The Combination Of A Computational Model Of The Circulatory System With Clinical Data And Its Application To The Assessment Of Cardiovascular Function

The assessment of cardiovascular function may provide important references for making appropriate therapeutic strategies and performing rational drug administration in the treatment of cardiovascular diseases.However,quantitative assessment of cardiovascular function remains difficult in current clinical settings.In the context,researchers have proposed to assess cardiovascular function through integration of a biomechanical model of the circulatory system and clinical data,and accordingly carried out a series of validatory works.However,previous studies rarely address issues related to the accuracy of assessment,which considerably limits the clinical applicability of the proposed assessment methods.The main purpose of the present thesis is to identify the key factors that affect the accuracy of patient-specific assessment of cardiovascular function(by means of integration of clinical data and cardiovascular model)and quantify their effects so as to provide theoretical evidence for guiding the clinical application of the assessment method.In the meantime,the assessment method was improved to enhance the relationships between assessment results and the information usable to clinical practice.Parameter sensitivity analysis was performed in combination with parameter subset selection to identify the secondary main parameters that are related closely to the model-based prediction of hemodynamic variables and the assessment of the main parameters(corresponding to the assessed cardiovascular function).Numerical experiments were carried out based on a series of virtual clinical data to first quantify the changes in assessment results induced by the measurement error(in a range of 0-5%)of a single clinical data and the variation in each secondary main parameter(rate of change being 0-30%).Subsequently,the effects of different combinations of measurement error and parameter variation on assessment results were investigated by means of a multivariable analysis method.Furthermore,the assessment system was modified so that the time constant of ventricular isovolumetric relaxation can be assessed from the pressure curve of the left ventricle and volumetric data,thus enriching the assessment of cardiac function and improving the efficiency of clinical data use.The main findings of the study included:(1)measurement errors of clinical data induced pronounced changes in assessment results,whereas,variations in the secondary main parameters had relatively less influence on the assessment results;(2)the errors in assessment results changed nearly linearly with the magnitudes of the measurement errors of clinical data or the variations in the secondary main parameters;and(3)the assessment errors resulting from coexisting measurement errors in multiple clinical data or simultaneous variation of multiple secondary main parameters were comparable to those obtained in univariate analysis.Based on the results obtained in the study,it was concluded that accurate clinical data measurement is a key step to guaranteeing the reliability of cardiovascular function assessment.The secondary main parameters only have limited influence on the assessment although they may vary significantly among patients,demonstrating that the adopted method for assessing cardiovascular function is robust and has promising potential for clinical use.Due to the fact that the present study has been performed using virtual clinical data,future studies would be required to further examine the clinical value of the method based on patient-specific clinical data.