Abstract
An important cochlear function is to separate out the frequency components in complex sounds. During transtympanic electrocochleography, objective measures of such frequency selectivity were obtained using two (simultaneous) tone on tone suppression paradigms: (a) iso-suppression curves (AP tuning curves): the frequency/intensity combinations to suppress, by a criterion percentage, the AP.to a near-threshold tone pip; and (b) iso-intensity curves: the reduction of an AP to a near-threshold tone pip, by a constant intensity suppressing tone as a function of frequency. Both methods yield 10-dB bandwidth measures related to cochlear frequency selectivity. The latter test, taking only 5-10 min, can be used routinely in clinical electrocochleography. The relationship between 10-dB bandwidth values and audiometric thresholds can be summarized as follows:
In normal ears (threshold elevations of less than 20 dB at the test frequency and ears with no otological problem other than small high-frequency threshold elevation), 10-dB bandwidths ranged from 0.25 to 0.55 octaves at 2 kHz, 0.17 to 0.39 octaves at 4 kHz, and 0.22 to 0.39 octaves at 8 kHz. At 4 kHz the mean value was 0.26 octaves, with 95% confidence for this mean between 0.21 and 0.30 octaves (small sample f distribution, V = 13).
In cases of cochlear hearing loss, threshold elevations of more than 40 dB were accompanied by abnormally wide 10-dB bandwidths (well outside the normal ranges defined above). These findings confirm that a deterioration in cochlear frequency selectivity occurs in pathological conditions of the cochlea.
In conductive hearing loss and in two cases of confi rmed retrocochlear lesions, normal 10-dB bandwidths were found despite audiometric threshold elevations in excess of 40 dB.
In two patients with (early) 's disease, and having low frequency threshold elevations but relatively normal thresholds at 4 kHz, 10-dB bandwidths were found to be significantly broader than those of the normal population.