1 |
Meshless deformable models for medical simulation applications. / CUHK electronic theses & dissertations collectionJanuary 2013 (has links)
在這篇論文中,我們提出了在醫學模擬應用的血管或傷口上作相互作用的粒子血流變模型框架。通過平滑粒子流體動力學(SPH)制定的非牛頓流體,進行了血液流變學的模擬。通過建模血管壁結構虛擬粒子,流體 - 結構相互作用(FSI)是一個純粹的拉格朗日(Lagrange)顆粒模型進行建模的血管或血液的交互。我們的建議的方法基於純粹的非網格方法,可用於常見的動脈瘤和血管狹窄等病症的建立上。如需模擬開放性傷口在手術部位中發生較大的變形情況時,我們則採用質量 - 彈簧系統進行血顆粒的交互,此交互框架可應用到幾個開放性手術模擬,如骨科或胃鏡檢查為基礎的手術。無論是常見的醫療圖像:如CT血管造影(CTA)、磁共振血管造影(MRA)或基於網格的數據也可以 作為系統輸入的數據。血栓形成與溶解模型也被集成到這個流固耦合框架中。實驗結果證明採用我們建議的粒子互相作用框架在模擬血管中的凝血過程是可行的。受益於簡潔的拉格朗日粒子交互作用模擬,我們的系統可以保持在互動幀速率中。 / 首先,我們在這篇論文中建議把無網格流變模擬框架應用於血管手術的建模中。於非牛頓粘性流動的假設下,我們建立了血液結構的一般模型方程:以平滑粒子流體動力學實現多血粘度模型與低彈性血管壁模型。血流動力學和軟組織都可以於相同的拉格朗日粒子為基礎下模擬。在這個意義上說,通過延伸平滑粒子流體動力學的密度和動量求和不管顆粒的性質下,本論文提出了一個有效的流體 - 固體交互作用模型。該模型是特別有利於整合多種類型的介質(包括固體或液體)的。在這方面,我們進一步提出了一個與流體相關的血塊凝集溶解模型,可以適用於許多不同種類的醫學模擬:例如血栓栓塞。 / 其次,本論文亦提出了如何基於粒子的血液建模框架的前提下,擴展到大變形的軟組織互動。我們是以耦合雙向階段性質量 - 彈簧系統與固體顆粒,去代替無網格粒子固體的建模,用以維持真正人體組織的高保真度,此方法可以實現類似軟組織的皮膚或真皮的交互式模擬。而耦合血顆粒與平滑粒子方面,則由一個聰明的碰撞模塊處理,使得利用模擬皮膚表面之上,可以模擬出真實的表皮出血現象。該模型的動態計算進一步以物理學處理單元加速;而渲染的模型則是通過一個強大的圖形處理單元為基礎的立方體運行(marching cubes)的方法來實現。該模型已應用於全身血液管理培訓中。 / In this thesis, we propose particle-based rheological modeling frameworks for blood-vessel and blood-wound interaction in medical simulation applications. The effect of blood rheology has been simulated through a smoothed particle hydrodynamics (SPH) formulation of non-Newtonian flow. By modeling the vessel wall structure as virtual particles, a pure Lagrange particle formulation for fluid-structure interaction (FSI) is proposed for modeling the blood-vessel or blood-device interaction. Our proposed framework synthesizes common vascular complication sites such as stenosis and aneurysm based on purely mesh-less approach. For larger deformation situations happened in surgical sites such as open wound, we adopt a mass-spring system to interact with the blood particles; the blood-wound interaction framework can be applied to several open surgery simulations such as orthopedics or endoscopy-based interventions. Input of the data can be obtained from either common medical modalities like computed tomographic angiography (CTA), magnetic resonance angiography (MRA) or processing mesh-based data. A thrombus (clot) formation-dissolution model is also integrated into this fluid-solid interaction framework. Results have demonstrated the feasibility of employing our proposed particle framework in simulating blood-vessel interaction in the clotting process which is essential to vascular procedure simulations. Having benefited from the elegant formulation of Lagrangian particle interaction; the simulation can be maintained at interactive frame-rates. / In this thesis, first, a meshless rheological modeling framework for medical simulation of vascular procedures is proposed. Instead of assuming a Newtonian non-viscous flow, we have built our model based on the general constitutive equation of blood. The multi-regime of viscosity in blood model with a hypoelastic model of vessel wall has been realized under a SPH formulation. The hemodynamic and the soft tissue can all be simulated under the same Lagrangian particle-based formulation. In this sense, an efficient formulation of fluid-solid interaction is proposed through extending SPH summations of density and momentum regardless the nature of particles. This model is particularly beneficial to the integration of multiple types of media (including solids or fluids). With this regards, we further propose a flow related clot aggregation-dissolution model which can be applicable to many different kinds of medical