Author: Kevin M Brom, Jaydev K. Dave, Chunming Gu, Nicholas J. Hangiandreou, Zaiyang Long, Donald J Tradup 👨🔬
Affiliation: Mayo Clinic 🌍
Purpose: Ultrasound scanner onboard quality checks for shear wave elastography (SWE) rely on operator judgement and are insufficient. This study aims to establish a protocol for quality assessment and apply it to a musculoskeletal (MSK) task.
Methods: The protocol begins by identifying SWE tasks and relevant test tools. In our application, an elastography phantom (model 049A, CIRS) was scanned by a GE Logiq E10 scanner using L2-9, ML6-15, and L8-18i probes. Three sizes of cylindrical targets (10.4, 4.05, and 2.53mm) and three nominal stiffnesses (71.5, 35.7, and 9.1kPa) were determined from literature review. Second, the acquisition process considers the number of repeats, imaging planes, and presets. Five repeats were acquired in longitudinal and transaxial directions using MSK_general and MSK_superficial presets. Next, the protocol includes several steps: assessing the percentage of target pixels meeting onboard quality threshold, verifying median shear wave speed (SWS) against a developed tolerance range, excluding artifactual pixels, and assessing temporal stability.
Results: Images were processed offline using MATLAB codes. Target masks were drawn and the percentage of target pixels passing the scanner's quality threshold was compared to an 80% benchmark. SWS tolerance was derived using the largest target with ML6-15 and L2-9 cross-referenced with phantom's nominal values. Artifacts were excluded using Density-Based Spatial Clustering of Applications with Noise evaluating max, min, average, 10th, and 90th percentile SWS. 170 of 547 images passed all criteria (66 L2-9, 42 L8-18i, 62 ML6-15). 90% of L8-18i MSK_superficial images were excluded. Linear regression of MSK_general transaxial data revealed significant effects of target size on SWE (p<0.01). ML6-15 MSK_general transaxial showed the best measurement but none were acceptable for 2.53 and 4.05mm targets.
Conclusion: The quality assessment protocol was developed and applied to a MSK task with benchmarking of probe and preset performance of specific targets presented.