Voiced Speech from Whispers for Post-Laryngectomised Patients

Patients who su_er larynx and voice box de_ciencies are typically unable to speak anything more than hoarse whispers without the aid of voice prostheses or rehabilitation techniques such as oe- sophageal speech. Speech therapists and researchers working in this _eld have, for many years, pursued the goal of rehabilitation of such patients so as to re- turn to them the ability to speak in a natural sound- ing voice. Typically due to removal of, or damage to, the voice box in a surgical operation such as la- ryngectomy, the pitch generation mechanism within these patients voice production systems, is lacking. Without a source of excitation for voiced speech, only hoarse, whisper like and sometimes not easily percep- tible sounds can be produced. This speech is obvi- ously di_erent to that from normal speakers, and will have lost many of the distinctive characteristics of the original speech. However, these patients typically re- tain the ability to whisper in a similar way to normal speakers. This paper aims to present an engineering approach to providing laryngectomy patients the capacity to regain their ability to speak with a more natural voice, and incidentally to allow them to conveniently use a mobile telephone for communication. The non-invasive and non-surgical techniques dis- cussed use auditory information coupled with signal analysis, formant insertion and smoothing and spec- trum enhancement within the reconstruction process. With these techniques, natural sounding speech is ob- tained from spoken whisper-speech building upon an analysis-by-synthesis system for vocal reconstruction with a modi_ed CELP codec.