在人們的日常生活當中，經常會使用到手的動作，尤其大拇指的使用率更高達50%。也因為這麼高的使用頻率，使得大拇指關節容易發生關節發炎、退化等病變症狀，尤其是大拇指基關節(Trapeziometacarpal Joint, TMC Joint)。此關節是大拇指最重要的一部分，它具有不同於其他手指關節的構造，使得它能夠進行多維的運動，例如彎曲伸直及外展內收，這是其他關節無法同時做到的，所以我們對它很感興趣。
為了研究與分析TMC關節的運動狀況，我們選用了電腦斷層掃描(CT)。CT影像對於人體高密度的結構具有較佳的解析度，因此經常用於骨關節、脊椎等部位的拍攝。我們利用CT針對大拇指關節運動的特定姿勢成像，然後分割並重建在各運動狀態下的拇指關節，藉以觀察分析關節的運動狀況以及量測關節面的生物力學參數。
在本篇的方法中，共有兩個重點：平均模型的建立，以及姿勢間的平均轉換關係。在平均模型的建立中，我們將藉由多個不同人的骨塊模型，建立一個統計性的平均模型。姿勢間的平均轉換，則是藉由先前的研究結果，以統計的方式計算出每個姿勢之間的平均運動參數，包含三個位移量及三個旋轉角度。完成以上兩個資訊的建立後，我們便藉由平均統計模型，以主動輪廓模型的方法快速地分割初始姿勢的骨塊模型，然後再將初始姿勢模型套用平均的運動參數，並以主動型狀模型的方法分割其餘所有姿勢的骨塊模型，完成一套自動化的分割。
整體來說，本篇研究包含骨塊自動分割技術、三維視覺模型重構、關節面參數量測、以及關節運動動態視覺化。本篇研究的技術與成果能夠將關節的狀況量化分析，藉以達到預防或是減緩病變的狀況，也能夠對大拇指基關節的生物力學探討有所幫助。

In human daily life, hands are often used in many actions, especially the thumb. It is used in up to 50% actions. Because of the frequently-used, makes it easily suffered from joint inflammation and arthrosis, especially Trapeziometacarpal Joint(TMC Joint). TMC joint is the most important joint of thumb. Its structure is totally different form any other finger joint, so it allows to move in multi-dimension, such as flexion, extension, adduction, abduction. These motion cannot be done at the same time by other joint, so we are really interested in TMC joint.
In order to investigate and analyze the motion of TMC joint, we choose computed tomography(CT). CT images have high anatomical resolution at the high density structure of human body, so it is often used in bone joint and spine scanning. In our experiment, we used CT scanning muti-posture of thumb motion for each subject.
There are two main points in our method, including mean model construction and average transformation between poses. First, we construct the statistical mean model by several intra case’ bone model. And then we calculate average transformation from previous studies. After acquiring these two information, we can segment the first pose’s bone model by deforming statistical mean model with Active Shape Model method to fit the bone region in its 3-D CT images. After the first pose's bone model is constructed, we can also obtain other poses’ initial bone model by applying the average transform to the first pose’s model, and then deform the initial model with Active Contour Model method to fit their own pose's 3-D images, finishing the segmentation. Overall, there are four points in our thesis, including automatically segment bony structure, 3-D visualize model construction, evaluating bio-mechanics of joint contact surface, and joint motion visualization. The technique and result of this thesis is helping for preventing from arthrosis and TMC joint bio-mechanics study.

[1] J. H. Youngleson, “The Management of the Contracted First Web Space,” S. A. Medical Journal, vol. 39, no. 32, pp. 716-719, 1965.
[2] http://www.3579999.com/view/22630.html
[3] Armstrong AL, Hunter JB, Davis TR. The prevalence of degenerative arthritis of the base of the thumb in post-menopausal women. J Hand Surg Br. 1994;19:340–341. doi: 10.1016/0266-7681(94)90085-X.
[4] Dahaghin S, Bierma-Zeinstra S, Ginai A, et al. Prevalence and pattern of radiographic hand osteoarthritis and association with pain and disability (the Rotterdam study) Ann Rheum Dis. 2005;64:682–687. doi: 10.1136/ard.2004.023564.
