Repeatability and Reproducibility of a Clinically Based QUS Phantom Study and Methodologies.

Area of Science:

• Medical Imaging
• Biophysics
• Ultrasound Technology

Background:

• Quantitative ultrasound (QUS) offers non-invasive liver disease assessment.
• Understanding the repeatability and reproducibility (R&R) of QUS metrics is crucial for clinical translation.
• Existing R&R data for liver QUS using specific clinical scanners is limited.

Purpose of the Study:

• To evaluate the R&R of quantitative ultrasound (QUS) estimates, specifically attenuation coefficient (AC) and backscatter coefficient (BSC).
• To detail the measurement and analysis methodology for QUS in a clinical setting.
• To assess QUS performance using a Siemens 3000 clinical ultrasound scanner.

Main Methods:

• Calibrated phantoms were scanned by two sonographers using two transducers over multiple sessions.
• Scans were performed within a clinical QUS liver study involving human participants with varying obesity and liver disease severity.
• Repeatability and reproducibility were analyzed using ANOVA with two random effect models.

Main Results:

• The measurement variance attributed to repeatability and reproducibility was small for both AC and BSC.
• AC variance: 2.4-3.2×10^-4 [dB/cm-MHz]²; 10log₁₀BSC variance: 0.23-0.27 dB².
• Reproducibility variance was significantly lower than repeatability variance.

Conclusions:

• The study demonstrates small measurement variance for QUS AC and BSC, supporting its reliability.
• QUS R&R was not significantly influenced by phantom properties across a relevant range for in vivo liver.
• The findings support the clinical utility of QUS for liver disease assessment with the evaluated scanner.