Comparison of different compression methods for ultrasound elastography in order to detect HIFU ablations in the liver: preliminary in vivo results.

Guiding HIFU treatments using elastic properties of ablated tissues is gaining attention since it is non-invasive and can provide new information. Good-quality elasticity imaging requires highly controlled compressions, which are often challenging to obtain with hand-held devices. Here we report the comparison of 4 compression methods for the detection of HIFU ablations created in the liver during an open procedure. This work was performed in vivo in 4 pigs. HIFU lesions of 20 cm
³
were produced in the liver during surgery using a previously validated toroidal HIFU device working at 3 MHz. Ablations were imaged with conventional ultrasound imaging and elastography. A 7.5 MHz sectorial array placed in the centre of the HIFU device was used for elastography. The probe was placed directly on the liver and compression was performed using four different methods: (i) the probe was used to manually exert a periodic pressure on the liver, (ii) the liver was compressed manually between a PVC cylindrical plate (diameter 120 mm, height 7 mm) and the probe, and (iii) the compression was performed by the probe fixed on a motorized arm. Seven days after the treatment, ablations were imaged again using the same methodology. The resulting images were analyzed and compared by calculating CNRe, SNRe and contrast. Dimension of ablation measured on sonograms and elastograms were compared with the dimensions measured on gross pathology. The mean elastography contrast was similar for each compression method (range: 20.9 - 21.9 dB). The quality of elastograms was also similar for all compression methods, CNRe ranged between 3.12 and 4.28, SNRe ranged between 1.12 and 1.17. The correlation between dimensions measured on elastograms and on gross pathology were higher in the case of elastography obtained by compression with the plate (r = 0.82). In the other cases the correlations were: r = 0.67 for manual compression, and r = 0.62 for motorized compression. Compared with sonograms, all elastography methods shown higher contrast and better estimation of ablation dimensions. Strains generated using hand-held compression with or without a plate produce satisfactory results. However hand-held compression may have broader possible applications.