An Evaluation Framework for an Ultrasound Imaging-Based Articulatory Observation Instrument
Yuxuan Li, Jianguo Wei, Qiang Fang, Kele Xu, Ruiteng Zhang, Xugang Lu · IEEE Transactions on Instrumentation and Measurement · 2025
Ultrasound tongue imaging (UTI) plays a crucial role in advancing our understanding of speech production dynamics, offering invaluable insights into the intricate movements of the tongue during natural speech. Achieving imaging consistency and clarity is crucial for subsequent tasks such as treating speech sound disorders, automatic speech recognition (ASR), and silent speech synthesis. Despite sustained efforts in the field, we find that challenges persist in ensuring imaging stability during natural speech production, with drifts and instability of the imaging plane significantly contributing to this challenge. A significant strategy for addressing these challenges is the implementation of imaging calibration helmets, which are regarded as pivotal instruments for improving stability during speech. However, the existing literature highlights the necessity for a standard method to quantify imaging plane stability, assess improvements in imaging instruments, and explore the consequential impact of stability enhancements on tracking sequence quality. In response to this gap, our article presents a novel evaluation framework, supported by a sensor-based imaging stability tracking system, to assess the performance of enhanced instruments in UTI. The framework establishes a robust methodology for further defining factors that impact imaging quality when using instruments. It also provides a comprehensive means to evaluate the effectiveness and performance of improved instruments, such as our proposed imaging helmet, in terms of imaging plane stability and the impact of enhanced stability on tracking consistency and imaging clarity. The ultimate goal is to elucidate the correlation between these stability enhancements and the previously mentioned image quality, thereby facilitating more reliable and accurate applications of UTI in diverse clinical linguistic and phonetics contexts.