Objectives: The electrically evoked compound action potential is one of the audiologic tools which is useful for early mapping. The errors must be considered at automated analysis. Materials and Methods: We analyzed all waveforms and compared manual results to the automated waveforms: Fifteen cochlear implant users were enrolled in this study. Differences in the thresholds of manual neural response telemetry (NRT) and automated NRT were analyzed by repeated measures ANOVA test, and to analyze the statistical difference in the thresholds at each electrode, a paired t-test was used. The ANOVA test was used to analyze differences in the threshold between the basal turn, middle and apical turn. Results: There was no statistical difference between the threshold of manual NRT and automated NRT (p>0.05). Each electrode was compared, a significant difference in threshold was studied, except two electrodes. Thresholds of approximately 2 current level (CL)similar to 12 CL lower threshold of manual NRT was confirmed at each electrode (p<0.05). The mean differences in the threshold between automated NRT and manual NRT was approximately 4 CL at the apical turn, 8 CL at the middle turn and 6 CL at the basal turn(p>0.05). Conclusion: The thresholds analyzed with visual examination had differences from thresholds analyzed by the automated NRT. We recommend 4 CL in the apical turn, 8 CL in the middle turn and 6 CL in the basal turn electrodes lower than the threshold of automated NRT. These findings might lead to better speech perception and rehabilitation for patients.
