation with acoustic stimuli.
Results:
(1) Averaged power spectrum had its peak at around 3 to 5Hz, when the acoustic stimuli were clearly perceived. On the other hand, it had its peak at around 9 to 11Hz, when the acoustic stimuli were not applied to the subjects.
(2) The most effective pass-band for detecting SVR using the non-recursive digital filter was 3 to 9Hz.
(3) The detectability of the digital filtered SVR in adults and infants was improved in comparison with the non-filtered SVR.
(4) In recursive digital filtering, the alpha-attenuation was observed more clearly with subjects given acoustic stimuli than with nonstimulated subjects.
