CT影像對於大多數骨科手術很有用，因為他們可以清楚的顯示骨結構的邊界，特別是對於骨折。 為了讓數字化的骨骼模型重新定位回到原本的位置，通常使用電腦輔助術前規劃模擬中的骨骼復位工具。 骨骼復位模擬的困難之處在於不容易去確認復位結果的精準度，因為每個破裂骨頭都有著不規則的形狀，因此其數字化模型的邊界可能不夠精準。 根據初步結果來看，從CT影像直接獲得的3D列印骨骼沒有辦法輕易和正確的組裝。 因此，為了減少兩個3D列印骨骼在組裝時的干涉表面，可以使用重疊網格檢測與重疊網格消除來去除3D模型上的重疊網格。 在此研究中，為了提高3D列印骨骼的組裝品質，我們提出了一個重疊網格檢測和消除的演算法來改善3D骨骼模型。 為了檢驗重疊網格檢測和消除演算法的可行性，我們使用了5個臨床病例的骨折模型。 在使用重疊網格消除之前，應該要先對骨折模型進行平滑化和去除雜亂網格。 計算了骨折模型在平滑化和去除雜亂網格前後的方均根誤差。 基於模擬結果可以得知，此演算法能夠獲得一個組裝品質良好的3D列印骨骼模型，其中重疊網格可以大幅減少，特別是在破裂區域。 臨床病例的方均根誤差範圍在0.149~0.255mm之間。 兩個3D列印骨骼在去除重疊網格後的誤差間隙範圍為0.86~1.11mm。 此演算法可以得到很好的結果，所有修改後的模型都可以很容易的組裝和拆卸。;The CT images are useful for most kinds of orthopedic surgery because they can reveal the 3D bone structure clearly, especially for bone fracture. In order to relocate the digitized bone model to their original position, bone repositioning tool of the computer-aided preoperative planning simulation is employed virtually. The difficulty in simulated bone repositioning is that it is not easy to check the accuracy of the repositioning result because each broken bone has an irregular shape and the boundary of each digitized model may not be accurate enough. According to the preliminary result, the 3D-printed bones obtained directly from CT images cannot be assembled easily and correctly. Hence, the overlapping detection and overlapping removal feature are applied to remove the overlapping meshes on 3D model in order to reduce the interference surface on the assembly of two 3D-printed bones. In this study, we propose an overlapping detection and removal algorithm to improve 3D bone model in order to increase the assembled quality of two 3D-printed bones. To evaluate the performance of overlapping detection and removal algorithm, five cases of patient clinical fracture bone models are proposed. The bone fracture models are performed smoothing and noise reducing before applied to overlapping removal process. The RMS error of the resulting bone fracture before and after smoothing and noise reducing is calculated. Base on the simulation result, the proposed algorithm is able to obtain a well-made assembled 3D-printed bone, where the overlapping surface can be minimized especially on broken area. The RMS error for entire clinical cases is in a range 0.149~0.255 mm. For the gap error of two 3D-printed bones after overlapping removal process is in a range 0.86~1.11 mm. The proposed algorithm has obtained satisfying result and all modified bone models can be assembled and disassembled easily.
