Objective vs subjective measure of the low contrast penetration test in ultrasound imaging

Abstract

Background: Current guidelines for evaluating low contrast penetration (LCP) in ultrasound (US) imaging, such as those from the American Association of Physicists in Medicine (AAPM) and the Institute of Physics and Engineering in Medicine (IPEM), primarily recommend qualitative assessment methods. These subjective approaches depend on visual interpretation, which introduces observer variability. As a result, there is growing interest in objective techniques to measure LCP in order to minimise inconsistency and improve reliability in quality assurance (QA) procedures. Aim: The primary objective of this study is to compare subjective and objective methods for assessing the LCP test in US QA. Specifically, it investigates how consistent and reliable each method is across different US probes and test repetitions. Research Methodology: Data was collected from thirty repeated QA tests performed on a single US system using three transducers: two linear, and one curvilinear. For each test, LCP was evaluated using both an automated method, based on the signal-to-noise ratio (SNR) from single phantom images (SPI), and five manual visual assessments performed by observers of varying US experience. Results: The automated method demonstrated significantly lower variability across repeated tests, for high-frequency linear probes. In contrast, manual results showed wider variation, especially among less experienced observers. The depth of penetration (DOP) values for the i18LX5 and i24LX8 probes were 46.5 ± 1.9 mm and 30.2 ± 0.9 mm, respectively, using the SPI method, compared to 66.9 ± 1.1 mm - 64.1 ± 7.7 mm, and 55.1 ± 1.2 mm - 42.2 ± 11.7 mm, respectively, from the manual methods. Agreement between the two methods was notably stronger for the curvilinear probe, likely due to its lower frequency and more gradual signal attenuation (SPI SD: 145.9 ± 3.8 mm, Manual SD: 143.4 ± 2.4 mm - 146.7 ± 12.5 mm). Conclusions & Recommendations: The objective SPI method provides a more consistent and repeatable measure of LCP than manual visual assessment, aligning well with existing literature. Variability in settings such as time-gain compensation (TGC) highlighted the need for standardised protocols. It is recommended to incorporate objective methods into routine QA to reduce subjectivity. Future research should investigate alternative data extraction methods, involve larger observer groups, and test varied US systems to further validate and improve LCP testing techniques.