醫學數據可視化及軟組織建模
軟件工程 黃鵬飛 指導教師 顧力栩
【摘要】 自百年前倫琴發現X 射線以來,醫學圖像可視化經歷了革命性的發展。醫學知識、物
理概念、數學方法和計算機科學理論有機地融合在醫學圖像處理之中。醫學虛擬現實是計算機科學
家孜孜不倦的追求。近年來,人體器官軟組織的三維建模更是計算機圖形學的重要研究課題。而軟
組織建模是實時手術虛擬和人機交互的基礎。
本文提出并討論了質點彈簧算法,它將形變物體表面建模成質點彈簧網絡,其計算量小,但是
算法精度不高。然后介紹了中心線描繪算法,這種新型算法通過中心線信息對表面進行重繪,從而
模擬形變,常被用作圖像的分割、配準。雖然它的實時性不是很理想,但是它重繪的思路在軟組織
建模中頗有價值。于是,筆者結合了簡化了的中心線描繪模型與質點彈簧模型,即在物體主要受力
形變區域采取質點彈簧模型,在其它區域采取中心線描繪模型。這樣既保證了局部形變的實時性,
同時也保證了全局形變的逼真性。由于醫學數據可視化是軟組織三維建模的前提,本文還將介紹可
視化相關的三維重建、圖像分割算法等問題。除了核心的三維建模部分,本文亦將介紹人機交互技
術、力反饋、碰撞檢測、實時性與精確性等手術虛擬相關問題。
【關鍵詞】 醫學圖像 網格 形變 質點彈簧 中心線描繪
【Abstract】 Since 100 years ago, when Roentgen discovered x-rays, medical image visualization has
revolutionarily developed. Medicine knowledge, physics concept, mathematics method and computer
science have interdisciplinarily worked together on the medical image processing. It is inherent for the
computer scientists to simulate the medical reality. In recent years, the 3D modeling of human organic
soft tissue has been an important research subject in computer graphics. And the soft tissue modeling is
the basis for the real-time surgery simulation and human computer interaction.
In this paper, we propose and discuss the mass-spring algorithm. It models the surface of deformable
object as the network of masses and springs. Its computational complexity is low but it is limited in
accuracy. And we propose a latest algorithm, medial representation. It redraws the surface based on the
centerline information to model the deformation and it is often used for image segemantation and
registration. Although its real-time property is not ideal enough, we can take advantage of its redrawing
strategy in soft tissue modeling. Therefore, we combine the mass-spring and medial representation
models so that we can achieve the real-time performance of local deformation and, at the same time, the
accuracy of entire deformation. Because medical data visualization is the precondition for soft tissue
modeling, we introduce the relating contents of 3D reconstruction and image segmentation. Besides 3D
modeling, which is the core part of surgery simulation, we also introduce the simulation’s relating issues
including human computer ineraction technique, force feeback, collision detection, accuracy and
real-time performance.
【Keywords】 Medical Image Mesh Deformation Mass-Spring Medial Representation
醫學數據可視化及軟組織建模
第 2 頁 共 63 頁
目錄
醫學數據可視化及軟組織建模 1
1 引言 3
1.1 研究背景····························································································· 3
1.2 研究目標····························································································· 3
1.3 內容組織····························································································· 3
2 醫學數據可視化技術 3
2.1 概述····································································································· 3
2.2 醫學數據三維重建·············································································· 3
2.2.1 醫學圖像數據格式······························································· 3
2.2.2 表面繪制技術······································································· 3
2.2.3 體繪制技術··········································································· 3
2.3 醫學圖像分割······················································································ 3
2.3.1 形態學方法··········································································· 3
2.3.2 快速步進法··········································································· 3
2.4 醫學可視化平臺開發實現·································································· 3
2.4.1 VTK 軟件開發包···································································· 3
2.4.2 ITK 軟件開發包···································································· 3
2.4.3 Atamai 手術導航軟件開發包·············································· 3
2.4.4 基于Atamai 的醫學可視化平臺········································· 3
3 三維物體表面網格化技術 3
3.1 概述····································································································· 3
3.2 VTK 數據格式···················································································· 3
3.3 行進立方體算法·················································································· 3
3.4 氣球算法····························································································· 3
4 人體軟組織三維建模 3
4.1 概述····································································································· 3
4.2 質點彈簧算法······················································································ 3
4.2.1 質點彈簧拓撲結構······························································· 3
4.2.2 質點彈簧系統原理······························································· 3
4.2.3 表面模型··············································································· 3
4.2.4 體模型··················································································· 3
醫學數據可視化及軟組織建模
第 3 頁 共 63 頁
4.3 中心線描繪算法·················································································· 3
4.3.1 中心線幾何結構··································································· 3
4.3.2 根據中心線繪制表面··························································· 3
4.4 質點彈簧算法與中心線描繪算法的結合·········································· 3
4.5 三維建模應用程序開發實現······························································ 3
5 醫學虛擬現實 3
5.1 概述····································································································· 3
5.2 交互作用技術······················································································ 3
5.2.1 三維圖形坐標選取······························································· 3
5.2.2 碰撞檢測··············································································· 3
5.2.3 力反饋··················································································· 3
5.3 實時性與真實性·················································································· 3
結論與展望 3
謝辭 3
主要參考文獻 3
附錄 3
醫學數據可視化及軟組織建模
第 4 頁 共 63 頁
1 引言
1.1 研究背景
醫學圖像處理與分析技術涉及很多學科,包括計算機圖像處理、計算機圖形學、
模式識別、虛擬現實以及醫學相關知識等。
關于醫學圖像的發展,是德國物理學家倫琴開創了人體圖像的先河。1895 年11
月8 日,倫琴在研究時無意中發現了X 射線, |
|