gms | German Medical Science

Objective: Auditory steady state responses (ASSR) may offer an alternative to brainstem auditory evoked potentials for continuous monitoring of the auditory during surgical procedures in the cerebello-pontine angle. After demonstrating the basic viability of recording ASSR under total intravenous anesthesia (TIVA), we evaluated the influence of technical and biological noise on ASSR detectability and amplitude estimation error.

Method: Simulated ASSR in real noise recorded during CPA surgery were constructed with known amplitude and modulation frequency (80, 90, 110Hz). Influence of four factors on detection of true and false positive responses were tested: ASSR-amplitude, modulation frequency, number of averaging sweeps and detection threshold. ASSR were detected using F-test statistics frequently applied in ASSR studies.

Results: Detection of ASSR was facilitated by high amplitude, increasing number of sweeps and a liberal detection threshold. High amplitude ASSR, which occur using high intensities, e.g. for monitoring applications, could be detected in 32-45% when only 16s of data were used. Percentage increased to 80-90% using 112s of data. Low amplitude ASSR, occurring using intensities slightly above the hearing threshold, could be detected in 0.8-25% even when using the maximum amount of data tested, i.e. 4 minutes. False positive rates ranged between 0.3-10.3% with stricter detection thresholds resulting in lower false positive rates. Number of averaging sweeps did not influence false positives. Amplitude estimation errors varied between -61% to +39% and were improved by increased number of averages but not by different detection thresholds. Strongest influence on all parameters however was the choice of modulation frequency. Using 80Hz yielded the lowest detection rate, doubled false positives and had the largest amplitude errors. Optimal modulation frequency was 110Hz.

Conclusions: Choice of stimulation and analysis parameters heavily influence detection and false positive rate, as well as amplitude estimation error. Optimal parameters (110Hz modulation, high amplitude) enabled detection of ASSR after only 16s in 45%. Due to specific noise characteristics, modulation frequency has a critical impact, which is currently not sufficiently recognized in ASSR studies.