Image-Based Metrics in Ultrasound for Estimation of Global Speed-of-Sound – Roman Denkin

Ultrasound imaging quality is heavily influenced by the accuracy of speed-of-sound (SoS) estimation in tissue. Common ultrasound devices often assume a constant tissue SoS of 1540 m/s, while existing SoS estimation methods typically rely on complex physical models of sound propagation in tissue. This study aims to develop robust SoS estimation techniques using image analysis-based metrics, many of which are well-explored within the photography industry for evaluating image focus. We evaluated 10 image analysis metrics, categorized into two primary groups: single-frame image quality and dual-frame image similarity assessments. Our innovative single-frame metric, Smoothed Threshold Tenengrad (ST-Ten), has demonstrated superior performance over its counterparts within the single-frame category. Dual-frame metrics, particularly the Correlation, showcased exceptional performance, surpassing both STen and other single-frame metrics. These findings suggest that image analysis-based metrics can provide a reliable and efficient approach for SoS estimation in ultrasound imaging.

About the speaker