Sound spectrogram analysis in patients receiving Kawahara’s surgical voice restoration for advanced carcinoma of the hypopharynx and cervical esophagus

Abstract

Loss of voice after total laryngectomy is a major concern that has prompted much effort to develop methods of surgical voice restoration. In 1992, Kawahara described a tracheocolic shunt for voice restoration using a revascularized ileocolic graft. With this method, aspiration through the phonatory shunt is prevented by the ileocolic valve (Bauhin's valve), while vocalization is allowed. The purpose of this study was to use sound spectrogram analysis to evaluate voices surgically restored in that way. Between 2002 and 2005, 10 consecutive patients underwent laryngopharyngoesophagectomy with Kawahara's surgical voice restoration for advanced carcinoma of the hypopharynx and/or cervical esophagus at Akita University Hospital, Japan. We then used sound spectrography to analyze and compare the voices of patients receiving Kawahara's voice restoration, against healthy volunteers and patients who underwent the same surgery without voice restoration and spoke using an electronic larynx. We also evaluated the intelligibility of conversation and performed a listening test. The sound spectrograms showed that when produced by the electronic larynx, consonant sounds and voice frequencies below 300 Hz were not clearly recognized. By contrast, in patients who received Kawahara's surgical voice restoration, consonant and vowel sounds at frequencies above and below 300 Hz were clearly recognized. Although conversation was intelligible with both Kawahara's surgical voice restoration and the electronic device, listeners judged the voice produced by the former to be superior. Thus, the voice produced by Kawahara's surgical voice restoration is superior to that produced by the electronic larynx.