Tracheoesophageal voice restoration after laryngectomy is possible with a variety of neopharyngeal reconstructions. We have used the tubed radial forearm free flap for neopharyngeal reconstruction since 1991. Six patients have undergone voice restoration with the Blom-Singer prosthesis and were available for quantitative and qualitative speech analysis. These patients were compared to five laryngectomy patients with standard pharyngeal closures and similar voice restorations. The free flap patients produced similar loudness levels compared to the standards with soft speech (52.06 dB and 47.19 dB, respectively) and loud speech (62.66 dB and 60.91 dB, respectively). The free flap patients demonstrated adequate intelligibility, with fundamental frequencies comparable to standards (124.82 Hz and 135.66 Hz, respectively), although with increased jitter (5.00% versus 1.96%). No differences were statistically significant, but evaluation by trained and naive listeners demonstrated significant differences in voice quality. This quantitative and qualitative and qualitative analysis of tracheoesophageal speech after radial forearm free flap reconstruction of the neopharynx demonstrates that acceptable voice can be achieved, but with limitations.

No abstract

The purpose of this study was to determine if measures of speaking fundamental frequency and its perturbation could be useful in differentiating talkers with no known vocal pathology and talkers with cancer of the larynx. Ten male subjects, five with a diagnosed malignancy of the larynx and five with normal voice, produced speech samples from which five voice production measures were obtained: the average speaking fundamental frequency (SFF), SFF variability during the reading of a sentence, the f 0 of a sustained vowel and a percent and magnitude jitter value. The perturbation factors, both directional and magnitudinal, during sustained vowels were found to be significant in discriminating normal talkers from those with laryngeal cancer. The speaking funda- mental frequency and its variability during the reading of a sentence improved the dis- criminant function.

Although tracheoesophageal voice restoration is accepted after reconstruction of the neopharynx with the pectoralis major myocutaneous flap, the character of such voice is not well described. Six patients reconstructed with the pectoralis major flap after laryngopharyngectomy underwent successful voice restoration with the Blom-Singer prosthesis. Voice was evaluated by a standardized protocol and compared with voices of control subjects treated with total laryngectomy and similar voice restoration. The patients with pectoralis major flaps produced similar intensity levels for soft voice (53.7 dB vs. 55.6 dB) and loud voice (61.3 dB vs. 65.3 dB) when compared with controls (p > 0.05). No significant differences (p > 0.05) were noted for fundamental frequency (F0) between patients with pectoralis major flaps and controls for soft (62.3 Hz vs. 85.4 Hz) and loud (109.8 Hz vs. 133.8 Hz) voice. Jitter was also comparable. Trained and naive listeners completed qualitative analyses for 10 parameters and judged that control patients had significantly better voice for most parameters. This finding demonstrates that dependable voice is attainable after pectoralis major flap reconstruction of the neopharynx. Although this voice does not differ significantly from voice after standard laryngectomy for acoustic parameters, perceptual analysis does reveal significant differences.

The purpose of this research was the development of a methodology for valid reliable measurement of laryngeal jitter during sustained phonation. We undertook both to develop and evaluate an appropriate procedure and to provide preliminary data on the degree of jitter present in vocal signals of normal larynges. The method utilized an ultrahighspeed photographic technique for measuring cycle-to-cycle variations in acoustic signals. Validation was accomplished by measurement of three different computer-produced (LADIC) synthetic signals of 65 consecutive cycles; each was programmed for predetermined wavelengths. The programmed values were compared with the measured values and extremely high correlations were obtained. Secondly, films and recordings were made while four healthy male subjects (aged 21–37) phonated the sustained vowel /a/ at each of four frequencies which were based on their phonational ranges. Although our measurements on human subjects failed to support the contention of Weber's law that an increase in jitter characteristically accompanies increase in fundamental frequency, we can conclude that (1) the methodology developed provides a valid means of measuring variations in quasiperiodic acoustical signals and (2) that we have been able to provide some tentative data concerning the amount of jitter that can be expected in the sustained phonation of normal male voices.