Choice of test stimulus matters for pitch matching performance: Comparison between pure tone and narrow band noise

Comparing pure tone and narrow band noise to measure tonal tinnitus pitch-match frequency

Tinnitus assessment is a precursor for individualized treatment and outcome measurement. In the recent years, several studies proposed two-alternative forced choice (2AFC) recursive matching as a method to determine tinnitus pitch-match frequency in a standardized reliable manner. Currently, pure tones are used as comparison stimuli to assess pitch-match frequency. In this study, we investigated the psychometric quality of the method comparing different sound types. We measured 20 chronic tinnitus patients in 2 runs on 3 days. To assess pitch-match frequency, we used 2AFC recursive matching and compared results between pure tones and narrow band noise (NBN). Test-retest reliability between runs and across sound types was high (α>0.9) and increased across days. Perceived matching difficulty and time to completion decreased over repetitions. Importantly, the difference of matched frequencies (DMF) between runs was significantly less for NBN. When patients matched the spectral bandwidth of a test tone to their tinnitus, consistency was high (α=0.86) and no patient indicated continuously a pure tone. In conclusion, we recommend using NBN sounds in 2AFC recursive matching to assess pitch-match frequency as a standardized reliable method. Such a procedure could be offered as smartphone-based application to monitor tinnitus symptomatology for individualized assessment and treatment outcome.

One Step Closer towards a Reliable Tinnitus Pitch-Match Frequency Determination Using Repetitive Recursive Matching

Introduction: The determination of the tinnitus pitch-match (PM) frequency is not straightforward but an important audiological assessment recommended for clinical and research purposes. We evaluated repetitive recursive matching using an iPod-based matching procedure as a method to estimate a patient’s PM frequency without audiometric equipment. Methods: One hundred and seventeen patients with chronic tonal tinnitus (uni- and bilateral tinnitus) measured their tinnitus in 10 sessions using a self-administered automated iPod-based procedure comprising a recursive 2 interval forced-choice test. Results: Mean SD of the PM frequency of all participants across sessions was 0.41 octaves. The internal consistency measured by Cronbach’s α was very high (0.8–>0.95). As an example, 7 PMs obtained excellent internal consistency (α = 0.93). The exclusion of the first and/or second session led to more definite PMs with a decreased SD. Outliers were identified by PMs departing 2 SDs (i.e., 0.94 octaves) from the mean variability (n = 5). Conclusion: Repetitive recursive matching together with recommendations for the exclusion of initial and redundant sessions as well as outlier identification and treatment can enable a reliable estimation of the PM frequency.