Baihui Yu

Assistant Professor
Biological Sciences Division
The University of Chicago

baihui.yu@uchicagomedicine.org

Dept. and Web

Radiology

Project Description

Evaluating Element Dropouts on Ultrasound Array Transducer Using a Random Hypoechoic Sphere Phantom

Ultrasound array transducers, essential components of diagnostic ultrasound systems, are routinely exposed to wear and tear during clinical use, making them vulnerable to defects such as element dropouts. While previous studies have primarily focused on detecting these element dropouts using uniformity phantom scans or more sophisticated electronic testers that evaluate individual elements of an array transducer, questions remain regarding the impact of the element dropouts and how to establish criteria for removing a transducer from clinical service. As a result, corrective actions for these commonly observed defects are often unclear.

IEC TS 62791:2022 specified a method to quantify the human-observer-related lesion detectability by measuring the lesion signal-to-noise ratio (LSNR) using the Random Hypoechoic Sphere Phantom (RHSP) with mechanical scanning devices. We developed an automated analysis algorithm for this task using freehand scanning in the previous work. The automated analysis method provides accurate and stable quantitative assessment of the lesion detectability for ultrasound performance evaluation.

This project aims to extend this method to evaluate the performance deviations from baseline caused by element dropouts. Quantitative metrics could be integrated into routine ultrasound quality control testing to guide corrective actions when necessary. Aim 1: using RHSP method to assess the impact on lesion detectability at the location under the element dropouts. Aim 2: investigating the potential benefits of compounding techniques to mitigate the effects of element dropouts. Aim 3: extending this method to applications with curved array transducers.

Requirement

Background in medical imaging, Python