膝關節軟骨分析之電腦輔助診斷系統的研發

Development of Computer-Aided Diagnosis System on Knee Cartilage Analysis

黃竹健(Chu-Jiang Huang)

指導教授 ： 蘇振隆

中原大學/工學院/生物醫學工程研究所/碩士(2010年)

https://doi.org/10.6840/cycu201001072

摘要

退化性關節炎(Osteoarthritis, OA)是世界上高齡人口常見疾病。在診斷中軟骨體積為重要之診斷依據。核磁共振造影(Magnetic Resonance Imaging, MRI)可提供體內組織之高對比度影像，且利用不同造影條件之磁振造影影像可獲得良好對比度之膝關節影像。因此本研究利用核磁共振影像開發出一套對於膝軟骨影像分割並提供軟骨體積計算之電腦輔助診斷系統。
本研究利用MRI 之3D GRE T2*影像作為輔助診斷系統之研究材料。於影像前處理利用中值濾波先行去除雜訊後，再使用Gamma轉換以增強高灰階度的軟骨及周邊組織之分界。接著利用兩不同之閥值為區域成長之限制參數，對於周邊組織先進行去除的動作以避免影像分割造成誤判，並進而獲得軟骨區域。本研究並以兩不同灰階之球體建構出弦月形之區域之假體，以程式繪製其切面影像來模擬膝關節軟骨，針對不同需求加入對應條件進行系統能力驗證。
本研究對於雜訊容忍、影像分割及體積計算方面進行假體驗證。開發之系統對包含24%以下影像雜訊皆能有效去除，且影響體積變異在0.5%以下。在體積計算中對於影像假體球半徑為50-100pixels，而切面間距分別為1.0及1.5mm之影像作驗證，其誤差皆於0.05mm3以下。在實際影像測試中則使用來自兩不同醫院之四組膝關節影像進行測試，其初步結果顯示本系統能有良好之分割且能獲得膝關節軟骨體積數據
本研究所開發之系統結合膝關節軟骨分割及體積計算，經由假體驗證可獲得與假體體積接近之數據，本系統提供使用者能有於判讀膝關節軟骨影像時的分割參考，及計算獲得軟骨體積數據。

關鍵字

膝關節軟骨 ； 核磁共振造影 ； 膝軟骨分割 ； 軟骨體積 ； 區域成長

並列摘要

Osteoarthritis (OA) is one of elder population general disease in the world. The most important diagnosis reference for OA is cartilage volume. Magnetic Resonance Image (MRI) provided high resolution images by using different imaging factor which could obtain specialized tissue contrast knee MRI image. Therefore, this study developed a computer aided diagnosis system to segment articular cartilage and evaluated cartilage volume based on MRI image.
3D T2* GRE MRI knee images were used to be our study material for aided diagnosis system. Because some boundary of the images enhanced cartilage tissue signal is not obvious, median filter was used to reduce noise, and then high gray level contrast was enhanced by gamma transform to make the boundary obvious. In first segmenting step, the tissue surround cartilage was erased to avoid wrong diagnosis of image segmentation, and then segment cartilage region could be got by using region growing. In this system, we presented two spheres with different gray level in virtual space, those two spheres were constructed a crescent region to simulate cartilage region. Furthermore, ability of this system was validated and volume difference was evaluated by adding different factors that might affect results.
In phantom validation, our system can removed less than 24% noise effectively, enhanced the contrast in high gray level and made volume difference below 0.5%. In volume evaluation, spheres radius 50 to 100 pixels with slice location 1.0mm and 1.5mm were used, the results showed that errors were under 0.05mm3. Four sets knee MRI image from two different hospitals were applied to test for real imaging. Preliminary results from this study proved that our system could segment cartilage actually and gain cartilage volume data.
By combining developed computer aided diagnosis system by this study with knee cartilage segmentation and cartilage volume evaluation, results of system were closed to phantom volume in phantom validation. Moreover, this system could assist user to diagnose knee cartilage region and calculate cartilage volume.

並列關鍵字

Magnetic Resonance Imaging ； Knee Cartilage ； Cartilage Volume. ； Region Growing

參考文獻

[25]李正輝，超順磁性氧化鐵造影劑於肝臟擴散加權影像的應用，中原大學醫學工程研究所碩士論文，中壢，民國九十八年

[4] C.G. Peterfy, C.F. van Dijke, D.L. Janzen, C.C. Gluer , R. Namba , and S. Majumdar,”Quantification of articular cartilage in the knee with pulsed saturation transfer subtraction and fat-suppressed MR imaging: optimization and validation,” Radiology, Vol.192, pp.485-491, 1994.

[5] M.A. Piplani, D.G. Disler , T.R. McCauley, T.J. Holmes, and J.P .Cousins, ” Articular cartilage volume in the knee: semiautomated determination from three-dimensional reformations of MR images,” Radiology , Vol.198 , pp. 855-859, 1996.

[6] Z.A. Cohen, D.M. McCarthy, S.D. Kwak, P. Legrand, F. Fogarasi, and E.J. Ciaccio, “Knee cartilage topography, thickness, and contact areas from MRI: in-vitro calibration and in-vivo measurements,” Osteoarthritis Cartilage, Vol.7 ,pp. 95-109 , 1999.

[7] C. Kauffmann, P. Gravel, B. Godbout, A. Gravel, G. Beaudoin, and J.P. Raynauld, ”Computer-aided method for quantification of cartilage thickness and volume changes using MRI: validation study using a synthetic model,” IEEE Trans. Biomed. Eng., Vol.50 , pp.978-988 , 2003