The first phantom study on the diagnostic accuracy of quantitative ultrasound elastography

Objective To evaluate the effects of the range and the frequency of the compression load on the accuracy for discerning target stiffness differences in ultrasound elastography. Methods Quantitative ultrasound elastography was achieved by integrating two compression force sensors, a laptop computer and a clinical ultrasound elastographic system. The force sensors and the ultrasound probe were assembled in a 3D printed mounting bracket for continuous monitoring of compression loads during ultrasound elastography. Both the force measurements and the elastographic maps were acquired and displayed on the laptop computer in real time. Four targets of the same diameter(10.4 mm), the same depth (3 cm) and different stiffness levels (8, 14, 45 and 80 kPa) were examined by a HITACHI preirus, L74M linear-array transducer. Each target was evaluated 45 times with two different method (i.e., freehand elastography and quantitative elastography), yielding 180 evaluations. The data were divided into the following three groups: group Ⅰ(80 kPa vs 45, 14 and 8 kPa), group Ⅱ(80, 45kPa vs 14, 8 kPa) and group Ⅲ(80, 45 and 14 kPa vs 8 kPa). Area under ROC curves(AUC) were calculated for different stiffness levels. Results In group Ⅲ, quantitative elastography yielded an greater AUC level than that of freehand elastography(P=0.0379). In group Ⅰ and group Ⅱ, two methods yielded the similar AUC levels (P=1.000). However, quantitative elastography was able to discern 8 kPa and 14 kPa targets (P<0.001), while freehand elastography was hard to differentiate them (P=0.258). Conclusions In comparison with freehand elastography, quantitative ultrasound elastography is able to improve the accuracy for discerning different target stiffnesses. Key words: Elasticity imaging techniques; Pressure; Phantom study; Diagnostic accuracy