Related Experiment Videos
Standardized shear wave elastography of ex vivo lymph nodes: a multistage evaluation of platform configurations,
Qingze Tan1, Jingjian Chang2, Hao Wang1
1Medical School, Nantong University, Nantong, Jiangsu, 226007, China; Department of Thoracic Surgery, Yancheng Third People's Hospital, The Sixth Affiliated Hospital of Nantong University, Yancheng, Jiangsu, 224000, China.
Rationale And Objectives:
Ex vivo lymph node (LN) shear wave elastography (SWE) enables mechanical phenotyping, but measurements depend on specimen support, coupling, loading, and platform configuration. We aimed to establish and evaluate a standardized acquisition and quality-control (QC) framework.
Materials And Methods:
This prospective, single-center, ex vivo methodological study included 141 LNs from 20 surgical patients in four non-overlapping cohorts. We compared water-tank coupling-pad (WT-CP) with water-tank nylon-mesh (WT-NM; n = 24), then compared the selected WT-CP configuration with gel-cup (GC) and sandwich coupling-pad (SCP; n = 36), and finally compared selected WT-CP with SCP in a final-stage paired comparison (n = 46). An independent cohort assessed reproducibility (n = 35). Outcomes included feasibility, QC burden, within-LN coefficient of variation (CV%), mean elasticity (Emean), confidence, and operator agreement.
Results:
WT-NM yielded lower Central and Whole-LN ROI Emean than WT-CP (both P < 0.001) without improving Central ROI CV% (P = 0.119). In platform screening, LN-level PASS was 36/36 for selected WT-CP, 31/36 for GC, and 35/36 for SCP; GC had the greatest QC burden. In the final-stage paired comparison, LN-level PASS was 46/46 for selected WT-CP and 45/46 for SCP. Central ROI CV% was 4.14 % versus 3.67 % (P = 0.761), whereas Emean was 17.02 versus 20.30 kPa (P < 0.001). Interoperator ICC(A,1), based on three-acquisition averaged measurements, was 0.900 for Central ROI and 0.915 for Whole-LN ROI (mean bias, 0.49 kPa; limits of agreement, -2.95 to 3.93 kPa). Intraoperator ICC(A,1) ranged from 0.946 to 0.963.
Conclusion:
Platform configuration affected feasibility, QC burden, and the absolute Emean scale; absolute values from different configurations should not be assumed interchangeable.
