Yi Shen

Starts:
10:30 am on Friday, November 9, 2018
Ends:
12:00 pm on Friday, November 9, 2018
Psychometrics, Machine-Learning, and Clinical Assessments Maximizing limited clinical time to gain as much information as possible on a patient’s auditory profile is essential for behavioral clinical tests in audiology. With modern Bayesian adaptive estimation techniques, significant improvement in data collection time can be gained while maintaining high reliability. In some cases, the amount of time saving could be an order of magnitude, reducing the testing time from hours to minutes. In this presentation, I will report a series of Bayesian adaptive tests recently developed and validated in my laboratory. These tests went beyond hearing threshold and focused on suprathreshold capabilities, including auditory spectral and temporal resolution, equal-loudness level contour, temporal masking release for speech recognition, and spectral relative weights for speech perception. These early efforts have promised a new era in behavioral hearing assessments, in which sophisticated auditory models and signal-processing algorithms can be fitted to individual patients within manageable amount of time. Applied Hearing Science Lab, Indiana University. http://www.indiana.edu/~ahslab/site/index.html