simulation e.g. thrombo-embolization. / Second, the proposed particle-based blood modeling framework has been extended to interact with large deformation of soft tissue. Instead of modeling the solid as meshless particles, a bi-phasic mass-spring system is coupled with solid particles so that an interactive simulation of skin or dermis like soft tissue can be realized with high fidelity to real human tissue. To couple with the SPH formulation of blood particles, a smart collision handling module is exploited so that a realistic bleeding simulation on top of the skin surface can be created. The dynamic computation of this model is further accelerated by the physics processing unit; while the rendering of the model is realized through a robust graphics processing unit based marching cube approach. The proposed model has been applied to provide general blood management training. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Chui, Yim Pan. / Thesis (Ph.D.)--Chinese University of Hong Kong, 2013. / Includes bibliographical references (leaves 98-113). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstracts also in Chinese. / Abstract --- p.ii / Chapter 1 --- Introduction --- p.1 / Chapter 2 --- Related works on physically based fluid-structure models --- p.7 / Chapter 2.1 --- Eulerian grid-based methods --- p.8 / Chapter 2.2 --- Lagrangian grid-based methods --- p.9 / Chapter 2.3 --- Lagrangian meshfree methods --- p.11 / Chapter 2.4 --- Fluid-structure interaction (FSI) --- p.12 / Chapter 2.5 --- Endovascular simulation --- p.14 / Chapter 2.6 --- Overview of Our Model --- p.15 / Chapter 3 --- Meshless blood-clot interaction --- p.16 / Chapter 3.1 --- Basic equations of fluid dynamics --- p.17 / Chapter 3.2 --- SPH basics --- p.18 / Chapter 3.3 --- SPH Rheological hemodynamics of blood --- p.20 / Chapter 3.4 --- SPH modeling of the hypoelastic vessel --- p.26 / Chapter 3.5 --- Fluid-solid interaction model --- p.28 / Chapter 3.6 --- Flow-related clot aggregation-dissolution model --- p.33 / Chapter 3.7 --- Time integration --- p.36 / Chapter 3.8 --- Hardware-friendly formulation --- p.37 / Chapter 3.9 --- Results --- p.39 / Chapter 3.9.1 --- Classical Dam-break problem --- p.41 / Chapter 3.9.2 --- Poiseuille flow --- p.43 / Chapter 3.9.3 --- Couette flow --- p.45 / Chapter 3.9.4 --- Mechanical model with material strength --- p.47 / Chapter 3.9.5 --- Hemoelastic feedback system --- p.49 / Chapter 3.9.6 --- Clotting in a stenosed vessel --- p.52 / Chapter 3.9.7 --- Timing results --- p.53 / Chapter 4 --- Meshless modeling of thrombo-embolization --- p.55 / Chapter 4.1 --- Modeling framework for thrombus formation within blood vessel . --- p.60 / Chapter 4.2 --- Geometric Modeling and Flow Simulation --- p.61 / Chapter 4.2.1 --- Data processing on vascular data --- p.61 / Chapter 4.2.2 --- Blood-Vessel particle distribution --- p.62 / Chapter 4.2.3 --- Blood-structure Interaction --- p.65 / Chapter 4.3 --- Visualization and Thrombosis Simulation --- p.66 / Chapter 4.3.1 --- Flow Visualization --- p.66 / Chapter 4.3.2 --- Thromb-Embolization Simulation --- p.68 / Chapter 4.4 --- Conclusion and discussion --- p.72 / Chapter 5 --- Lagrangian modeling framework for bleeding simulation --- p.76 / Chapter 5.1 --- SPH-based bleeding model --- p.78 / Chapter 5.2 --- Biphasic Soft-tissue deformation --- p.79 / Chapter 5.3 --- Interaction between blood and soft tissue --- p.83 / Chapter 5.4 --- Integrated training for blood management --- p.87 / Chapter 6 --- Discussion and Conclusion --- p.93 / Bibliography --- p.98
|
Page generated in 0.1014 seconds