| Group |
2022-14 |
Status |
completed |
| Title |
Software design for automatic voice analysis in the context of neurodegenerative diseases’ characterization |
| Supervisor |
H. Benali, L. Jeancolas |
| Description |
Voice and speech impairments are common in many neurodegenerative diseases (such as in Parkinson or Alzheimer). They provide some valuable and objective information regarding the type and severity of the disease at a low cost. Several signal processing and machine learning methods have been developed to characterize vocal disruptions in neurodegenerative diseases but are usually suited to a particular disease and tuned using specific databases.
During this project, the students will adapt and improve algorithms validated in previous experiments and publications1, 2, to make them compatible with different types of audio recordings and diseases.
The aim of this project is to develop a user-friendly software or toolbox that could be easily shared and used by other research teams (with different data) and even eventually used by medical doctors. Such a tool will be helpful for the early detection and monitoring of neurodegenerative diseases or for the assessment of the efficiency of medical treatments.
References:
[1] Jeancolas L., et al. "Voice Characteristics from Isolated Rapid Eye Movement Sleep Behavior Disorder to Early Parkinson s Disease." Parkinsonism & Related Disorders, 95 (2022), https://doi.org/10.1016/j.parkreldis.2022.01.003.
[2] Jeancolas L., et al. “X-vectors: New Quantitative Biomarkers for Early Parkinson s Disease from Speech." Front. Neuroinform. 15:578369. doi: 10.3389/fninf.2021.578369
|
| Requirement |
- Good level in programming (Matlab, Python, Bash, C++ ...)
- Good understanding of oral and written French language: the documentation of the audio processing algorithms that will be used and the audio databases are in French
- Good knowledge in signal processing and machine learning
|
| Tools |
All the necessary resources will be provided to the students. |
| Number of Students |
5 or 6 |
| Students |
S. Arsan, K. Chatta, S. Delisle, J. Moarbes, L. Nguemtchouang, C. Sfeir |
| Comments: |
Contact: habib.benali@concordia.ca; laetitia.jeancolas@concordia.ca |
| Links: |
|