This study compared acoustic and electroglottographic (EGG) jitter from [a] vowels of 103 dysphonic speakers. The EGG recordings were chosen according to their intensity, signal-to-noise ratio, and percentage of unvoiced intervals, while acoustic signals were selected based on voicing detection and the reliability of jitter extraction. The agreement between jitter measures was expressed numerically as a normalized difference. In 63.1% (65/103) of the cases the differences fell within +/-22.5%. Positive differences above +22.5% were associated with increased acoustic jitter and occurred in 12.6% (13/103) of the speakers. These were, typically, cases of small nodular lesions without problems in the posterior larynx. On the other hand, substantial rises in EGG jitter leading to differences below -22.5% took place in 24.3% (25/103) of the speakers and were related to hyperfunctional voices, creaky-like voices, small laryngeal asymmetries affecting the arytenoids, or small-to-moderate glottal chinks. A clinically relevant outcome of the study was the possibility of detecting gentle laryngeal asymmetries among cases of large unilateral increase in EGG jitter. These asymmetries can be linked with vocal problems that are often overlooked in endoscopic examinations.

Jitter, or the amount of cycle-to-cycle variation in the fundamental frequency, is a characteristic of the vocal folds' vibration that may affect electroglottographic (EGG) and acoustic signals in similar ways, because the translaryngeal conductance and the oral pressure are modulated by the same physiological mechanism. Despite the apparent simplicity in jitter computation and the relative facility in recording and analyzing EGG signals, only a few studies comparing EGG and acoustic jitter have been reported. This can be attributed to the remarkable sensitivity of measures of acoustic jitter to such features as the type of sound being analyzed, the equipment used for data acquisition, or the algorithms used to identify glottal cycle boundaries. To assess the agreement achievable by measures from these signals in pathological voices, acoustic and EGG waveforms of sustained vowels (/i/, /a/, /u/) produced by 15 dysphonic patients were analyzed by jitter extraction methods on the basis of peaks, zero crossings, and a waveform matching technique. The agreement, expressed as normalized absolute differences between acoustic- and EGG-derived jitter, was poorer for /i/ and /u/ than for /a/ vowels. For /a/ vowels, a method of acoustic jitter estimation is also proposed that combines peaks and zero crossings and resulted in increased consistency with the zero crossing-based EGG measures (mean normalized absolute difference: 10.95%, SD: 6.44%; range: 23.81%). The proposed method, which has a built-in-test intended to reject unreliable estimates, may provide more confidence in acoustic measures in dysphonic speakers and lead to a better understanding of the relationships between acoustic and EGG signals.

No abstract

No abstract

No abstract

No abstract

No abstract

This paper reports an investigation of the influence of electroglottographic ͑EGG͒ baseline fluctuation on the precision of F 0 estimation based on zero crossings. A passage read by 15 subjects suffering from various laryngeal pathologies was filtered with zero phase shift by several bandpass filters with a flat frequency response between 60 and 5000 Hz. A simple F 0 extraction algorithm processed the filtered EGG signals providing good results even with data from severely affected larynxes. In general, all investigated errors, including a measure of jitter, reduced significantly when the attenuation in the rejection bands varied from 0 to 40 dB, with slight but consistent improvements thereafter. The best results, obtained with 80 dB of attenuation, indicated that unvoiced ͑or silent͒-to-voiced errors and voiced-to-unvoiced ͑or silent͒ errors were under 1.17% and 2.26%, respectively, and at least 91.8% of the estimated F 0 values had a precision better than 0.103%, considering an estimated error of 0.058% in the reference F 0 contours; the incidence of gross errors due to F 0 doubling or halving was under 0.44%. These results qualify the method as a very precise and robust F 0 estimation algorithm for pathological voices.