Noise-Augmented Speech Therapy for Parkinson's Disease

Overseen By Gemma Moya-Galé, Ph.D.

Age: 18+

Sex: Any

Trial Phase: Phase < 1

Sponsor: Teachers College, Columbia University

Must be taking: Antiparkinsonian

Disqualifiers: Atypical parkinsonism, Neurosurgery, others

No Placebo Group All trial participants will receive the active study treatment (no placebo)

Approved in 1 Jurisdiction This treatment is already approved in other countries

Compensation: Varies

Number of Visits: 22

Duration: 4 weeks

42 Participants Needed

What You Need to Know Before You Apply

What is the purpose of this trial?

This trial tests a new method to enhance speech therapy for people with Parkinson's disease. By using noise-augmented feedback, which adds background noise to improve speech clarity, through automatic speech recognition, the study aims to determine if it can help individuals speak more clearly and engage more in social activities. Participants will undergo speech treatment sessions at home, guided by a speech therapist, with some sessions monitored online. This trial may suit someone with Parkinson's disease who experiences speech difficulties and has not recently received voice-focused treatment. As an Early Phase 1 trial, it offers the opportunity to explore innovative treatment methods at an initial stage.

Will I have to stop taking my current medications?

The trial does not specify if you need to stop taking your current medications, but it requires that your Parkinson's medication schedule is stable. This likely means you should continue your current medications without changes.

What prior data suggests that this noise-augmented feedback protocol is safe for individuals with Parkinson's disease?

Research shows that noise-enhanced speech recognition technology is being studied to help people with Parkinson’s disease improve their speech. In a previous study, a web-based app called "Understand Me for Life" provided feedback to help users speak more clearly.

The app uses Google Cloud’s speech-to-text technology and was tested with individuals both with and without speech difficulties. Tests included background noise, such as people talking, to determine if the app could still understand and improve speech in noisy environments.

Regarding safety, this treatment involves using a phone and computer for speech exercises. No negative effects are known from using this technology for speech therapy. This trial is in the early stages, focusing on ensuring the technology is safe and effective. Early-stage trials generally aim to confirm safety, so any serious side effects would be unexpected. ^{1 2 3 4 5}

Why are researchers excited about this trial?

Researchers are excited about noise-augmented automatic speech recognition therapy for Parkinson's disease because it offers a novel approach to improving speech clarity. Unlike traditional speech therapy, which relies heavily on repetitive exercises and vocal coaching, this method integrates background noise to enhance the brain's processing of speech. This innovative technique could lead to more effective communication for patients, addressing a critical need in managing Parkinson's disease symptoms.

What evidence suggests that noise-augmented automatic speech recognition is effective for improving speech in Parkinson's disease?

Research has shown that noise-enhanced speech recognition technology might help people with Parkinson's disease speak more clearly. In this trial, participants in the immediate treatment arm will receive speech treatment right after their second pre-treatment assessment, while those in the deferred treatment arm will receive the same treatment one month later. A small study found that a speech app improved how well people with Parkinson's were understood by providing feedback in noisy settings. This method also improved the speech recognition system's ability to understand them by 30%. Early results suggest that the app's feedback can enhance communication, making conversations easier and more effective. Overall, these initial findings are promising for helping people with Parkinson's speak better and engage more in social activities. ^{1 3 6 7 8}

Who Is on the Research Team?

Gemma Moya-Galé, Ph.D.

Principal Investigator

Teachers College, Columbia University

Are You a Good Fit for This Trial?

This trial is for individuals with Parkinson's Disease who are interested in improving their speech. Participants will need to attend initial and follow-up assessments in a lab, engage in at-home treatment sessions using phone and computer, and be monitored online by a speech therapist.

Inclusion Criteria

Participants scoring ≥ 25 in the MMSE

Participants scoring ≤ 24 in the BDI-III

My Parkinson's medication dose has been stable.

Exclusion Criteria

I have undergone deep brain stimulation surgery.

I have been diagnosed with a form of parkinsonism that is not typical or another neurological condition.

I have a speech or voice issue not caused by Parkinson's disease.

Timeline for a Trial Participant

Screening

Participants are screened for eligibility to participate in the trial

1 day

1 visit (in-person)

Pre-treatment Assessment

Participants complete 2-4 initial speech assessments in-person

1-2 weeks

2-4 visits (in-person)

Treatment

Participants engage in speech treatment sessions at home using their phone and computer, with some sessions monitored online

4 weeks

16 sessions (home-based, 4 monitored online)

Post-treatment Assessment

Participants come to the lab for two post-treatment assessments

1 week

2 visits (in-person)

Follow-up

Participants are monitored for safety and effectiveness after treatment

6 months

1 visit (in-person)

What Are the Treatments Tested in This Trial?

