Application of Bayesian Active Learning to the Estimation of Auditory Filter Shapes Using the Notched-Noise Method.
Josef Schlittenlacher, Richard E Turner, Brian C J Moore
PMID 33073723WHAT IT FOUND
A new Bayesian active learning test estimated auditory filter shapes in hearing-impaired adults in under 30 minutes.
It matched the accuracy of standard adaptive procedures across a wide frequency range, offering a faster way to measure individual hearing characteristics.
Key findings
01The Bayesian active learning method produced reliable and valid auditory filter shape estimates that were consistent with standard adaptive procedures.
02The test could be completed in less than 30 minutes while providing estimates across a three-octave frequency range.
03The method was compared against a standard 2-up/1-down procedure, showing similar accuracy per trial but covering a wider range of frequencies.
STILL TO COME
How it was doneWhat they foundWhat it means for SLPs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
The study included only 11 participants, all with hearing impairment, so results may not generalize to normal-hearing individuals. The comparison to the standard method showed the BAL approach was slightly less accurate per trial, though it covered a wider frequency range. The test duration of 48 to 61 minutes is still substantial for clinical practice, although the authors propose a shorter 30-minute version based on simulation data. The method estimates filter shapes but does not directly measure the parameters of the underlying auditory filters during the test, which might have allowed for even faster estimation.
Declared interests
Funding was provided by the Engineering and Physical Sciences Research Council and the National Institute for Health Research.
The easy way to misread this
Do not assume this test is currently available for clinical use. It is a research protocol that demonstrated feasibility and accuracy in a small group of 11 participants. It is not yet a standard diagnostic tool or a replacement for current hearing aid fitting methods.