Augmented Control of a Hands-Free Electrolarynx

During voiced speech, the larynx acts as the sound source, providing a quasi-periodic excitation of the vocal tract. Following a total laryngectomy, some people speak using an electrolarynx which employs an electromechanical actuator to perform the excitatory function of the absent larynx. Drawbacks of conventional electrolarynx designs include the monotonic sound emitted, the need for a free-hand to operate the device, and the difficulty experienced by many laryngectomees in adapting to its use. One improvement to the electrolarynx, which clinicians and users frequently suggest, is the provision of a convenient hands-free control facility. This would allow more natural use of body language, as well as enabling the user to perform other manual tasks whilst speaking. An additional enhancement is the redevelopment of the devices' actuator to remove the self-radiated buzzing sound which is emitted during operation. The overall goal of this research is to investigate the implementation of a more natural sounding, hands-free electrolarynx, with an activation method which uses accelerometers that are placed on the chest and abdomen to detect torso based breathing prior to the initiation of speech. Results illustrate from intelligibility tests carried out that the speech produced using our novel actuator is substantially more intelligible to all the listeners than that produced using the conventional electrolarynx. Also, the abrupt change in breathing pattern which occurs at the onset of speech is clearly observable in signals recorded by the accelerometers, facilitating a possible automatic activation of an electro-larynx which will be investigated further in user trials.