Laryngectomy: Improving speech after surgery
First published on 18 October 2008
Updated on 18 June 2013
What did the project achieve?
Researchers are developing a new way to restore speech after surgery to remove the voice box.
“Some people with throat cancer, injuries to the throat and destructive throat infections have to have their voice box – or larynx – removed,” explains Dr James Gilbert. “After surgery, people lose their ability to speak. This can lead to feelings of isolation and depression. Although there are a number of ways to restore speech, they all have limitations.”
“We are developing a new technique, which reconstructs speech using tiny magnetic implants and sensors that assess how the mouth and tongue move as people try to talk,” continues Dr Gilbert. “The speech that’s created will be similar to each person’s original voice.”
Laboratory trials have shown that the new system can recognise individual words from a vocabulary of 57 different words with an accuracy of over 98 per cent. It can also distinguish accurately between groups of similar words such as mime, mine, mile and mice.
“While many technical challenges remain, our research has demonstrated that this new technique has great potential,” says Dr Gilbert. “Our development work continues, with the ultimate aim of creating a system that improves the lives of large numbers of people.”
This research was completed on 31 March 2012
Surgery to remove the voice box – known as laryngectomy – leaves people unable to speak. Methods to restore speech have limitations. One, for example, gives people a ‘Dalek-like’ voice. Another depends on an artificial valve in the throat, which must be replaced every few months. Researchers are developing an exciting, new computer-based device, which reconstructs someone’s speech by assessing the way their mouth moves. They hope it will improve speech quality and cut the number of interventions required.
What's the problem and who does it affect?
Life without a voice box
Some people with throat cancer, injuries to the throat, and destructive throat infections have to have their voice box – or larynx – surgically removed. After surgery, people have to breathe through a hole – or stoma – in their neck and they lose their ability to speak.
It is extremely important that people learn to talk again as soon as possible after their operation. Speech is vital to our normal social interactions, it helps people adjust to their new way of life after surgery and it may stop them from becoming depressed.
There are three main ways to restore speech, but they all have limitations. Firstly, a silicone valve may be inserted into the throat. This allows fluent, though gruff speech, but the valve has to be replaced every few months, which some people find uncomfortable and distressing, and valves are not suitable for everyone. Secondly, an electronic device held next to the neck can help people speak, but their new voice is normally monotonic and ‘Dalek-like’. The third way involves swallowing air and belching, forming the sound into words. This is difficult to learn, fluent speech is impossible and some people find it embarrassing.
What is the project trying to achieve?
When we speak, our voice box vibrates to create a sound, which is ‘shaped’ into words by movement of the palate, nasal cavity, tongue and lips.
The existing methods used to help people talk again after a laryngectomy focus on finding alternative ways to do the job of the voice box.
The researchers working on this project are trying a totally different approach. They are developing a device that reconstructs speech by interpreting the movement of the mouth, lips and tongue.
The new approach involves positioning a series of tiny magnets in the mouth, implanted into the teeth and soft tissue under local anaesthetic. The magnets create a magnetic field, which changes as the mouth moves during speech. These changes correspond to particular sounds or words.
Sensors incorporated into everyday attire, such as a pair of glasses, necklace or tie, detect the changes in the magnetic field. A computer interprets the changes and generates the appropriate words. Assessing speech before surgery would enable the device to reconstruct each person’s own normal voice.
What are the researchers' credentials?
|Project Leader||Dr J M Gilbert PhD CEng MIET|
|Location||Department of Engineering, University of Hull, in conjunction with the Department of Computer Science, University of Sheffield, and Department of Otolaryngology and Head and Neck Surgery, Hull Royal Infirmary.|
|Grant awarded||18 July 2008|
|Start date||1 April 2009|
|End date||31 March 2012|
|Grant code||AP1164, GN1723|
We do not provide medical advice. If you would like more information about a condition or would like to talk to someone about your health, contact NHS Choices or speak to your GP. Please see our useful links page for some links to health information, organisations we are working with and other useful organisations. We hope you will find these useful. We are not responsible for the content of any of these sites.
The researchers working on this project have many years experience of working with people who have undergone a laryngectomy. They have made real progress towards improving the performance of existing speech restoration devices by, for example, improving the composition of artificial valves used in the throat.
The project team combines the expertise of two university groups who have world-leading expertise in analysing and reconstructing what people with speech difficulties are trying to say. This expertise, along with input from the leading clinicians who are working with the team, should make it possible to bring their groundbreaking, new, speech-reconstruction device a big step closer.
Who stands to benefit from this research and how?
The next generation
The researchers’ ultimate aim is to develop a ‘next-generation,’ speech-reconstruction device for people who have lost their ability to talk after having their voice box removed.
In this project, the researchers are developing a prototype device, which they are testing on themselves. Initially, the vocabulary will be limited and the accuracy of voice reconstruction imperfect. As the research progresses, the team aims to improve performance so the device can generate fluent speech, with the intended emphasis and intonation using each user’s original voice. Another key aim is to ensure the device is acceptable to users.
The researchers hope their new device will help overcome the limitations of existing systems. They believe it may prove beneficial to more people, may give people better quality speech and may avoid the need for the repeated surgical procedures required with artificial valves in the throat. Based on their experiences in their own region of the country, the research team estimates there are approximately 4,000 valve users living in the UK and around 16,500 valve changes at an estimated cost to the NHS of nearly £10 million.
In the longer term, this new approach to speech reconstruction could help others with speech difficulties too, including people with cerebral palsy.