Zhao Yue, Guo Shuxu, Luo Min, Shi Xue, Bilello Michel, Zhang Shaoxiang, Li Chunming
School of Electronic Engineering, Jilin University, Changchun, Jilin, China.
Department of Radiology, Fujian Provincial Hospital, Fuzhou, Fujian, China.
Magn Reson Imaging. 2018 Jun;49:94-100. doi: 10.1016/j.mri.2017.03.002. Epub 2017 May 15.
In this paper, we present a level set method for multiple sclerosis (MS) lesion segmentation from FLAIR images in the presence of intensity inhomogeneities. We use a three-phase level set formulation of segmentation and bias field estimation to segment MS lesions and normal tissue region (including GM and WM) and CSF and the background from FLAIR images. To save computational load, we derive a two-phase formulation from the original multi-phase level set formulation to segment the MS lesions and normal tissue regions. The derived method inherits the desirable ability to precisely locate object boundaries of the original level set method, which simultaneously performs segmentation and estimation of the bias field to deal with intensity inhomogeneity. Experimental results demonstrate the advantages of our method over other state-of-the-art methods in terms of segmentation accuracy.
在本文中，我们提出了一种水平集方法，用于在存在强度不均匀性的情况下从液体衰减反转恢复（FLAIR）图像中分割多发性硬化症（MS）病变。我们使用一种用于分割和偏置场估计的三相水平集公式，从FLAIR图像中分割MS病变、正常组织区域（包括灰质和白质）、脑脊液和背景。为了节省计算量，我们从原始的多相水平集公式中推导了一种两相公式，以分割MS病变和正常组织区域。所推导的方法继承了原始水平集方法精确定位物体边界的理想能力，该方法同时进行分割和偏置场估计以处理强度不均匀性。实验结果证明了我们的方法在分割精度方面优于其他现有先进方法。
