Zurich Ultrasound Research and Translation (ZURT), Institute of Diagnostic and Interventional Radiology, University Hospital Zurich, Rämistrasse 100, 8091, Zürich, Switzerland.
Institute of Diagnostic and Interventional Radiology, University Hospital Zurich, Rämistrasse 100, 8091, Zürich, Switzerland.
Eur Radiol. 2020 Oct;30(10):5272-5280. doi: 10.1007/s00330-020-06885-8. Epub 2020 May 8.
To compare speed of sound (SoS) ultrasound (US) of the calves with Dixon magnetic resonance imaging (MRI) for fat content quantification.
The study was approved by the local ethics committee. Fifty calf muscles of 35 women (age range 22-81 years) prospectively underwent an US and subsequent MRI (Dixon sequence) examination as well as body weight and impedance fat measurements. SoS (in m/s) was calculated positioning a reflector on the opposite side of a conventional US machine probe with the calf in between. Fiducial nitroglycerin markers were placed on the calf at the reflector and US probe end positions for later registration of the US sonification volumetric section. An automatic segmentation algorithm separated MRI adipose tissue, muscle and bone regions. MRI fat fraction of the entire leg slice (total) and intramuscular and adipose tissue fat fraction were calculated and correlation analysis and correlation coefficient comparison were performed.
Median SoS demonstrated a very strong (r = - 0.83 (95% CI - 0.90; - 0.72); p < 0.001) correlation with MRI total fat fraction, a strong (r = - 0.61 (95% CI - 0.76; - 0.40); p < 0.001) correlation with MRI adipose tissue fat fraction and a moderate (r = - 0.54 (95% CI - 0.71; - 0.31); p < 0.001) correlation with MRI intramuscular fat fraction. Impedance body fat percentage correlated strongly with SoS (r = - 0.72 (95% CI - 0.85; - 0.51); p < 0.001) and MRI total fat fraction (r = 0.61 (95% CI 0.34; 0.78); p < 0.001). For electrical impedance, significantly lower correlations (p = 0.033) were found for MRI total fat fraction compared with SoS.
Correlations of SoS with Dixon MRI fat fraction measurements were very strong to moderate.
• Correlations of speed of sound with Dixon MRI fat fraction measurements of the same body location were very strong to moderate. • Speed of sound measurements showed a high repeatability. • Speed of sound provides a sufficient discrimination range for fat fraction estimates.
比较小腿的声速(SoS)超声(US)与 Dixon 磁共振成像(MRI)在脂肪含量定量方面的差异。
本研究经当地伦理委员会批准。35 名女性(年龄 22-81 岁)的 50 块小腿肌肉前瞻性地接受了 US 和随后的 MRI(Dixon 序列)检查,以及体重和阻抗脂肪测量。SoS(以 m/s 为单位)通过在小腿之间放置在常规 US 机探头对面的反射器来计算。在小腿上的反射器和 US 探头端位置放置了基准硝化甘油标记物,以便以后对 US 声像图容积部分进行注册。自动分割算法将 MRI 的脂肪组织、肌肉和骨骼区域分开。计算整个腿部切片的 MRI 脂肪分数(总)以及肌内和脂肪组织脂肪分数,并进行相关性分析和相关系数比较。
中位数 SoS 与 MRI 总脂肪分数呈很强的相关性(r=-0.83(95%置信区间-0.90;-0.72);p<0.001),与 MRI 脂肪组织脂肪分数呈很强的相关性(r=-0.61(95%置信区间-0.76;-0.40);p<0.001),与 MRI 肌内脂肪分数呈中度相关性(r=-0.54(95%置信区间-0.71;-0.31);p<0.001)。阻抗体脂百分比与 SoS 呈很强的相关性(r=-0.72(95%置信区间-0.85;-0.51);p<0.001)和 MRI 总脂肪分数(r=0.61(95%置信区间 0.34;0.78);p<0.001)。对于电阻抗,与 SoS 相比,MRI 总脂肪分数的相关性明显较低(p=0.033)。
SoS 与 Dixon MRI 脂肪分数测量值之间的相关性非常强到中等。
同一身体部位的声速与 Dixon MRI 脂肪分数测量值之间的相关性非常强到中等。
声速测量具有较高的可重复性。
声速提供了足够的脂肪分数估计分辨范围。