International Journal of Computing Technology Vol II No 1 September 2011 148 Extraction of 3D Heart Shape Using CT Image Abstract— In recent years the causes of heart diseases are increasing tremendously not only in Bangladesh but also in the whole world. Noninvasive imaging techniques like Computed Tomography (CT), Magnetic resonance imaging (MRI), PET etc are gaining abrupt popularity for heart diseases diagnosis. In the proper diagnosis and treatment, extraction of the 3D heart shape is often necessary. For any kind of operations in heart the extraction of heart shape is also necessary. This paper proposes an algorithm which can extract the 3D heart shape easily. We apply it for a CT image size of (512 512 200) voxels and the result is promising. Keywords-Heart extraction; Algorithm; Coronary heart disease; CT. I. INTRODUCTION Cardiovascular disease (CVD) is the most prevalent cause of death in the whole world. CVD includes high blood pressure, coronary heart disease (CHD), congestive heart failure, stroke, and congenital cardiovascular defect. Among these, CHD accounts maximum percent of deaths of CVD in the world [1]. In modern medical system 3D imaging plays the major role for the above mentioned heart diseases. It makes the diagnosis easy and perfect. If a particular small organ of interest can be extracted then it will be easier and more perfect for the physicians to diagnose. This work proposes an algorithm which extracts the heart using CT images. The use of imaging techniques in cardiac applications has grown immensely in the recent years. For coronary artery imaging the methods can be divided into two categories: invasive imaging and noninvasive imaging [2]. Invasive imaging mainly includes X-ray angiography, intravascular ultrasound (IVUS) which are all catheter-based imaging techniques with contrast agents or imaging sensors carried out to the desired location inside the vessel by a catheter. The intrinsic limitation of the X-ray Angiography, however, is the ability to acquire only 2D projections of the vessels, which are indeed 3D anatomical structures. As a result the, the stenosis or lesion may be underestimated by the 2D X-ray angiography images, and the exact 3D geometry of the coronaries cannot be recovered from a single 2D image. However the invasive imaging techniques are not suitable for observational or routine follow up examinations performed in the same patient on a regular basis. With the development of engineering technology, noninvasive imaging modalities have become more and more attractive, and more frequently used in clinical settings. Among the noninvasive imaging modalities, computed tomography angiography (CTA) is the most widely used for patients at present, and magnetic resonance imaging (MRI) is currently more preferable for conducting advanced research and studies [3][4]. To determine the coronary artery diseases (stenosis, congenital abnormalities, coronary embolism, syphilitic aortitis, occlusion of a coronary artery, polyarteritis and coronary artery spasm) and heart chamber related diseases extraction of coronary artery and heart chamber is extreme necessity [5]. Extraction of heart shape is prerequisites of coronary artery and heart chamber extraction. In [6] different morphological operations are used for different shapes extraction. Figure 1. typical X-ray angiography II. PROPOSED ALGORITHM Let A be an array which contains the pixels of original image, X be an array with same size of A and contains zero in each cell initially. Let B be the structuring element and let Md. Mosaddik Hasan 1 Md. Motiur Rahman 1 Dr. Md. Shorif Uddin 2 1 Dept. of Computer Science & Engineering Mawlana Bhashani Science and Technology University,Santosh, Tangail-1902, Bangladesh shohug_0301012@yahoo.com 2 Dept. of Computer Science & Engineering Jahangirnagar University Savar, Dhaka-1342, Bangladesh