為達模擬骨骼肌肉手術與診斷目的之容積操作與表面容積化演算法

Volume manipulations and voxelization algorithms in applications of diagnoses and surgical simulations of musculoskeletal system

周賢彬(Shyan-Bin Jou)

指導教授 ： 蔡明達

中原大學/工學院/電子工程研究所/博士(2002年)

https://doi.org/10.6840/cycu200200614

摘要

在本論文中，我們舉出手術模擬的重要性，也調查了手術模擬相關的技巧，了解了因為容積可以解析到解剖結構內部資訊，所以準確度較表面模型為高。然後我們調查了容積模型學及操作容積實體、表面的方法，探討了這些方法於手術模擬的不足。
所以我們提出一個新的容積資料結構，及新的在容積內描述表面的演算法。也提出了改進操作容積實體的效率的演算法。而利用這些演算法及資料結構，我們開發了可以模擬操作容積化實體的功能，特別是應用在模擬骨骼肌肉手術。這些功能包括碰撞偵測、實體間的布林運算、切割一個解剖骨塊(即容積中的實體)、以及辨識，移動骨塊和癒合兩個骨塊等。
這篇論文以數個手術模擬測試以上提到的功能。首先是一個下顎骨突出矯正手術：這個實例中顯示了，複雜的幾何和拓樸的改變可以藉由以上提到的功能，得到良好的模擬。在第二個例子中，這些功能結合一個利用三唯幾何限制的分析與評估方法，用以決定選擇最佳的商業上人工膝關節的尺寸和最佳的植入位置。這個分析與評估方法所獲致的結果，被用來當作人工膝關節置換術的模擬手術的重要依據。第三個例子是結合診斷椎間盤突出病例的方法，這個方法利用了中心式貝氏曲線的特性來近似斷層片中的椎間盤邊界。我們求出椎間盤邊界的中心，並以相鄰像素半徑的變化率得出特徵點，再以特徵點的所在位置，決定出椎間盤突出的等級和算出突出物的體積。這個椎間盤突出的診斷方法，可以有效的偵測出可能的突出位置與大小，提供後續手術模擬中，所需的切割位置與辨識的資訊。
第一個實驗結果顯示，我們所提出的資料結構演算法和模擬的功能，完全符合骨骼肌肉手術模擬系統的需求。第二及第三個實驗結果顯示，基於容積操作的模擬方法可以藉由相關演算法的輔助，進一步擴充到診斷、驗證以及提供可能的處置對策。

關鍵字

手術模擬 ； 表面容積素化 ； 容積顯像

並列摘要

In this thesis, we reveal the importance of surgical simulation, survey the related simulation techniques and realize that volume model is more accurate than others as it can resolve the inner information of anatomic structure. Then, we survey varied volume models with extended data structures and methodologies for manipulating solids, surfaces and find out the insufficiencies of them.
We propose a new volume data structure and algorithms for surface representation in a volume, also promote the algorithms that can efficiently manipulate volumetric solids. With these algorithms and data structure, we develop some simulation functions to manipulate volumetric solids and especially simulate musculoskeletal surgery, including collision detection and Boolean operations between solids, tear an anatomic structure (a solid), recognize, reposition the structure and fuse two separate structure, etc.
This study testifies the functions with several surgical examples. First, a corrective osteotomy for correcting a maxilla extrusion is demonstrated. This example shows a complex musculoskeletal surgery with complicated topologic and geometric changes can be well simulated by our functions. In the second example, the functions combine with an algorithm of evaluating and analyzing 3D geometric constraints for determining commercial knee prosthesis. The evaluation and analysis results are used as parameters for simulating the knee arthroplasty. In the last example, functions combining HIVD (Herniated Inter-Vertebral Disc) detection method is demonstrated. This method accommodates the radial B-spline to approximate the spline curve composed by disc boundary on slice. We calculate the center of inter-vertebral disc, determine the concave and convex features on the disc based on the change rate of radii and determine the HIVD degree based on the feature characteristics. This method provides effective information for the followed-up HIVD surgical simulation.
The first experimental result shows that proposed data structure, algorithms and simulation functions completely fulfill the requirements of simulating the musculoskeletal surgery. The second and last experimental results show the simulation method based on volumetric manipulation can be extended to a diagnosis, verification and treatment tool by combining the related algorithms.

並列關鍵字

volume visualization ； surface voxelization ； surgery simulation

參考文獻

1. Albert, T.A. and Slaaf, D.W., 1995, “A Rapid Regional Filling Technique for Complex Binary Images”, Computer & Graphics, Vol.9 no.4, pp.541-549.

2. Anne, G.O., 1994, Diagnostic Neuroradiology. Mosby, pp. 836-872.

3. Anouchi, Y.S., Whiteside, L.A., Kaiser, A.D. and Milliano, M.T., 1993, “The Effects of Axial Rotational Alignment of the Femoral Component on Knee Stability and Patellar Tracking in Total Knee Arthroplasty Demonstrated on Autopsy Specimens”, Clin. Orthop., Vol.287, pp.170-177.

4. Berger, R.A., Rubash, H.E. and Seel, M.J., 1993, “Determining the Rotational Alignment of the Femoral Component in Total Knee Arthroplasty Using the Epicondylar, Axis”, Clin. Orthop., Vol.173, pp.40-47.

5. Baumrind, S., Moffit, F. and Curry, S., 1983, “The Geometry of Three-Dimensional Measurement from Paired Coplanar X-Ray Images”, Am. J. Orthod., Vol.84, pp.313-326.