The Study Of Segmentation Algorithm Of Heart CT Image

Image segmentation is one of the important technology of computer-aided diagnosis.The purpose of medical image segmentation is set apart different areas with special meaning in images, and make segmentation results close as possible as anatomical structure, consequently provide reliable basis for clinical diagnosis and pathological research. Generally, image segmentation method can't get ideal segmentation effect if it's directly applied in medical image, because the complexity of the human anatomy, the irregular shape of organs and individual differences. And that because the characteristics of medical image, segmentation algorithm is generally does not make an example. Therefore, targeted medical image segmentation method must be researched. This paper aims to study the effective segmentation method for segmentation processing of the heart CT image, including coronary artery segmentation and ventricular and atrial segmentation.Based on reviewing the domestic and foreign medical image segmentation method, this paper expounds the aim and significance of heart CT image segmentation. According to the characteristics and the special needs of medical image, this paper present the regional growth method first based on centerline. And detailed introduces the extraction of center line and regional growth method based on centerline, then design the image segmentation method of atrium and ventricular. In addition, to improve this method in the light of the existing problems. Finally realize the algorithm in MATLAB and get a good segmentation results.