2312 - A Vision Test for Ultrasound System: Quantitative Quality Assessment Using a 3D Printed Eyechart
Presenter(s)
X. Zhu1, W. Nie1, A. Copeland1, M. Joo1, N. Zhang2, M. Correa2, J. Fan1, and D. W. Kim1; 1Inova Schar Cancer Institute, Fairfax, VA, 2Inova Schar Cancer Institute, Falls Church, VA
Purpose/Objective(s): Brachytherapy often relies on ultrasound (US) imaging for accurate target and normal tissue localization. However, current QA standards, such as AAPM TG128, lack quantitative measures of image quality. This study introduces a novel, easy-to-implement, and objective method for US QA using a 3D printed eyechart, analogous to how an eye chart evaluates visual acuity in human vision. This approach enables longitudinal tracking of imaging performance in US-guided brachytherapy.
Materials/Methods: A custom 3D-printed eyechart was printed using Polylactic Acid (PLA) with character sizes ranging from 5 to 10 mm. The printed rows were submerged in a water bath, with US images acquired in both transverse and sagittal planes. Characters were positioned at distances from 10 to 50 mm from the probe surface, in 10 mm increments. An open-source optical character recognition (OCR) software processed the acquired images for detecting the characters, and detection accuracy was evaluated via confidence levels.
Results: At a confidence level of 0.35, the smallest detectable 3D-printed character was 7 mm, with partial recognition of “DEFPOTEC” (only correctly identifying P and T). Recognition accuracy improved as character size increased, demonstrating a direct correlation between ultrasound image quality and feature detectability.
Conclusion: This 3D printed eyechart offers a quantitative, objective, and reproducible method to evaluate US imaging performance in radiotherapy. It serves as a practical "vision test" for US systems, enabling consistent QA and potential longitudinal monitoring of image quality over time.