Ultrasound Based Mechanical Properties Quantification Of Biological Tissues And Its Applications In Diseases Diagnosis

The estimation of the mechanical properties of biological tissues are of great potential in the tissue abnormalities diagnosis.Imaging techniques are commonly used in the analysis of tissue stress-strain relationship,which is the fundamental for noninvasive and quantitative evaluation of its mechanical properties in vivo.Among other imaging modalities,ultrasound imaging has the priorities of lower cost,no ionizing radiation,real-time capability and higher portability that make it as one of the most widely used modality in the research filed of biomechanics.Therefore,this thesis focuses on the ultrasound based mechanical properties quantification of biological tissues and its applications in diseases diagnosis,the contributions of this thesis are as follows:(1)Lymphedema diagnosis based on the Sub-Herts Analysis of Viscoelasticity(SAVE)of tissues.SAVE is an ultrasound-based framework for quantitative estimation of mechanical properties using the creep response of soft tissues.The mechanical properties of the subcutaneous tissue and muscle from 24 patients’ forearms were analyzed using SAVE following:firstly,a custom made automated compression device was used to apply an uniaxial ramp-and-hold force on the studied tissues which is the requirement of creep test,and the ultrasound scanning was simultaneously conducted during the compression;Secondly,an autocorrelation based doppler equation(ACDE)was used to track the uniaxial tissue deformation induced by the compression;Then from the acquired B-mode image,2 rectangular boxes were respectively defined in the region of subcutaneous tissue and muscle as ROIs;Finally,based on the results obtained from previous steps,standard linear solid(SLS)model was used to calculate the viscoelastic parameters of the studied tissues.Our results showed that the elastic parameters of the affected arm increased significantly compared to the unaffected arm,and the viscous parameters also increased insignificantly indicating the capability of our method in the diagnosis of lymphedema.(2)A vectorized normalized cross correlation(VNCC)algorithm used in the fast tracking of myocardial motion.Compared to the subcutaneous tissue and muscle studied in(1),the myocardial motion is inherently more complex,the myocardial border is more irregular,and the accurate stress distribution on the myocardium is nearly impossible to obtain in vivo,which adds extra difficulties in the estimation of mechanical properties of myocardium.Therefore,a VNCC method was proposed to track the 2D myocardial motion and reduce the computational time.Then a displacement compensated cross correlation coefficient(DCCCC)was proposed to evaluation the motion tracking accuracy of VNCC and an optimal selection of VNCC algorithmic parameters were conducted using proposed DCCCC to improve the performance of VNCC.Finally,we used the myocardial deformation tracked by VNCC as the quantitative biomarker for the evaluation of myocardial mechanics.The ultrasound data acquired from the hearts of healthy mice was analyzed using the abovementioned method,the results shows that the proposed VNCC is capable of reduce the computational time of myocardial motion tracking without losing the accuracy,the feasibility of the above-mentioned modifications in the quantitative evaluation of myocardial function is also validated by the results.(3)Myocardial Infarction(MI)diagnosis based on the estimation of myocardial mechanics using VNCC.Myocardial motion abnormalities estimation is of great potential in early diagnosis of MI.Based on the method describe in(2),a semiautomatic regional segmentation of myocardium method was proposed for the accurate analysis of regional myocardial function which is required by the clinic examination for the MI diagnosis.Then,a quantitative biomarker for the analysis of transmural motion of myocardium was also proposed,called transmural motion index(TMI).Finally,the segmental and transmural myocardial mechanics were analyzed using above mentioned methods.Ultrasound data was acquired from 3 groups of rats(7 per group):sham,MI and ischemia-reperfusion(IR)groups and used for the myocardial mechanics analysis.Our results suggest that the myocardial contractility reduced significantly during acute ischemia phase,and this myocardial motion loss was gradually recovered from acute ischemia phase to remodeling phase after the rat model establishment.The results indicatie that our method is of great potential in non-invasive,novel and early MI detection.