Voice-based Disease Recognition
Voice-based Disease Recognition projects are part of the Human-Centered Monitoring (HCM) research. This particular project involves research into voice analaysis and acoustic event detection to automatically recognize dysarthria. Dysarthria is a speech motor disorder that is a neurological sympthom arising from cerebral dysfunction. It is an indicator of certain neurological conditions including Parkinson's Disease, Multiple Sclerosis, Myasthenia Gravis, and cerebrovascular disease. In addition to identifying cues in dysarthric speech to enable clinical classifications, we plan to develop a speech-based index to assess when an individual may require medical intervention to e.g. prevent choking due to dysphagia. Uniquely, this research combines analyses of phonetic, prosodic and voice quality cues alongside neurological expertise.
- Self-monitoring disease status: Patients monitor the status of their disease from home in collaboration with specialists.
- Remote intervention strategy: MDs determine the efficacy of therapies with periodic voice analyses performed from the patient's home or care facility.
- Acoustic event detection
- Phonetic, prosodic, and spectral analyses
- Melody and intonation characterization scripts
- Dysarthria Identification Interview Technique
- SPARK (Speech-based recognition of Parkinson’s Disease) pilot
- Intonation patterns of Dutch languages
- Ethernet Sound Interface (ESI-121-24)
- Dysarthria Identification Interview Technique (DE/NL)