Semi-automatically tracking articulatory features in 2D dynamic sagittal MRI images

Motivation: The motivation for this research is from the need to improve diagnosis and treatment of speech disorders, as well as advancements in speech synthesis. Goal(s): By analyzing the movement of the lips, tongue and velum during speech, we aim to gain insights into the coordination required for accurate pronunciation. Approach: UsUsing coordinate data from corresponding frames of the corresponding phrase across five samples, we measure y-coordinate difference, calculating the velocity for the feature's movements. Results: Results showed that 14 mm tongue displacement during "at," with velocities of 54mm/s and 52mm/s. Velum contact occurred in 78% of total frames from 17500 frames. Impact: These results provide insights into tongue and velum movement during speech, potentially aiding clinicians in diagnosing speech disorders and improving therapeutic techniques. This study also suggests a question on the neural control of articulation and speech production accuracy.