期刊名称：International Journal of Image Processing (IJIP)
电子版ISSN：1985-2304
出版年度：2021
卷号：15
期号：4
页码：48-67
语种：English
出版社：Computer Science Journals
摘要：Magnetic resonance angiography (MRA) is an emerging magnetic resonance imaging method for the detection and diagnosis of cerebrovascular diseases including cerebral small vessel disease (CSVD). However, the challenges to extract cerebrovascular structures are recognised, especially from the time-of flight MRA (TOF-MRA) images due to the intricate vascular structures and inherent noise. This paper presents a comprehensive review on image processing pipeline which have been successfully applied on CSVD images such as Computed Tomography (CT) scan, Computed Tomography Angiography (CTA), Digital Subtraction Angiography (DSA), Magnetic Resonance Angiography (MRA), and Magnetic Resonance Imaging (MRI), review on various denoising filters in CSVD images such as Nonlocal Mean (NLM) filter, Multiscale filter, Anisotropic Diffusion filter (ADF), Bilateral filter (BF), Smoothing filter, 3D Steerable filter, Moving Average filter, Trilateral filter, Wiener filter, Blockmatching and 3D filtering (BM3D), Non-linear quasi-Newton method (L-BFGS), and Histogram Equalization (HE). This review also features edge preserving filter (EPF) techniques to reduce noises while preserving the edges from TOFMRA images including ADF, BF, NMF, Mean Shift filter (MSF), and Sigma filter (SF).
关键词：Cerebrovascular Segmentation;Time-of-flight Magnetic Resonance Angiography;Signal-to-noise Ratio;Vessel Enhancement;Cerebral Small Vessel Disease
