计算机辅助泪道手术规划技术的研究

发布时间：2019-04-08 10:19

【摘要】：泪道位于鼻眶筛区,是泪液的排出系统,由泪点、泪小管、泪总管、泪囊和鼻泪管五部分组成,其中鼻泪管是泪液从眼睛排入鼻腔的通道。当鼻眶筛区发生骨折后,会导致鼻泪管受到挤压或骨折,从而使其引流功能出现异常,导致人眼视力下降甚至造成严重的功能障碍。由于该区解剖结构复杂,骨折后可造成多部位、不同类型的阻塞,因此,术前充分了解伤情是成功实施手术、最大限度恢复泪道引流功能的关键。随着科学技术的不断进步,计算机辅助技术在医学领域的应用已经越来越广泛,但是,目前国内外还没有科研单位、商业医学软件公司针对泪道手术提出完整的计算机辅助规划方案,本文通过与上海交通大学附属第九人民医院合作,对计算机辅助的泪道手术规划技术进行了以下几点研究: (1)二维坐标点到三维坐标点的转换。在三维空间中只能在物体的表面标记采样点,为此,本文提出了一种二维坐标点到三维坐标点的转换方法,实现了在三维物体内部标记采样点。 (2)对体数据沿鼻泪管走行方向进行重采样。因为鼻泪管的走行方向与CT扫描方向不垂直,CT切片不能准确反映鼻泪管的结构和形状信息,本文利用二维坐标点到三维坐标点的转换关系,计算求得鼻泪管内部的中轴线,然后对鼻泪管附近的体数据沿垂直于中轴线的方向进行重采样,得到新的鼻泪管二维切片序列。 (3)鼻泪管和泪囊的分割与重建。得到沿鼻泪管走向的重采样二维切片序列之后,本文使用主动轮廓模型算法对切片序列进行了分割,然后分别使用基于轮廓线的表面重建和基于体素的表面重建完成了正常鼻泪管以及泪囊的三维重建。对于骨折的鼻泪管,由于其收敛轮廓是不封闭的曲线,本文分别采用Snake模型和人工交互式方法完成了鼻泪管的二维分割,然后利用最短对角线法由轮廓线重建鼻泪管的三维模型,同时,针对传统最短对角线法在连接三角片时容易形成钝角三角形和误差不断累积的情况,本文对其做了改进。 (4)自动分割方法的提出。针对主动轮廓模型分割方法交互操作过多的不足,本文提出了一种自动分割方法,即利用相邻切片间的连续性,将上一个切片的收敛轮廓作为下一个切片的初始轮廓,来完成切片序列的自动分割。 (5)计算机辅助分析和测量。完成泪道的三维重建后,可以清晰显示泪道内外的解剖结构、鼻泪管骨折情况、泪囊移位距离,在此基础上,实现了可视化立体测量、标记和正常与骨折鼻泪管的镜像对称等。本文提出的泪道辅助规划技术将CT成像技术和图像分割技术以及科学计算可视化技术结合起来,可以提高手术的安全性、准确性和微创性,对于辅助医生诊断具有重要意义。
[Abstract]:The lacrimal duct, located in the naso-orbital ethmoid region, is a lacrimal drainage system consisting of lacrimal spots, lacrimal canaliculi, common lacrimal duct, lacrimal sac and nasolacrimal duct. When the naso-orbital ethmoid region is fractured, the nasolacrimal duct will be squeezed or fractured, and the drainage function will be abnormal. The visual acuity of the human eye will be decreased and even serious dysfunction will be caused by the fracture of the nasolacrimal duct. Because the anatomical structure of the area is complex, it can cause multiple sites and different types of obstruction after fracture. Therefore, the key to successful operation and maximum recovery of lacrimal duct drainage function is to fully understand the injury before operation and to restore the drainage function of lacrimal duct to the maximum extent. With the continuous progress of science and technology, the application of computer-aided technology in the field of medicine has become more and more extensive. However, at present, there are no scientific research institutes at home and abroad. The Commercial Medical Software Company has put forward a complete computer-aided planning scheme for lacrimal duct surgery. In this paper, through cooperation with the Ninth people's Hospital affiliated to Shanghai Jiaotong University, The techniques of computer-aided surgery planning for lacrimal passage were studied as follows: (1) the transformation of two-dimensional coordinate points to three-dimensional coordinate points. In three-dimensional space, sampling points can only be marked on the surface of objects. Therefore, a method of transforming two-dimensional coordinate points to three-dimensional coordinate points is proposed in this paper, which can mark sampling points in the interior of three-dimensional objects. (2) resampling the volume data along the direction of nasolacrimal duct. Because the direction of the nasolacrimal canal is not perpendicular to the scanning direction of CT, CT slice can not accurately reflect the structure and shape of the nasolacrimal duct. In this paper, the inner axis of the nasolacrimal duct is calculated by using the transformation relationship between the two-dimensional coordinate points and the three-dimensional coordinate points. Then the volume data near the nasolacrimal duct were resampled along the direction perpendicular to the central axis and a new two-dimensional slice sequence of the nasolacrimal duct was obtained. (3) Segmentation and reconstruction of nasolacrimal duct and lacrimal sac. After obtaining the resampling two-dimensional slice sequence along the nasolacrimal duct, the active contour model algorithm is used to segment the slice sequence. Then three-dimensional reconstruction of normal nasolacrimal duct and lacrimal sac was completed by surface reconstruction based on contour line and surface reconstruction based on voxel. For the fractured nasolacrimal duct, the two-dimensional segmentation of the nasolacrimal duct is accomplished by using the Snake model and the artificial interactive method, because the constringency profile of the nasolacrimal duct is not closed. Then the shortest diagonal method is used to reconstruct the three-dimensional model of the nasolacrimal canal from the contour line. At the same time, the traditional shortest diagonal method can easily form the obtuse triangle and the error accumulates continuously when the triangle is connected, this paper improves it. (4) the automatic segmentation method is proposed. In this paper, an automatic segmentation method is proposed, which takes the convergence contour of the previous slice as the initial contour of the next slice by using the continuity between adjacent slices, aiming at the deficiency of the interactive operation of the active contour model segmentation method. To complete the automatic segmentation of the slice sequence. (5) computer aided analysis and measurement. After three-dimensional reconstruction of lacrimal duct, the anatomic structure inside and outside the lacrimal duct, the fracture of nasolacrimal duct and the distance of lacrimal sac displacement can be clearly displayed. On this basis, visual stereometry, marking and mirror symmetry between normal and fractured nasolacrimal duct can be achieved. The lacrimal passage assisted planning technique proposed in this paper combines CT imaging technology with image segmentation technology and scientific computing visualization technology, which can improve the safety, accuracy and micro-originality of surgery. It is of great significance for assisting doctors to diagnose.
【学位授予单位】：上海交通大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：R779.6;TP391.7

【相似文献】