Virtual Voice: Artificial Larynx

News

Not a member yet? Register for full benefits!

The larynx, or voicebox is the organ responsible for much of the production of soundwaves that become speech. Sound itself is generated in the larynx, and the pitch and volume of the resultant sounds are controlled there. In addition, the force the air is expelled from the lungs, adds to volume. Without a larynx, any form of speech is not possible.

The longer the vocal cords themselves, the deeper the sound produced, as with any string instrument.

After the larynx proper, the vocal tract heads on up to the head, where it opens out into the oral cavity (mouth). The oral cavity is the most important component of the vocal tract because its size and shape can be varied by adjusting the relative positions of the palate, the tongue, the lips, and the teeth.

In other words, the movement of the roof of the mouth, the tongue, the lips, and the teeth all contribute to the final phonetic shape of the sounds coming out.

When the larynx goes wrong

A whole raft of issues can affect the larynx. These can create hoarse speech, pain in the throat, and breathing difficulties. They can also cause temporary or permanent loss of speech. Some of the more pertinent are outlined below:

Acute laryngitis

Acute laryngitis is a sudden inflammation and swelling of the larynx. Excessive shouting or an infection will cause the tissues to swell.

Chronic laryngitis

Chronic laryngitis occurs over time, and the tissues can swell permanently if a great deal of yelling is performed, or as a long-term reaction to cigarette smoke, or pollution.

Presbylarynx

Presbylarynx is an axe-related atrophy of the tissues in the larynx. This reduces the voice to a weak strength and limited use.

Perforated Airway

If the airway is breached below the larynx, as performed in some emergency surgeries when the windpipe is crushed, then air no-longer flows properly over the larynx, and all voice is lost.

Two forms of cancer of the larynx result in it's removal.

Any form of surgery on or about the larynx can result in nerve damage, and thus inoperability.

Vocal cord paresis is weakness of one or both vocal folds

Further Speech Disorders

Disease in the lungs themselves, or in the bronchii, preventing forceful resperation of air sufficient to excite the larynx - weak or nonexistant speech.
Diseases and disorders of the brain: alogia, aphasias, dysarthria, dystonia and speech processing disorders, impaired motor planning, nerve transmission, phonological proccessing or perception of the message results in unintelligable speech.
Difficulty hearing, which leads to difficulty hearing and thus correcting the person's own voice.
Pronounciation issues such as lisps, cleft palate, or nerve damage. Also damage to lips and teeth resulting in poor sound shaping.

Any one of these issues leaves the person either temporarily, or permanently, without a voice to truly call their own.

Creating an Artificial Larynx

To create an artificial larynx, you must model all the variable elements of a larynx system:

Volume of airflow from the lungs
Speed of airflow from the lungs
Position and shape of soft tissue in the larynx
Length ofthe vocal cords
Size of the oral cavity
Shape of the oral cavity
Shape and clefting of the palette
Movement range ofthe palette
Shape of the tongue
Movement range of the tongue
Shape of the teeth
Movement range of the teeth
Shape of the lips
Movement range of the lips

Only when you have all these elements, can an artificial larynx be modelled. It is not really possible to create one without all elements, as a voice is a result of modulations in airflow, and it is the airflow itself you seek to recreate with the artificial larynx. In order to model the airflow correctly, you require all elements which would act on that airflow.

The algorithms created from this arrangement are required far more than visualisation when it comes to creating a virtual model of the larynx. A visual, VR recreation of the larynx and trachea, fully animated, anatomically correct with lifelike textures may look really accurate, but it won't model the airflow accurately, unless each element is also modelled mathematically.

A visual artificial larynx oral cavity is a nice idea, but not if it slows the processing system down sufficiently to distort the actual speech output.

To model every flap of tissue, every muscle, bone, and potential variable in the speech system, is difficult, being computationally expensive. However, we are just now, at the time of writing, entering into the technical capability to perform such.

Lesser Simulations

Rudimentary Visualisation

To give an idea of how close we are getting, even on what is currently considered low end equipment, Animated Speech Corporation, an educational company operating out of San Francisco in the US, produce a complete model of the organic speech production system in visualised segments to help children with speech difficulties.

"Children with autism, hearing impairments, developmental delays and other language problems struggle every day with communication, conversation and reading. Animated Speech's software is designed to give speech pathologists, educators and parents who work with these children an important new tool to build comprehension and vocabulary skills."

Whilst this system is purely visual and does not accurately model actual airflow, irt does show an easily visible hint at what a more powerful model running on an embedded system is capable of.

Accurate Tissue Recreation

The Dextroscope is a holographic imaging system, dedicated to the accurate recreation of bodily tissues, at functional precision. The current workstations are five feet tall, a little too bulky to be used as an artificial larynx on the move, and a tad oversized for the avarage home setup, to give a 'natural' voice over the phone or online. However, it is also a considerable degree of overkill for the requirements - actually too powerful.

However, its existence, now, does domonstrate with crystal clarity how technically ready we are, to create an artificial larynx system.

Dextroscope Imaging of functional, virtual human tissue

In Closing

An artificial larynx - and entire vocal system has the potential to give voice to the voiceless, and a much more natural and maintainable voice than esophageal speech methods used to produce tinny, and artificial voices for people without a functional larynx.

It opens up the possibility of normal interaction with others, whether you have had your larynx removed, you have a really bad cold, or your own voice is just way too high to make a convincing mountain troll.