[5] http://ovrt.nist.gov/projects/vrml/h-anim/jointInfo.html
[6] J. M. Moran, J. H. Hemann, and A. S. Greenwald, “Finger Joint Contact Areas and Pressures,” Journal of Orthopaedic Research, vol. 3, no. 1, pp. 49-55, 1985.
[7] L. J. Soslowsky, E. L. Flatow, L. U. Bigliani, R. J. Pawluk, G. A. Ateshian, and V. C. Mow, “Quantitation of In Situ Contact Areas at the Glenohumeral Joint: A
[8] G. Windisch, B. Odehnal, R. Reimann, F. Anderhuber, H. Stachel, “Contact Areas of the Tibiotalar Joint,” Journal of Orthopaedic Research, pp. 1481-1487, Nov. 2007.
[9] G. A. Ateshian, J. W. Ark, M. P. Rosenwasser, R. J. Pawluk, L. J. Soslowsky, and V. C. Mow, “Contact Areas in the Thumb Carpometacarpal Joint,” Journal of Orthopaedic Research, pp. 450-458, 1995.
[10] Kuo, Li‐Chieh, et al. "A quantitative method to measure maximal workspace of the trapeziometacarpal joint—normal model development." Journal of orthopaedic research 22.3 (2004): 600-606.
[11] L.C. Kuo, W.P. Cooney III, K.N. An, K.Y. Lai, S.M. Wang, and F.C. Su, “Effects of Age and Gender on the Movement Workspace of the Trapeziometacarpal Joint,” Proceedings of the Institution of Mechanical Engineers, Part H: Journal of Engineering in Medicine, vol. 223, no. 2, pp. 133-142, Feb. 2009.
[12] Su, Fong-Chin, et al. "In vivo analysis of trapeziometacarpal joint arthrokinematics during multi-directional thumb motions." Clinical Biomechanics 29.9 (2014): 1009-1015.
[13] Goto, Akira, et al. "In Vivo Pilot Study Evaluating the Thumb Carpometacarpal Joint During Circumduction." Clinical Orthopaedics and Related Research®472.4 (2014): 1106-1113.
[14] Pal, Nikhil R., and Sankar K. Pal. "A review on image segmentation techniques." Pattern recognition 26.9 (1993): 1277-1294.
[15] T. F. Chan and L. A. Vese, ”Active Contours Without Edges,” IEEE Trans. on Image Processing, vol. 10, no. 2, pp. 266-277, Feb. 2001.
[16] M. Kass, A. Witkin, and D. Terzoulos, “Snake: Active contour models,” International Journal of Computer Vision, vol. 1, no. 4, pp. 321-331, 1987.
[17] Cootes, Timothy F., et al. "Active shape models-their training and application."Computer vision and image understanding 61.1 (1995): 38-59.
[18] Powell, Michael JD. "An efficient method for finding the minimum of a function of several variables without calculating derivatives." The computer journal 7.2 (1964): 155-162.
[19] Guan-Po Chen, "In Vivo Study of Trapeziometacarpal Joint Gliding and Contact using 3D Images," Master Thesis, July. 2011.
[20] Weickert, Joachim. Anisotropic diffusion in image processing. Vol. 1. Stuttgart: Teubner, 1998.
[21] Ng, Hui-Fuang. "Automatic thresholding for defect detection." Pattern recognition letters 27.14 (2006): 1644-1649.
[22] Gerig, G., Kubler, O., Kikinis, R., & Jolesz, F. A. (1992). Nonlinear anisotropic filtering of MRI data. Medical Imaging, IEEE Transactions on, 11(2), pp.221-232.
[23] van den Boomgaard, Rein. "Algorithms for Non-Linear Diffusion: Matlab in a Literate Programming Style." Intelligent Sensory Information Systems, University of Amsterdam., The Netherlands.
[24] W. E. Lorensen and H. E. Cline, “Marching Cubes: A High Resolution 3D Surface Construction Algorithm,” Computer Graphics, vol. 21, no. 4, pp. 163-169, July. 1987.
[25] MeshLab Visual Computing Lab - ISTI - CNR http://meshlab.sourceforge.net/ @misc{Meshlab author = {Visual Computing Lab ISTI - CNR}, title = {MeshLab}, note = {http://meshlab.sourceforge.net/} }