CREATIS, CNRS UMR 5220, INSERM U1044, UCB Lyon1, INSA Lyon, University of Lyon, France; Jiangsu University, Xuefu Road No. 301, Zhenjiang, Jiangsu 2012013, PR China.
Med Image Anal. 2014 Jan;18(1):1-8. doi: 10.1016/j.media.2013.08.003. Epub 2013 Sep 4.
Computed Tomography Angiography (CTA) plays an essential role in the diagnosis, treatment evaluation, and monitoring of cerebral aneurysms. Segmentation of CTA medical images of giant intracranial aneurysms (GIA) provides quantitative measurements of thrombus and aneurysms geometrical characteristics allowing 3D reconstruction. In fact, GIA demonstrated neuroradiological features and propensity of partial or total spontaneous intra-aneurysmal thrombosis generating a thrombus. Despite intensive researches on medical image segmentation, aneurysm (Lumen, Thrombus, and Parent Blood Vessels) segmentation remains as a difficult problem that has not been yet resolved. In this paper, we proposed a Lattice Boltzmann Geodesic Active Contour Method (LBGM) for aneurysm segmentation in CTA images in order to estimate both the volumes of the thrombus and the aneurysm. Although the noise in the CTA images is very strong and the edges of the thrombus are not so different than the surrounding tissues, the aneurysms are segmented effectively. Based on these results, a method using a dome-neck aspect ratio (AR) parameter for the evaluation of the Spontaneous Thrombosis (ST) phenomena demonstrates the promising potentiality of this LBGM for clinical applications.
计算机断层血管造影（CTA）在脑动脉瘤的诊断、治疗评估和监测中起着至关重要的作用。对巨大颅内动脉瘤（GIA）的 CTA 医学图像进行分割，可提供血栓和动脉瘤几何特征的定量测量，从而进行 3D 重建。事实上，GIA 表现出神经影像学特征和部分或完全自发颅内瘤内血栓形成的倾向，从而产生血栓。尽管对医学图像分割进行了深入研究，但动脉瘤（管腔、血栓和母血管）分割仍然是一个尚未解决的难题。在本文中，我们提出了一种基于格点玻尔兹曼测地线主动轮廓模型（LBGM）的 CTA 图像中动脉瘤分割方法，以估计血栓和动脉瘤的体积。尽管 CTA 图像中的噪声非常强，并且血栓的边缘与周围组织没有太大区别，但仍能有效地分割出动脉瘤。基于这些结果，一种使用穹顶-颈部比（AR）参数来评估自发性血栓形成（ST）现象的方法表明，这种 LBGM 具有潜在的临床应用价值。
