Acoustic Analysis of Malayalam Digits: Comparing Healthy and Dysarthric Speech

Dysarthria is a common neurological condition that causes difficulty with exact articulation, poor coordination of respiratory and phonatory subsystems, and irregular pauses. Improving the intelligibility of dysarthric speech is a significant task in the current context. The analysis described in this paper focuses on isolated Malayalam digits. The evaluation uses the PRAAT software tool to extract intensity, pitch, duration, and formant frequencies of speech samples. The acoustic parameters such as jitter, shimmer, and Harmonics to Noise Ratio (HNR) are also investigated. This work explores the acoustic patterns of dysarthric speech across three severity levels (mild, moderate, and severe) in comparison to a healthy control group. The lexical and sub-lexical variations in these parameters help to differentiate the acoustic properties of healthy individuals from those with dysarthria. These investigations characterize the temporal dynamics and spectral properties of normal and dysarthric speech samples, providing deeper insight into their acoustic distinctions.