Communication and Voice Banking With MND

Communication and voice banking with MND: the thing you cannot do later

About 80 per cent of people living with MND will have changes to their speech. MND Australia’s own figure, and its 2025 report puts communication difficulty at over 80 per cent as well.

Most of this guide is about practical options. But the single most important point in it is a timing one, so it goes first.

Voice banking has to be done while you can still speak. It is a recording process — you read material aloud so that a synthetic voice can be built from your own. If you wait until speech has deteriorated, the recordings capture the deterioration, and if you wait longer than that you cannot do it at all. There is no retrospective version. A person who banks their voice in the month after diagnosis and never needs it has lost an afternoon. A person who waits to see whether their speech will be affected and then finds it is has lost their voice permanently.

So the instruction, for anybody newly diagnosed with MND, whatever their current speech: do it now. Before the wheelchair conversation, before the NDIS plan, before anything that feels more urgent. It is the one thing on the entire MND to-do list where the window closes quietly and completely.

The rest of this guide covers what happens to speech, what voice banking and message banking actually involve, the range of communication options from pen and paper to eye gaze, how a system gets built and by whom, and what funding applies. Communication in other conditions is covered elsewhere — communication and AAC for the general framework, and aphasia after stroke for a quite different problem, since aphasia is a language disorder and MND speech change is a motor one.

What happens to speech

The term is dysarthria — a motor speech disorder. The language is intact. The person knows exactly what they want to say and can understand everything said to them. What fails is the machinery of producing it.

In MND the pattern is usually mixed dysarthria, combining two types:

Flaccid features, from lower motor neurone weakness: breathy or weak voice, nasal speech because the soft palate does not seal, imprecise consonants, short phrases because there is not enough breath.

Spastic features, from upper motor neurone involvement: strained or strangled voice quality, slow rate, harsh tone, effortful speech.

The progression, broadly. Early on, speech sounds slightly slurred, or tired by the end of the day, or a bit nasal. People notice on the phone before they notice in person. Then rate slows and volume drops. Then listeners start asking for repetition, and the person starts shortening what they say to avoid the effort. Eventually speech becomes unintelligible to all but the most familiar listeners, and finally absent.

Bulbar-onset MND starts here rather than in a limb, which means communication planning is the first priority rather than a later one. Around a quarter to a third of people have bulbar onset, and it is more common in women and in later onset.

Three things are usually preserved, and they are the foundation of everything that follows.

Language and comprehension. Unless there is significant frontotemporal involvement, the person understands normally. This cannot be stressed enough, because people with unintelligible speech get treated as though they are cognitively impaired, spoken about in the third person, and spoken to loudly and slowly. It is the commonest and most dehumanising error in MND care.

Hearing. Unaffected by MND.

Eye movement, for a long time and for most people. Which is why eye-gaze technology works, and why nobody with MND should be treated as unreachable.

Voice banking and message banking

Two different things, often confused, and both worth doing.

Voice banking creates a synthetic voice from recordings of your speech. You record a set script — the number of sentences depends on the service, from a few dozen to several hundred — and software builds a personal synthetic voice. That voice is then loaded onto a speech-generating device, so when you type or select a message later, it comes out sounding like you. Not a perfect impersonation, but recognisably you rather than a generic computer voice.

Modern voice-building needs far less material than it used to, and some services can work with surprisingly little or with older recordings. If speech has already changed, ask anyway — it may still be possible, and some systems can enhance degraded recordings.

Message banking records your actual voice saying specific things, kept as audio files and played back at the touch of a button. It captures what a synthetic voice cannot: intonation, laughter, the way you say your children’s names, the phrases that are specifically yours. A swear word delivered properly. A joke you always make. “I love you,” in your own voice, with your own emphasis.

Families consistently report that message banking matters more to them than voice banking does, and they rarely know about it in time.

What to bank, if you do nothing else:

• Names, said the way you say them
• “I love you”
• “Yes,” “no,” “I’m fine,” “thank you,” “please,” “sorry”
• “I’m in pain,” “I need to be moved,” “I’m too hot,” “I need suction” — the functional ones
• Your laugh
• Phrases that are yours and nobody else’s
• Something for each person who matters
• Your own swearing, which people are shy about recording and then grateful for

How to arrange it. Through a speech pathologist, which is the best route because they will advise on what to record and how. Several state MND associations run voice banking support, and MND Queensland runs a voice banking clinic specifically. MND and Me Foundation has published an Australian guide to voice banking at home. Your state association, through 1800 777 175, will tell you what is available where you are.

Do it in a quiet room, in a few sessions rather than one, at a time of day when speech is at its best — usually morning. Fatigue degrades the recordings.

The range of options

A communication system is rarely one thing. Most people end up with several, used in different situations, and the mix changes as the condition progresses.

Voice amplification. For somebody whose speech is still intelligible but too quiet to be heard. A small amplifier with a headset microphone. Simple, cheap, and often the first useful intervention — it buys months of normal conversation. State MND association loan libraries stock these.

Low-tech, which is underrated. Pen and paper. A whiteboard. An alphabet board. A communication board with common phrases and needs. These are fast, need no charging, work in the bathroom and in the car, and work when the device is being repaired. Everybody should have a low-tech fallback, however good their technology.

Phone and tablet apps. Text-to-speech on a device the person already owns. Good for early needs and highly portable.

Speech-generating devices. Dedicated communication devices with built-in speech output, designed for people with physical limitations. Accessed by touch, switch, head mouse, or eye gaze. These can carry a banked voice and banked messages.

