A Novel UTI Plane Stability Calibration and Compensation Method Based on Skull Structural Relationships for Articulatory Data Observation

The tongue is a crucial articulator, where even minor deformations in shape and position significantly affect speech production. Ultrasound tongue imaging (UTI) has emerged as a crucial method to track complex tongue movements, improving our understanding of how the vocal tract geometry impacts speech production. Superior imaging results also benefit various UTI-based tasks, such as silent speech recognition and speech sound disorder treatment, which rely on imaging plane stability. Sustaining this stability remains challenging, given its impact on UTI consistency and clarity, essential for reproducibility and recognition accuracy. To solve the challenges, we propose a stability calibration and compensation method based on skull structural relationships. This paper disentangles the stability-influencing factors into specific and comprehensible study questions by incorporating the skull structural relationships. It clarifies the mechanism through which the proposed solutions improve stability and image quality. This method is assessed using the improved evaluation framework, which is upgraded by the methods we newly proposed in this study. The evaluation results showcase the effectiveness of our proposed method, showing superior performance over the benchmark. Crucial features, such as the abrupt edge gray change in the UTI images, are provided a preliminary and quantitative representation by integrating our calibration and evaluation methods. The results may further clarify the relationship between the stability, vocal actions, UTI images, energy variations, and improvements. This facilitates more reliable and accurate applications of UTI-based tasks.