Influence of F0 and Sequence Length of Audio and Electroglottographic Signals on Perturbation Measures for Voice Assessment

Objective. Within the functional assessment of voice disorders, an objective analysis of measured parameters from audio, electroglottographic (EGG), or visual signals is desired. In a typical clinical situation, reliable objective analysis is not always possible due to missing standardization and unknown stability of the clinical parameters. The aim of this study was to investigate the robustness/stability of measured clinical parameters of the audio and EGG signals in a typical clinical setting to ensure a reliable objective analysis. In particular, the influence of F-0 and of the sequence length on several definitions of jitter and shimmer will be analyzed. Patients and Methods. Seventy-four young healthy women produced a sustained vowel /a/ and an upward triad with abrupt changeovers. Different sequence lengths (100, 150, 500, and 1000 ms) of sustained phonation and triads (100 and 150 ms) were extracted from the audio and EGG signals. In total, six variations of jitter and four variations of shimmer parameters were analyzed. Results. Jitter(%), Jitter(11p), and Jitter(PPQ) of the audio signal as well as Jitter(mean), Shimmer, and Shimmer(11p) of the EGG signal are unaffected by both sequence length and F-0. Conclusions. Influence of F-0 and sequence length on several perturbation measures of the audio and EGG signals was identified. For an objective clinical voice assessment, unaffected definitions of jitter and shimmer should be preferred and applied to enable comparability between different recordings, examinations, and studies.