Distortion product otoacoustic emission delay measurement in human ears.

The measurement of distortion product otoacoustic emission (DPOAE) delay using the phase gradient method is discussed. A short f2 sweep paradigm is described. Its merits are discussed and comparisons made with other methods. Using sweeps of less than 4%, consistent and reliable measurements were obtained from 12 normal ears, except for a few isolated measurements. Investigation of these apparent anomalies indicated that the short f2 sweep method was sensitive to the localized effects of narrow distortion product intensity notches and spontaneous emissions. It is possible that middle ear pressure instability or some local change in the cochlea sometimes compounded the situation. The intersubject range (63% of the mean) of DPOAE delay values was greater than the range of values observed due to measurement variability (5.6% of the mean) or intrasubject variability (8.5% of the mean) demonstrating that the method can detect real differences between ears. However, comparison with previous work shows that caution is needed in interpreting these data, as all DPOAE phase methods provide only partial latency values. The physical interpretation depends on both the stimulus sweep used and the origin of DPOAES. Our data shows that f1 sweep derived latency is lower than f2 sweep data by more than 20%. Caution also needs to be exercised when comparing cochlear traveling wave delay, as inferred from electrophysiological data, with that obtained from DPOAE delay measurements.