Automated Blood Vessel Segmentation of Fundus Images Based on Region Features and Hierarchical Growth Algorithm
نویسندگان
چکیده
Automated blood vessel segmentation plays an important role in the diagnosis and treatment of various cardiovascular and ophthalmologic diseases. In this paper, a novel unsupervised segmentation algorithm is proposed to extract blood vessels from fundus images. At first, an enhanced vessel image is generated by morphological reconstruction, and then divided into three parts: preliminary vessel regions, background regions and undetermined regions. Secondly, a list of region features of blood vessels are defined and used to remove the noise regions in preliminary vessel regions and extract the skeletons of blood vessels. An intermediate vessel image is obtained by combing the denoised preliminary vessel regions with vessel skeletons. After that, a novel hierarchical growth algorithm, namely HGA, is proposed to label each pixel in undetermined regions as vessel or not in an incremental way, which takes the advantage of color and spatial information from the intermediate vessel image and background regions. Finally, postprocessing is performed to remove the nonvessel regions. The proposed algorithm has low computational complexity and outperforms many other state-ofart supervised and unsupervised methods in terms of accuracy. It achieves a vessel segmentation accuracy of 96.0%, 95.8% and 95.1% in an average time of 10.72s, 15.74s and 50.71s on images from three publicly available retinal image datasets DRIVE, STARE, and CHASE DB1, respectively.
منابع مشابه
A Semi-Automated Algorithm for Segmentation of the Left Atrial Appendage Landing Zone: Application in Left Atrial Appendage Occlusion Procedures
Background: Mechanical occlusion of the Left atrial appendage (LAA) using a purpose-built device has emerged as an effective prophylactic treatment in patients with atrial fibrillation at risk of stroke and a contraindication for anticoagulation. A crucial step in procedural planning is the choice of the device size. This is currently based on the manual analysis of the “Device Landing Zone” fr...
متن کاملDetection of Blood Vessels in Color Fundus Images using a Local Radon Transform
Introduction: This paper addresses a method for automatic detection of blood vessels in color fundus images which utilizes two main tools: image partitioning and local Radon transform. Material and Methods: The input images are firstly divided into overlapping windows and then the Radon transform is applied to each. The maximum of the Radon transform in each window corresponds to the probable a...
متن کامل[Automatic detection of vessels in color fundus images].
PURPOSE The main purpose of the paper is to evaluate an automated method for blood vessels segmentation in color fundus images, due to its important role in the diagnosis of several pathologies such as diabetes. The final objective is to introduce the algorithm into a Computer Aided Diagnosis (CAD) tool that would be available in those local medical centers without specialists. METHOD An auto...
متن کاملAn Automated MR Image Segmentation System Using Multi-layer Perceptron Neural Network
Background: Brain tissue segmentation for delineation of 3D anatomical structures from magnetic resonance (MR) images can be used for neuro-degenerative disorders, characterizing morphological differences between subjects based on volumetric analysis of gray matter (GM), white matter (WM) and cerebrospinal fluid (CSF), but only if the obtained segmentation results are correct. Due to image arti...
متن کاملOn Supervised Methods for Segmentation of Blood Vessels in Ocular Fundus Images
Information about the retinal blood vessel network is important for diagnosis, treatment, screening, evaluation and the clinical study of many diseases such as diabetes, hypertension and arteriosclerosis. Automated segmentation and identification of retinal image structures had become one of the major research subjects in the fundus imaging and diagnostic ophthalmology. Automatic segmentation o...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید
ثبت ناماگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید
ورودعنوان ژورنال:
- CoRR
دوره abs/1701.00892 شماره
صفحات -
تاریخ انتشار 2017