Interventions

• Noise-augmented Automatic Speech Recognition

Trial Overview The study tests if noise-augmented feedback through automatic speech recognition can help people with Parkinson's speak more clearly (intelligibility) and participate more in social situations. It involves regular home-based treatments over four weeks, plus assessments.

How Is the Trial Designed?

Group I: Immediate treatment Experimental Treatment 1 Intervention

Participants will receive speech treatment immediately after their second pre-treatment assessment.

Group II: Deferred treatment Active Control 1 Intervention

Participants will receive speech treatment one month after their second pre-treatment assessment. This arm receives the same speech treatment as Arm 1.

Noise-augmented Automatic Speech Recognition is already approved in United States for the following indications:

Approved in United States as Understand Me for Life for:

• Speech treatment for Parkinson's disease

Who Is Running the Clinical Trial?

Teachers College, Columbia University

Lead Sponsor

Trials

27

Recruited

6,100+

Temple University

Collaborator

Trials

321

Recruited

89,100+

Published Research Related to This Trial

In a pilot study involving 9 patients with Parkinson's disease and dysarthria, augmented visual feedback significantly improved the size of articulatory movements during speaking tasks, with 8 out of 9 participants showing increased movement amplitude after training.

The majority of participants were able to retain the learned skill 48 hours after training, indicating that this method could be a feasible approach for enhancing speech in individuals with hypokinetic dysarthria.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Game-Based Augmented Visual Feedback for Enlarging Speech Movements in Parkinson's Disease. Yunusova, Y., Kearney, E., Kulkarni, M., et al.[2019]

In a study involving 10 patients with Parkinson's disease, the use of white masking noise at 90 dB significantly increased their voice intensity, suggesting a potential new treatment approach.

This method may be particularly beneficial for patients who do not respond well to traditional drug therapies or behavioral speech therapies, especially with the development of portable voice-activated maskers.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Can the Lombard effect be used to improve low voice intensity in Parkinson's disease? Adams, SG., Lang, AE.[2019]

In a randomized control trial with 64 participants with Parkinson's disease, the LSVT LOUD treatment significantly improved speech intelligibility in noisy environments compared to the LSVT ARTIC treatment and no treatment.

The study highlights that focusing on prosodic features, like vocal volume and pitch, through LSVT LOUD can enhance the clarity of speech for individuals with Parkinson's disease, thereby potentially improving their quality of life.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Single Word Intelligibility of Individuals with Parkinson's Disease in Noise: Pre-Specified Secondary Outcome Variables from a Randomized Control Trial (RCT) Comparing Two Intensive Speech Treatments (LSVT LOUD vs. LSVT ARTIC). Schulz, G., Halpern, A., Spielman, J., et al.[2021]

Background noise significantly affects the intelligibility of both disordered (dysarthric) and control speech, but the impact is similar for both types, meaning that increased noise reduces clarity equally regardless of speech disorder.

The study found no interaction between the type of speech and the level of background noise, indicating that dysarthric speech is inherently less intelligible, but the presence of noise does not worsen this effect beyond what is already caused by the speech disorder itself.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Combining degradations: The effect of background noise on intelligibility of disordered speech. Yoho, SE., Borrie, SA.[2019]

The PD Check-In model significantly helped maintain vocal intensity in people with Parkinson's disease (PWPD) for up to 24 months after intensive LSVT LOUD treatment, with sustained improvements in vocal intensity measures like maximum sustained phonation.

Participants reported high satisfaction with the PD Check-In program (93.75% for PWPD and 79.99% for communication partners), indicating that the program's benefits extend beyond just acoustic outcomes, supporting their self-management and motivation for speech practice.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Clinical outcomes of PD Check-In, a model for supported self-managed maintenance of speech in Parkinson's disease: A Phase 1 study. Finnimore, A., Theodoros, D., Rumbach, AF.[2023]

A study involving 20 participants with Parkinson's disease tested different biofeedback configurations using a portable voice accumulator (VoxLog) to improve vocal loudness during speech.