Eye-gaze systems. A camera tracks where the eyes are looking, and the person selects letters or phrases by looking at them and dwelling briefly. This is the technology that keeps communication possible when nothing else does — no speech, no hand function, no head control. It takes practice, it works best with a good seating and positioning setup, and it needs the screen in the right place, which is an occupational therapy and equipment problem as much as a speech pathology one.

Switch access and partner-assisted scanning. Where even eye gaze is unreliable, a single reliable movement — a finger, an eyebrow, a blink — can operate a switch that scans through options. Partner-assisted scanning is the no-technology version: the partner reads out groups of letters or options and the person signals at the right one. It is slow and it works, and every family should know how to do it as a backup.

Yes/no signals, gestures, writing, texting. Agreed early, written down, and used consistently by everybody.

Emergency and identification tools. MND Australia suggests an “I have MND” wallet card or phone wallpaper, and a “My care needs” form for emergencies. For somebody with unintelligible speech arriving at an emergency department, a card explaining that they have MND, that their comprehension is normal, how to communicate with them and who to call is the difference between being treated as a person and being talked over.

Building a system that keeps working

The mistake is treating this as a purchase. It is not; it is a process that has to stay ahead of the progression.

Get a speech pathologist involved early, before there is a problem to solve. They assess, trial options, and — crucially — plan the next stage before the current one fails. A person who is switched to eye gaze while they still have some hand function learns it far more easily than somebody introduced to it in crisis.

Get an occupational therapist involved too. MND Australia’s guidance describes building a communication system with support from both. The OT handles mounting, positioning, seating, screen placement and access method, which determine whether the device is actually usable. A beautifully specified eye-gaze system mounted at the wrong angle is an expensive ornament.

Plan the next step while the current one still works. The practical rule for MND communication: whatever is working now, know what replaces it and have it trialled.

Train everyone. The device is only useful if the people around the person know how to wait for it. A banked message is only useful if a support worker knows where the button is.

And keep the low-tech fallback current. Devices break, batteries die, systems get sent for repair. An alphabet board in the bag means a bad day instead of a silent week.

How to talk to someone whose speech is going

Short, because it matters more than any of the technology.

Give them time. Whatever a device or a weak voice is producing, it is slower than you are used to. Resist filling the silence. Resist finishing sentences. The impulse to help by guessing is the single most frustrating thing a person with dysarthria experiences.

Ask, do not assume. If you did not understand, say so honestly: “I got the first part, can you give me the last word again?” That is far better than nodding, and people can tell.

Yes/no questions when energy is low. Open questions when there is time.

Do not raise your voice. Hearing is fine. Volume does not help and it signals that you think something is wrong with them.

Speak to them, not about them. Not to the carer in the room. Not in the third person while they are present. This is the complaint people with MND make most often about hospital and clinic encounters.

Keep the eyes visible and the screen in position. If you are using eye gaze, do not stand in front of it.

And assume full comprehension and full adult intelligence unless you have specific information to the contrary. Someone who cannot speak and cannot move is still following everything said in the room, including the bits said as though they were not there.

Funding and where to get it

For NDIS participants, communication devices are assistive technology, and speech-generating devices with eye-gaze access sit in the higher-cost categories requiring an AT assessment and a quote. The things that speed this up: an occupational therapy and speech pathology joint assessment, a clear statement of the progression and why the device is needed before it is needed, and asking for AT funded as a flexible category rather than a single named item, so the plan can respond to change.

Ask explicitly for the assessment, trial, training, mounting and maintenance to be funded, not just the device. A device without training and mounting is frequently unused.

For people over 65 on Support at Home, the assistive technology allowance is around $15,000 within a maximum package of about $108,000, and a speech-generating device competes against the wheelchair, the hoist and the bed for that money. This is one of the clearest places where the age boundary bites. See MND over 65.

And regardless of funding system: the state MND association. Loan libraries stock voice amplifiers and text-to-speech tools and lend them free, without any funding approval, for as long as needed. Some run voice banking clinics. This is the fastest route to a working solution and almost nobody uses it first. 1800 777 175.

For the general picture on equipment speed, see equipment and home modifications with MND.

Writing the communication profile

One document worth making early, because it travels with the person and does more than any amount of explaining in the moment.

A one-page communication profile, kept at the front of the support plan, given to every new worker on their first shift, and sent to hospital with the person. What it covers:

How I communicate. Which methods, in which situations, and in what order to try them. “Device first; if the battery is flat, the alphabet board in the side pocket of the chair; if I am in bed, partner-assisted scanning.”

How long to wait. An explicit instruction, because the single most common failure is people filling the silence. Something like: “I will take up to a minute to compose a sentence. Please wait. Do not guess the ending.”

My yes/no signals. Exactly what they are and what they look like. Written down, because an eyebrow raise means nothing to somebody who has not been told.

What I want to be called, and what I do not. Not “love” or “sweetheart” unless that is genuinely welcome.

How you will know I have not understood you, and how you will know I am in pain, too hot, uncomfortable or frightened. For somebody with no speech, this section prevents more harm than any other.

Where my banked messages are, and what is in them.

My positioning requirements for the device — screen height, angle, distance, and the fact that standing in front of it makes it useless.

Five things about my life. My job, my kids’ names, the team I follow, the music I like, where I grew up. This transforms what a stranger can do with somebody in the first ten minutes of a shift, and it is the section people are most glad they wrote.

Who knows me best, and their number.

For somebody with unintelligible speech arriving in an emergency department at 11pm among staff who have never met them, this page is the difference between being understood and being managed. It takes an hour to write and it should be written while writing is still easy.