Application And Implementation Of Implicit Modeling In Blood Flow Analysis Model Construction

Hemodynamics is a crucial field in biomedical engineering,providing valuable data to improve the success and effectiveness of clinical interventional surgery.Accurate geometric modeling of blood vessels is essential for studying hemodynamics,which has evolved from CAD-based simulations to non-invasive medical imaging.However,complex blood vessel morphology and numerous branches can lead to missing or inaccurate construction in the reconstruction process,particularly for larger vessels,requiring more time.Thesis addresses these challenges by exploring the use of implicit modeling to quickly and accurately construct blood flow analysis models,with significant contributions including:(1)The present thesis enhances the implicit reconstruction algorithm of RBF with ellipsoidal constraints.Firstly,the concept of “error-driven”methodology is introduced to mitigate the influence of noise on the reconstructed results and thus,augment the precision of the reconstruction model.Both visual and geometric evaluations demonstrate that the proposed algorithm improves the accuracy of the reconstruction up to 88.4%,which is significantly higher than that achieved by the original algorithm.Secondly,a novel lightweight fusion strategy is presented to reduce the computational complexity of the algorithm and enhance its efficiency.Additionally,parallel computing optimization is implemented to accelerate the scientific computation of the algorithm.Experimental results demonstrate that the improved algorithm’s reconstruction speed is enhanced by two orders of magnitude compared to the original algorithm,and it attains a good trade-off between reconstruction accuracy and speed.(2)Hemodynamic preprocessing is a critical step in ensuring the reliability and accuracy of hemodynamic analysis.Its primary task is to generate precise vascular geometric models and boundary conditions that support accurate computations for hemodynamic analysis.However,current boundary demarcation procedures heavily rely on manual operations,which are time-consuming,inefficient,and constitute a bottleneck for the advancement of hemodynamic research.To address this issue,thesis proposes an automatic boundary division algorithm based on the advantages of the ”divide and conquer” reconstruction strategy,which aims to achieve automatic partitioning and naming of the inlet and outlet boundaries of the vascular model,thereby providing more reliable and accurate boundary conditions.Experimental results show that the automatic partitioning algorithm can accurately and rapidly partition the inlet and outlet boundaries of the vascular model,resulting in a higher-quality mesh than the commonly used interactive partitioning method.(3)Through the integration of the algorithms discussed in thesis,a rapid vascular modeling visualization system has been designed and developed.The system is characterized by its ease of operation and high efficiency,providing medical researchers with a comprehensive vascular modeling scheme that encompasses the entire process from medical image import to vascular model reconstruction.Additionally,the system allows for a streamlined one-stop operation from vascular modeling to boundary division,while also offering a simple yet effective blood flow simulation analysis function.