The optimal configuration for effective feedback was found to be activating when the speaker's volume fell below 3 dB of their average loudness, with a response time of 500 ms, achieving a feedback frequency of 21.2%.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Motor-Learning-Based Adjustment of Ambulatory Feedback on Vocal Loudness for Patients With Parkinson's Disease. Gustafsson, J., Ternström, S., Södersten, M., et al.[2022]

This pilot study evaluated the performance of Google Cloud speech-to-text technology for recognizing speech from individuals with dysarthria (a speech disorder) compared to healthy speakers, particularly in noisy environments with multi-talker babble.

The findings suggest that while there are challenges in AI performance due to noise, current tools can be used to effectively measure speech intelligibility, indicating potential for developing assistive technologies for people with Parkinson's disease.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Automatic Speech Recognition in Noise for Parkinson's Disease: A Pilot Study. Goudarzi, A., Moya-Galé, G.[2022]

The PD Check-In program, developed as a self-management intervention for people with Parkinson's disease (PD) after LSVT LOUD treatment, showed that participants maintained vocal intensity above pre-treatment levels for up to 24 months, indicating potential long-term benefits for speech maintenance.

While the program had a positive impact on vocal intensity, the psychosocial effects on quality of life varied among participants, suggesting that further evaluation of PD Check-In is needed to fully understand its efficacy and benefits.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

PD Check-In: The development and trial of a supported self-management program for people with Parkinson's disease following intensive speech intervention. Finnimore, A., Theodoros, D., Rumbach, AF.[2022]

Individuals with Parkinson's disease (PD) use their voice 50%-60% less in daily life compared to matched healthy controls, with a significant decrease in voice sound level, averaging 8.11 dB lower for females and 6.7 dB lower for males.

The study highlights that environmental noise levels further exacerbate the differences in voice use, suggesting that assessments of voice function should consider real-life situations rather than just controlled clinical settings.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Voice Use in Daily Life Studied With a Portable Voice Accumulator in Individuals With Parkinson's Disease and Matched Healthy Controls. Gustafsson, JK., Södersten, M., Ternström, S., et al.[2020]

Citations

1. michaeljfox.org michaeljfox.org/grant/understand-me-life-using-noise-augmented-automatic-speech-recognition-improve-intelligibility

Understand Me for Life: Using Noise-augmented Automatic ... Impact on Diagnosis/Treatment of Parkinson's disease: Our project represents the first study of a self-administered, speech-treatment app that provides feedback ...

2. michaeljfox.org michaeljfox.org/grant/supplement-understand-me-life-using-noise-augmented-automatic-speech-recognition-improve

Using Noise-augmented Automatic Speech Recognition ... Impact on Diagnosis/Treatment of Parkinson's disease: Our project represents the first speech treatment study for individuals with PD to implement feedback on ...

3. pmc.ncbi.nlm.nih.gov pmc.ncbi.nlm.nih.gov/articles/PMC9634525/

Automatic Assessment of Intelligibility in Noise ... The goal of this study was to pilot-test and validate the use of a customized web-based app to assess speech intelligibility in noise in individuals with ...

4. frontiersin.org frontiersin.org/journals/artificial-intelligence/articles/10.3389/frai.2021.809321/full

Automatic Speech Recognition in Noise for Parkinson's ... This pilot study reports on the performance of Google Cloud speech-to-text technology with dysarthric and healthy speech in the presence of multi-talker babble ...

5. slcl.illinois.edu slcl.illinois.edu/news/2024-10-21t201245/automatic-speech-recognition-learned-understand-people-parkinsons-disease

Automatic speech recognition learned to understand people ... Listening to people with Parkinson's disease made an automatic speech recognizer 30% more accurate, according to initial findings from the ...

6. pmc.ncbi.nlm.nih.gov pmc.ncbi.nlm.nih.gov/articles/PMC8727902/

Automatic Speech Recognition in Noise for Parkinson's ... This pilot study reports on the performance of Google Cloud speech-to-text technology with dysarthric and healthy speech in the presence of multi-talker babble ...

7. jmir.org jmir.org/2022/10/e40567/

Automatic Assessment of Intelligibility in Noise ... This study aimed to develop, pilot-test, and validate the use of a web-based app, Understand Me for Life, to automatically measure speech ...

8. mdpi.com mdpi.com/2306-5354/12/7/728

Speech-Based Parkinson's Detection Using Pre-Trained ... This study investigates the efficacy of fine-tuning pre-trained Automatic Speech Recognition (ASR) models, specifically Wav2Vec 2.0 and HuBERT, for PD detection ...