Imagine a thin, flexible necklace that detects the slightest vibrations in the throat and transforms them into coherent words — restoring speech for patients with dysarthria. This is not science fiction: a new study introducing the “Smart Neck System” could become a new hope for restoring communication abilities in people who have lost their speech due to stroke or other neurological disorders.
Speech impairment is one of the most common complications after a stroke, affecting two out of every three patients. Some cases are temporary, but many patients must adapt to a difficult reality and relearn how to speak. For such patients, various alternative communication tools have long been used — such as eye-tracking screens or implantable neuromodulators. However, screens can be cumbersome, and surgical implantation carries significant risks.
In recent years, scientists have begun developing so-called “silent speech interfaces” — devices that detect what a person wants to say using only muscle movements or tiny skin vibrations, without sound. Unfortunately, most of these technologies can only recognize individual words or require pauses between phrases, making real-time communication difficult.
A new study published in the latest arXiv paper directly addresses this challenge. A team of researchers from the London Medical College and Peking University has developed an AI-powered wearable system that decodes micro-signals from throat sensors in real time and converts them into natural, emotionally expressive speech.
To discuss this topic in more detail, we spoke with speech therapist Eka Chavchavadze.
What types of speech disorders do you encounter in patients after a stroke? How do aphasia and dysarthria differ, and what are the treatment approaches for each?
As you know, speech is a complex neuromuscular process that involves the coordinated function of breathing, phonation, resonance, articulation, and prosody. Dysfunction in any of these subsystems can lead to issues with voice, speech naturalness, intelligibility, and effective communication.
After a stroke, patients may develop speech disorders such as aphasia, dysarthria, or apraxia.
Aphasia results from cortical damage in the brain areas responsible for speech formation and comprehension. It disrupts the linguistic structure of language — articulation of words, sentence construction, synthesis and analysis of phonemes, and understanding of directed speech. Depending on type and severity, all linguistic components — semantics, grammar, phonology, syntax, and morphology — can be affected, impacting both spoken and written language.
Dysarthria results from damage to neuromotor pathways or peripheral nerves that control speech. Patients retain cognitive and linguistic skills but struggle to perform and control articulatory movements with the necessary speed, strength, accuracy, tone, and duration. Their speech may be difficult to understand, though they can fully comprehend and analyze both spoken and written language.
Apraxia of speech is a motor planning disorder where patients have difficulty initiating and coordinating speech movements. They struggle to start speech or move smoothly between sounds, often making inconsistent errors depending on word complexity. Pure speech apraxia is rare and usually co-occurs with Broca’s aphasia.
What assistive tools or digital technologies are currently available for speech therapists treating dysarthria? How easily can patients use them?
Today, high-tech tools are available — for instance, smartphone applications that allow patients to control speech tempo, volume, and rhythm. Therapists also use speech synthesizers, digital recorders, speech generators, and virtual-articulatory therapy platforms.
Recently, AI technologies have become increasingly integrated into speech therapy, offering personalized and adaptive rehabilitation. AI enables automatic speech analysis, objective evaluation of dysarthric speech characteristics, progress tracking, and dynamic adjustment of therapy programs.
New neural models can reconstruct and synthesize a patient’s pre-dysarthric voice, helping restore their personal vocal identity and natural sound.
Additionally, virtual reality–based speech practice provides a motivating, immersive environment, allowing patients to test their communication skills in simulated real-life settings such as cafés, pharmacies, clinics, supermarkets, or offices. Studies confirm that such therapy can significantly speed up the recovery of natural speech.
Following COVID-19, tele-rehabilitation and teletherapy have gained importance, allowing therapists to assess speech parameters and progress remotely.
Another emerging field is neurospeech prosthetics, including BCI (brain–computer interface) technologies that convert neural activity into speech or text — a revolutionary direction in communication recovery.
All these methods and technologies, once properly taught, can be successfully used by patients even with severe forms of dysarthria.
In recent years, innovative technologies have emerged in speech rehabilitation — from implantable neural interfaces to non-invasive wearable devices like the “Smart Necklace.” How would you assess their potential for speech recovery in dysarthric patients after a stroke?
In the near future, the “Smart Necklace” could become one of the most effective aids for many patients with dysarthria — both due to its ease of use and its ability to not only stimulate standard articulatory movements but also adjust the emotional tone of speech, something previous devices could not achieve.
I truly hope that the “Smart Necklace” will soon become available in our country as well.
Do you think the “Smart Neck System” will be effective for all forms of dysarthria?
It’s too early to make definitive predictions — time will tell. However, based on my observations, it is likely to be especially effective in mixed and ataxic forms of dysarthria, where speech tends to be slow, strained, and nasal, with disrupted prosody. Positive results are also expected in hyperkinetic and hypokinetic forms.
In contrast, in spastic and flaccid forms — caused by facial or hypoglossal nerve paresis and often accompanied by dysphagia or chewing difficulties — the device may be less effective.
What psychological challenges do patients face after losing their ability to speak due to a stroke?
The loss of speech ability after a stroke is deeply distressing. Communication difficulties often lead to anxiety and lowered self-esteem, while the inability to express feelings, desires, and needs can result in irritability, depression, and sometimes refusal of therapy.
Therefore, early diagnosis of depression and emotional support from family, friends, and caregivers are critically important — along with timely involvement of psychiatrists, neuropsychologists, and psychotherapists when necessary.
The study’s results are impressive: the “Smart Neck System” accurately recognized nearly all words and sentences spoken by stroke patients with speech disorders — achieving over 95% accuracy for individual words, and only minor deviations for full sentences.
Researchers also found that the device maintained effectiveness over time — requiring only slight calibration for each patient’s unique muscle patterns, after which speech precision was restored.
This technology stands out not only for its accuracy but also for its ability to detect emotion and enable natural communication — factors that are crucial for real human dialogue.
Though still in the research phase, this innovation offers real hope that in the future, recovering speech after a stroke or neurological injury may no longer depend solely on long, intensive therapy sessions.
