Bankier Alexander A, O'Donnell Carl R, Mai Vu M, Storey Pippa, De Maertelaer Viviane, Edelman Robert R, Chen Qun
Department of Radiology, University of Vienna, Vienna, Austria.
J Magn Reson Imaging. 2004 Dec;20(6):961-6. doi: 10.1002/jmri.20198.
To test the hypothesis that, in magnetic resonance (MR) imaging of healthy individuals, equal relative changes in lung volume cause equal relative changes in MR signal intensity of the lung parenchyma.
In two experimental runs, 10 volunteers underwent spirometrically monitored MR imaging of the lungs, with MR images acquired at 10 incremental lung volumes ranging from total lung capacity to 10% above residual volume. Average signal intensity, signal variability, and signal intensity integrals were calculated for each volunteer and for each lung volume. The effect of lung volume on signal intensity was quantified using linear regression analysis complemented by the runs test. Slopes and intercepts of regression lines were compared with an analysis of covariance. Slopes of the lines of best fit for lung volumes and signal intensities from the two runs were compared to the slope of the line of identity. Comparisons between the two runs were visualized using Bland and Altman plots.
The slopes of the 10 individual regression lines yielded no significant differences (F = 1.703, P = 0.101; F = 1.321, P = 0.239). The common slopes were -0.556 +/- 0.027 (P = 0.0001) for the first and -0.597 +/- 0.0031 (P = 0.0001) for the second experimental run. Both slopes displayed no significant nonlinearity (P = 0.419 and P = 0.067). There was a strong association between changes in lung volumes (rs = 0.991, P = 0.0001) and changes in signal intensity (rs = 0.889, P = 0.0001) in the two experimental runs. Lines of best fit for lung volume and signal intensities were not significantly different from the slope of the line of identity (P = 0.321 and P = 0.212, respectively).
Equal changes in lung volume cause equal changes in MR signal intensity of the lung parenchyma. This linear and reproducible phenomenon could be helpful in comparing pulmonary MR signal intensity between individuals.
验证以下假设，即在健康个体的磁共振（MR）成像中，肺容积的相对变化相等会导致肺实质MR信号强度的相对变化相等。
在两次实验中，10名志愿者接受了肺活量监测下的肺部MR成像，在从肺总量到残气量以上10%的10个递增肺容积下采集MR图像。计算每名志愿者和每个肺容积的平均信号强度、信号变异性和信号强度积分。使用线性回归分析并辅以游程检验来量化肺容积对信号强度的影响。通过协方差分析比较回归线的斜率和截距。将两次实验中肺容积与信号强度的最佳拟合线斜率与恒等线斜率进行比较。使用布兰德-奥特曼图直观显示两次实验之间的比较结果。
10条个体回归线的斜率无显著差异（F = 1.703，P = 0.101；F = 1.321，P = 0.239）。第一次实验的共同斜率为-0.556±0.027（P = 0.0001），第二次实验为-0.597±0.0031（P = 0.0001）。两条斜率均无显著非线性（P = 0.419和P = 0.067）。在两次实验中，肺容积变化（rs = 0.991，P = 0.0001）与信号强度变化（rs = 0.889，P = 0.0001）之间存在强关联。肺容积与信号强度的最佳拟合线与恒等线斜率无显著差异（分别为P = 0.321和P = 0.212）。
肺容积的相等变化会导致肺实质MR信号强度的相等变化。这种线性且可重复的现象有助于比较个体之间的肺部MR信号强度。
