Lockwood Heather, McLeod Graeme A
School of Medicine, University of Dundee, Dundee, UK.
Institute of Academic Anesthesia, School of Medicine, University of Dundee, Dundee, UK.
Ultrasound. 2022 Nov;30(4):346-354. doi: 10.1177/1742271X221091726. Epub 2022 May 9.
Shear wave elastography (SWE) presents nerves in colour, but the dimensions of its colour maps have not been validated with paired B-Mode nerve images. Our primary objective was to define the bias and limits of agreement of SWE with B-Mode nerve diameter. Our secondary objectives were to compare nerve area and shape, and provide a clinical standard for future application of new colour imaging technologies such as artificial intelligence.
Eleven combined ultrasound-guided regional nerve blocks were conducted using a dual-mode transducer. Two raters outlined nerve margins on 110 paired B-Mode and SWE images every second for 20 s before and during injection. Bias and limits of agreement were plotted on Bland-Altman plots. We hypothesized that the bias of nerve diameter would be <2.5% and that the percent limits of agreement would lie ±0.67% (2 SD) of the bias.
There was no difference in the bias (95% confidence interval (CI) limits of agreement) of nerve diameter measurement, 0.01 (-0.14 to 0.16) cm, = 0.85, equivalent to a 1.4% (-56.6% to 59.5) % difference. The bias and limits of agreement were 0.03 (-0.08 to 0.15) cm, = 0.54 for cross-sectional nerve area; and 0.02 (-0.03 to 0.07), = 0.45 for shape. Reliability (ICC) between raters was 0.96 (0.94-0.98) for B-Mode nerve area and 0.91 (0.83-0.95) for SWE nerve area.
Nerve diameter measurement from B-Mode and SWE images fell within a priori measures of bias and limits of agreement.
剪切波弹性成像（SWE）以彩色显示神经，但彩色图像的尺寸尚未与配对的B超神经图像进行验证。我们的主要目标是确定SWE与B超神经直径的偏差和一致性界限。我们的次要目标是比较神经面积和形状，并为人工智能等新型彩色成像技术的未来应用提供临床标准。
使用双模换能器进行11次联合超声引导区域神经阻滞。两名评估者在注射前和注射期间每隔一秒在110对配对的B超和SWE图像上勾勒神经边缘，共持续20秒。偏差和一致性界限绘制在布兰德-奥特曼图上。我们假设神经直径的偏差将小于2.5%，一致性百分比界限将在偏差的±0.67%（2个标准差）范围内。
神经直径测量的偏差（一致性95%置信区间（CI）界限）无差异，为0.01（-0.14至0.16）厘米，P = 0.85，相当于差异1.4%（-56.6%至59.5%）。横截面神经面积的偏差和一致性界限为0.03（-0.08至0.15）厘米，P = 0.54；形状的偏差和一致性界限为0.02（-0.03至0.07），P = 0.45。评估者之间的可靠性（ICC）对于B超神经面积为0.96（0.94 - 0.98），对于SWE神经面积为0.91（0.83 - 0.95）。
B超和SWE图像的神经直径测量值在预先设定的偏差和一致性界限范围内。
