Former NFL star Chris Johnson was diagnosed with ALS. An expert explains how he is still able to talk using new tech
Once a first round draft pick, Chris Johnson is now largely unable to move and talk. But innovative new technology has allowed him to keep his voice. A Northeastern voice tech expert explains how.
Former National Football League star running back Chris Johnson recently revealed that he has been diagnosed with the neurodegenerative disorder known as ALS, or amyotrophic lateral sclerosis.
The first-round pick in the 2008 NFL draft, Johnson, 39, retired in 2017 and told “Good Morning America” that he felt he was in the prime of his life last year. Now, he can hardly move and has lost his ability to speak.
“Just over a year ago, I was picking up my 7-year-old daughter so she’d make a wish with her birthday cake,” Johnson told “Good Morning America. “Today, I couldn’t do that.”
Northeastern Global News, in your inbox.
Sign up for NGN’s daily newsletter for news, discovery and analysis from around the world.
ALS affects the nerve cells in the brain and spinal cord that allow people to voluntarily control their movements and breathing. Without a cure, those with ALS typically die from respiratory failure within three to five years, according to the National Institutes of Health. The disease, also known as Lou Gehrig’s disease after the American baseball player who was diagnosed with it in the 1930s, has become more well-known due to the 2014 viral Ice Bucket Challenge that raised funds for treatment. The diagnoses of high-profile people such as astrophysicist Stephen Hawking and actor Eric Dane have also helped bring more attention to it.
Despite the rapid degeneration that Johnson has experienced, he was able to recreate his voice using innovative new technology and is still able to “speak.”
Rupal Patel, a professor of communication sciences and disorders at Northeastern University, explained to Northeastern Global News the science behind this tech. Patel founded VocaliD, a startup that pioneered innovative methods of creating personalized synthetic voices. Although the tech is vital for those, like Johnson, who can no longer speak, given how quickly conditions like ALS can develop, Patel said this technology is important for everyone to understand.
“I think this points to the fact that everyone should bank their voice,” Patel said.
This interview has been edited and condensed for clarity.
How did Chris Johnson recreate his voice so that he can keep speaking despite his condition?
Essentially, he must have banked his voice ahead of time, so he recorded sentences before (his) speech got worse. In an ideal situation, you record some sentences where you’re told what to read aloud. If you’re not able to record or the disease has gotten way worse, then you could go and fetch some old recordings or interviews or whatever you have and use that as the input.
What we do these days is we use modern large voice models, which are essentially deep learning models, that then use the speech you have to learn the patterns of how one produces the different consonants and vowels and the melody of one’s speech. It essentially creates your speech signature and then recreates that as an algorithm or mathematical equation.
Now the computer knows all the ways you say these different sounds and the sound combinations so that when you want to say something novel that you haven’t recorded it can then use this mathematical equation to generate the speech. It then gets integrated into the device that you use to communicate.
How important is it for people in cases like Johnson’s to be able to continue using their own voices?
It’s super important because it’s a sense of identity and a sense of who you are. Now, even though much of all that’s left is eye motor control, he’s still being heard as himself to the world. That’s really empowering at a time where you’ve sort of lost control of everything else. It’s not like you’re sounding like a robot. You’re still sounding like who you are.
Editor’s Picks
How is he able to communicate through that device using only his eye movements?
The whole system is called an augmentative and alternative communication system, or an AAC device.
It’s not really reading his brain. On the computer screen that he’s looking at there are symbols or there’s a letter board. There’s a forward-facing camera that’s looking at his eyes and it then triangulates where he’s looking at the screen and if you dwell on the screen long enough, then that’s a selection. My guess is he’s probably using something with symbols because it’s faster. If you try to type out every single word you want to say, it just takes really long.
Essentially, there will be some fast phrases like, “I want water,” but there will also be a way to create phrases and messages that are more on the fly. Things like “I’m itchy” or “I had a dream yesterday,” something you want to share that’s not a quick-fire message.
When you select those messages, the spoken message that’s programmed in behind that is spoken out loud. In order to get his voice to speak that, all that’s really necessary is that you swap out the generic voice that the system came with to this new voice that you’ve built using his previous recordings. Then it integrates fully with the system.
Would you recommend people more generally consider recording and banking their voice?
When I started VocaliD [in 2014], we needed seven hours of audio to create a high-quality synthetic voice. By the time VocaliD was acquired in 2022, we needed five minutes of audio. Now, they need closer to three to five minutes.
It’s so easy to do. For three minutes, that’s just an investment in your future.
