Related Experiment Video
Updated: Oct 7, 2025
11:00
Targeted Training of Ultrasonic Vocalizations in Aged and Parkinsonian Rats
Published on: August 8, 2011
19.9K
Automatic Speech Recognition in Noise for Parkinson's Disease: A Pilot Study.
Alireza Goudarzi1, Gemma Moya-Galé2
1Factorize, Tokyo, Japan.
Frontiers in Artificial Intelligence
|January 10, 2022
Summary
This study tested Google Cloud speech-to-text for Parkinson's patients with speech disorders. Results show AI performance varies with noise, impacting assistive technology development for speech intelligibility.
Area of Science:
- Speech Technology
- Artificial Intelligence
- Assistive Technology
Background:
- Artificial intelligence (AI) sophistication has grown, but its performance in noisy environments, especially with disordered speech, remains a challenge.
- Generalizing AI behavior to real-world conditions is difficult, particularly for individuals with speech impairments affecting intelligibility.
- Developing effective assistive technology for people with Parkinson's disease requires understanding AI performance with dysarthric speech.
Purpose of the Study:
- To evaluate the performance of Google Cloud speech-to-text technology with dysarthric and healthy speech under varying noise conditions.
- To assess the potential of automatic speech recognition (ASR) as an assistive tool for individuals with Parkinson's disease.
- To explore the impact of multi-talker babble noise intensity on AI speech recognition accuracy.
Main Methods:
- A pilot study was conducted using Google Cloud speech-to-text technology.
- The system was tested with speech samples from individuals with dysarthria (associated with Parkinson's disease) and healthy controls.
- Speech samples were analyzed in the presence of multi-talker babble noise at different intensity levels.
Main Results:
- Google Cloud speech-to-text technology demonstrated variable performance with dysarthric speech in noisy conditions.
- The intensity of multi-talker babble noise significantly affected the accuracy of speech recognition for both groups.
- Despite limitations, the study identified controllable aspects of AI performance relevant to real-world intelligibility measurement.
Conclusions:
- AI speech recognition systems show promise for assistive technology in Parkinson's disease but require careful calibration for noisy environments.
- Performance variability highlights the need for further research into robust AI solutions for individuals with speech disorders.
- Current AI tools can be adapted to measure speech intelligibility in challenging, real-life scenarios.
Related Concept Videos
01:24Parkinson's Disease: Treatment
435
Neurodegenerative disorders, such as Parkinson's Disease (PD), involve the gradual and irreversible destruction of neurons in particular brain areas. These disorders exhibit standard features like proteinopathies, selective vulnerability of some neurons, and an interaction of intrinsic properties, genetics, and environmental influences in neural injury.
Parkinson's Disease is primarily a result of the loss of dopaminergic neurons in the substantia nigra pars compacta. The cornerstone of...
Parkinson's Disease is primarily a result of the loss of dopaminergic neurons in the substantia nigra pars compacta. The cornerstone of...
435
01:15Parkinson's Disease: Overview
808
Neurodegenerative disorders are progressive diseases that cause irreversible damage and loss to neurons in specific brain areas. Examples of these disorders include Parkinson's disease, Alzheimer's disease, Multiple Sclerosis (MS), and Amyotrophic Lateral Sclerosis (ALS). These disorders share characteristics such as proteinopathies, selective neuronal vulnerability, and a complex interplay between genetic and environmental factors. The primary therapeutic goal for these conditions is...
808