Abstract
PURPOSE: Sound pressure recordings at different positions at the head such as ear canal, nasal cavity, or on the skin are effective tools to verify, fit, or follow up the output of bone conduction devices (BCDs) intra- and post-operatively. Here, we investigated the possibility of using a surface microphone (SM) as a non-invasive alternative to laser Doppler vibrometry (LDV) measuring cochlear promontory (CP) vibrations in human heads. METHODS: A percutaneous BCD (Ponto system) was implanted at the standard position in five human (four males/one female) cadaver heads (ten ears). CP vibration was measured using LDV in response to the BCD stimulation. Simultaneously, the sound pressure level (SPL) emitted by the skin was measured by the SM attached to the forehead of the specimens. A linear regression model estimated the vibration amplitudes based on the measured SPL. RESULTS: A frequency-independent linear regression between recorded SM SPL and CP velocity showed a significant correlation (slope = 1.012; r(2) = 0.535, p < 0.001). An enforced fixed slope (constant = 166.7 dB) of one resulted in a mean absolute error of MAE = 7.7 ± 2.7 dB across frequencies. Although the initial linear model with a fixed slope of one showed frequency-dependent deviations, applying a frequency-specific correction significantly improved the prediction accuracy (r(2) = 0.557, MAE = 6.5 ± 1.5 dB). CONCLUSION: Microphone-based recording of acoustic surface emissions offers a non-invasive alternative to LDV to assess BCD output.