Voice Features and Cognitive Status in Korean Speakers
Author Information
Author(s): Lee Jiho, Kim Nayeon, Ha Ji-Wan, Kang Kyunghun, Park Eunhee, Yoon Janghyeok, Park Ki-Su, Han Jae-Ho
Primary Institution: Neopons Inc.
Hypothesis
Can voice acoustic features be used to detect cognitive impairment in Korean speakers?
Conclusion
The study suggests that certain voice features may indicate cognitive status, but demographic factors significantly influence these relationships.
Supporting Evidence
- The Deep Neural Network model achieved a PR-AUC score of 0.737 for severe vs. no impairment.
- Demographic factors like age and education significantly influenced voice feature relationships with cognitive status.
- Voice features such as DDA shimmer and stdevF0 were identified as important indicators of cognitive impairment.
Takeaway
Researchers looked at how people's voices can show if they have trouble thinking. They found some voice patterns that might help tell if someone has cognitive issues.
Methodology
Speech data from 223 Korean patients were collected and analyzed using machine learning models to classify cognitive status based on voice features.
Potential Biases
The study did not control for age and education, which could affect the interpretation of voice features.
Limitations
Demographic differences in age and education levels between groups may confound results.
Participant Demographics
Participants were Korean patients with suspected cognitive impairment, with varying ages and education levels.
Statistical Information
P-Value
p<0.001
Statistical Significance
p<0.001
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website