Quasi-Static Elastography And Its Investigation Of Focused Ultrasound Induced Tissue Lesions

BACKGROUNDHigh intensity focused ultrasound has been a novol noninvasive modality for clinical treatment of tumor diseases and non-tumor diseases. Its principle is to direct low intensity ultrasound from outside into the living body to form a high intensity focused ultrasound target, where high intensity ultrasound produces hyperthermia, cavitation and mechanical effects. The temperature at the focus rise to 60℃above, inducing nonreversible damage, coagulative necrosis and protein denaturation, thus ablated the lesion tissue at the focus.Monitoring is a key factor to the success of high intensity focused ultrasound (HIFU) therapy. If the HIFU focus in the tissue is like a scalpel, the monitoring is like the eyes of the doctor. Safe and effective therapy can only be achieved by master of the physiological and pathological structure and real-time management of ablated region and ablation gradient. The main modality for HIFU treatment monitoring is now B-mode ultrasonography and Magnetic resonance imaging (MRI). B-mode ultrasonography has advantage of real-time monitoring, simplicity and easy combination with HIFU system. But it can't measure the tissue temperature in real-time, thus can't precisely reflect the real progress of coagulative necrosis at the target. MRI can measure the tissue temperature in real-time, thus can tell the temperature at the target and the progress of coagulative necrosis very well. But MRI system is very hard to be combined with HIFU system, and it is too costly for the patients. Thus, it is necessary to investigate a new modality for monitoring HIFU therapy and evaluating the outcome.The lesion produced by HIFU mainly has the appearance of nucleus breakdown or pyknosis, protein denaturation, cellular structure destruction and surrounding tissue swelling, et al. These pathological changes result into the elasticity change of the tissue, which appears to be that HIFU ablated tissue is significantly stiffer the surrounding tissue. Because ultrasound elasticity shows promise in monitoring HIFU therapy, many studies about elastography monitoring HIFU therapy have been done in the last decade.Former studies have confirmed that elastography can investigate HIFU induced coagulative necrosis inside the tissue and determine the size and boundary of the coagulative necrosis [19, 20]. While previous studies mainly focused on the coagulative necrosis produced by HIFU, in this study, comparatively low intensity focused ultrasound has been used to form lesions inside the tissue with small dosage of acoustic power. Quasi-static elastography had been carried out and the lesions inside the tissue had been successfully detected. During the experiments, a series of improvements on the algorithms had been made to improve the motion detection accuracy of quasi-static elastography, thus producing satisfying imaging results.OBJECTIVES1. To enhance the motion detection accuracy and improve the result of elastography by optimizing the algorithm of elastography.2. To carry out phantom experiments to conform the effects of the elastography and compare the effects of different compression ratio on the experimental results.3. To prove the feasibility by carrying out experiments of quasi-static elastography for the investigation of FUS induced tissue lesions.METHODS1. To optimize the algorithm of elastography.1.1 To use two dimensional cross-correlation algorithm to eliminate the decorrelation effects by tissue lateral displacement.1.2 To use parabola fitting interpolation to determine the non-integer displacement.1.3 To use piecewise linear approximation to remove and replace the decorrelation values in the displacement.1.4 To use two dimensional median filter to smooth the axial displacement data and eliminate high frequency noise.1.5 To use low pass digital differentiator (LPDD) and LPDD based on wavelet transforms to effectively reduce the high frequency noise during the axial difference of the displacement data.2. To carry out phantom experiments to conform the effects of the elastography.2.1 To make a phantom with a hard cylinder inclusion and a three layer phantom with a soft middle layer, and to carry out experiments of elastography for the investigation of phantom inclusions.2.2 To use a standard phantom (CIRS Model 049A) to compare the imaging results of elastography under different compression ratio.3. To form coagulative necrosis inside tissue using CZF gynaecology focused ultrasound therapy system (Chongqing Haifu Co.ltd). Then to investigate the lesion inside tissue produced by focused ultrasound using a Sonix RP B-mode ultrasonography system (Ultrasonics Co.). The results of elastography were compared with B-mode images and the sectional plane photography of the real lesion to determine the outcome and feasibility of elastography investigating FUS induced lesions inside tissue.RESULTS1. Decorrelation effects in axial displacement estimation could be effectively reduced by two dimensional cross-correlation algorithm and relatively ideal axial displacement curve could be achieved. Parabola fitting interpolation could enhance the accuracy of displacement estimation and reduce systematic error in displacement estimation. Piecewise linear approximation could effectively detect the decorrelation values in axial displacement data and replace them with results of linear approximation. Two dimensional median filter could eliminate high frequency noise and smooth the axial displacement data. Low pass digital differentiator (LPDD) and LPDD based on wavelet transforms could effectively reduce the high frequency noise during the axial difference of the displacement data, which had significantly better results than that of direct difference algorithm. And LPDD based on wavelet transforms had a more ideal result than LPDD SG-Ⅰand LPDD SG-Ⅱ.2. Phantom experiments confirmed that quasi-static elastography could successfully detect the hard inclusions in phantoms and tell the stiffness difference in three layer phantom with a soft middle layer. The experiments on standard phantom indicated that different compression ratio had a big influence on the experiment results and a suitable compression ratio must be chosen to have a better experiment result. Experiments showed that the compression ratio between 0.5%-0.8% would be preferable.3. Biological tissue experiments showed that quasi-static elastography could successfully detect FUS induced tissue lesion, while the non-immediate B-mode image could not distinguish lesions and normal tissue. The size and shape of the lesion in elastography accorded very well with the sectional plane photography across the real lesion. Strain variation inside the lesion could reflect different degree of tissue damage.CONCLUSION1. Motion detection accuracy in elastography can be effectively improved by taking a serious of algorithm improvements, which can improve the outcome of elastography.2. Quasi-static elastography can successfully detect FUS induced lesions inside tissue and clearly show the boundary of the lesion, while the non-immediate B-mode images could not distinguish the tissue lesions by FUS and surrounding normal tissue. Elastography could even distinguish different degree of lesions produced by FUS inside tissue.