A unified approach to combine temperature estimation and elastography for thermal lesion determination in focused ultrasound thermal therapy

Sonogram-based temperature estimation and elastography have both shown promise as methods of monitoring focused ultrasound (FUS) treatments to induce thermal ablation in tissue. However, each method has important limitations. Temperature estimates based on echo delays become invalid when the relationship between sound speed and temperature is nonlinear, and are further complicated by thermal expansion and other changes in tissue. Elastography can track thermal lesion formation over a wider range of elasticity, but with low specificity and high noise. Furthermore, this method is poor at small lesion detection. This study proposes integrating the two estimates to improve the quality of monitoring FUS-induced thermal lesions. Our unified computational kernel is tested on three types of phantoms. Experiments with type I and type II phantoms were conducted to calibrate the thermal mapping and elastography methods, respectively. The optimal settings were then used in experiments with the type III phantom, which contains ex vivo swine liver tissue. Three different spatial-peak temporal-average intensities (Ispta; 35, 133 and 240 W cm-2) were delivered with a sonication time of 60 s. The new procedure can closely monitor heating while identifying the dimensions of the thermal lesion, and is significantly better at the latter task than either approach alone. This work may help improve the current clinical practice, which employs sonograms to guide the FUS-induced thermal ablation procedure.