cARDIAC CYCLE PHASE DETECTION USING AN ARTIFICAL NEURAL NETWORK

this paper proposes a new hybrid approach to estimate the cardiac cycle phases in 2-D echo-cardiographic images as a first step in cardiac volume estimation. The cardiac cycle phase identification has done by making out the anatomical information of the heart with the dataset of both the normal and infant cardiac images of the heart. These dataset here helps in extracting the information about the given image and also differentiates them under the two categories either the heart left ventricle is in the diastolic state or systole state. For identifying this state the mitral valve position of the heart is considered. In the first step, the noise in the image was removed using the median filter and wavelet transforms for the canny edge detection as the second step. In third step, feature extraction, mean and the standard deviation value for the dataset images were calculated. To classify the two states of the heart, the feed forward back propagation neural network was used in the fourth scenario. By training the neural network, the images of the heart were classified into diastole and systole and could be measured in both the manner and compared.