Abstract :
Nearly two million non-speaking people today rely on speech output devices to communicate needs and desires. Patients who have undergone laryngectomy have to use one of three speech rehabilitation systems: artificial larynx, esophageal voice, and tracheoesophageal voice. There are major shortcomings associated with these options, including difficulty with use, complications, low speech intelligibility, etc. The paper presents results of a study in feasibility of developing a prosthesis that will enhance speech quality and intelligibility through integration of advances from smart materials, thin film electromechanical systems, and neural-based data processing. Laboratory experiments and resulting speech enhancements are presented and discussed.
Keywords :
neural nets; prosthetics; speech intelligibility; enhanced voice intelligibility; laryngectomees; laryngectomy; neural-based data processing; neural-based prosthesis; nonspeaking people; prosthesis; smart materials; speech intelligibility; speech output devices; speech quality; speech rehabilitation systems; thin film electromechanical systems; Esophagus; Larynx; Muscles; Neck; Neural prosthesis; Prosthetics; Safety; Speech enhancement; Speech processing; Surgery;
