ABRpresto: An algorithm for automatic thresholding of the Auditory Brainstem Response using resampled...

Area of Science:

• Neuroscience
• Audiology
• Biomedical Engineering

Background:

• Auditory Brainstem Response (ABR) is crucial for assessing cochlear health in research and clinics.
• Estimating ABR threshold, the minimum sound level evoking a response, is challenging due to waveform shifts and low signal-to-noise ratio.
• Current standard practice relies on subjective visual evaluation of ABR waveforms.

Purpose of the Study:

• To develop and validate an objective algorithm for precise ABR threshold estimation.
• To improve upon existing methods for ABR threshold determination in preclinical studies.

Main Methods:

• Developed a novel algorithm using cross-correlation of two independent response averages for each stimulus level.
• Responses were split into two groups, median waveforms calculated, and normalized cross-correlation computed.
• Repeated cross-correlation 500 times, analyzed distributions, and fitted data with sigmoid or power law functions to estimate thresholds.

Main Results:

• The algorithm achieved 92% accuracy within ±10 dB of human-rated thresholds on extensive mouse data.
• Demonstrated superior performance compared to several previously published algorithms on the same dataset.
• Successfully replaced manual ABR threshold estimation in preclinical studies, enhancing speed and objectivity.

Conclusions:

• The developed cross-correlation algorithm provides a robust, accurate, and objective method for ABR threshold estimation.
• This algorithmic approach significantly improves efficiency and reliability in auditory research.
• The findings support the adoption of this algorithm for standardized preclinical auditory assessments.