Study on cerebral vascular extraction and three-dimensional reconstruction based on MRI

Abstract

Information extraction and three-dimensional reconstruction of blood vessel are keystones in visualization of medical images. Firstly, an improved Sobel algorithm combined with fuzzy theory is presented, which is used to extract the cerebral information of serial images and can solve the contradiction between preservation of detail information and high de-noising ability effectively. Then, an improved ray casting volume rendering algorithm based on fusion segmentations is used to improve the rendering speed with a better rendering quality. The experiments show that the presented new method can preserve more complete blood vessel 3-D information and obtain good de-noising effect. This new method can produce better rendering quality and speeds up the volume rendering.