To correctly diagnose various cardiac disorders, detailed observation of the heart chambers, wall movements and valve functions is necessary. Echocardiography can provide information about the size and shape of the heart as well as the extent of any tissue damage. A number of segmentation algorithms are proposed for cardiac chamber analysis. Here, two methods are considered to extract all four cardiac chambers simultaneously and in a fully automatic manner. In the first method, a phase symmetry approach and a level segmentation algorithm are used. The second approach considers a novel boundary-based object detector for precise and robust tracking of all four cardiac chambers. Key words: echocardiography, echocardiogram, phase symmetry, level set segmentation I. INTRODUCTIONEchocardiography allows estimation of overall cardiac function and observation of cardiac chambers based on B-mode and M-mode imaging. It uses two-dimensional, three-dimensional and Doppler ultrasound for creation of cardiac images . Echocardiographic images are an essential part of clinical cardiovascular research. Noninvasive assessment of cardiac structure by echocardiography can provide insight into disease mechanisms and therapeutic benefits. Segmentation is the process of partitioning parts of an image. The goal of segmentation is to change the representation of an image. It is used to locate objects and boundaries of an image. There are different types of segmentation that can make chamber extraction easier. Most cardiac segmentation methods are semi-automatic and also require user interaction[1]. Initially, segmentation methods were only successful in segmenting the left ventricle. Andrzej Skalskit[2] p...... middle of paper ......images are heavily affected by speckle noise. Two distinct processes constitute echo formation: specular reflection and scattering. Reflection is the interaction between sound waves and objects larger than the wavelength of the pulse and scattering means the interaction between sound waves and objects smaller than the wavelength of the pulse. pulse. Overlapping scatter echoes results in speckle. To remove speckle noise, a Speckle Redusing Anisotropic Diffusion (SRAD) filter is used.2B. MethodAfter preprocessing, pixels with almost identical values ​​are grouped by k means clustering. Here a cluster size of 20 is used. Next, the region and boundary features are captured. The region feature can quickly and approximately locate objects, while the boundary features provide more precision. Region features include color and texture. Limit function includes edge and frame difference.