Heart video tracking system on long axis view

The heart is one of the vital organs in the body, Cardiac performance can be analyzed by a doctor through a cardiac examination with tests echocardiography (heart ultrasound), which usually contains information about the heart movement. One of parameter used in the analysis is the distance between the walls of the heart. The analysis results of the heart condition based on the echocardiography video is depend on accuracy and experiences of the doctor. This condition has a risk of differences the analysis results between doctors so that can causing patients confuse with their health status. This study was conducted to address the differences in diagnosis because of misreading of a heart condition in a video echocardiography. The system built in this study is to search the distance between the wall of the cardiac motion on echocardiography video parasternal long axis (PLAX) are semi-automatic. Histogram equalization, bilateral filter, and median filter are used to improve image quality. Bilateral filter Method is a filter that has three parameters calculation, which is based on the diameter, intensity of the color and depth range. After that, distance features is search using optical flow farneback, the farneback method will process all pixels in an image with a specific range of pixels to obtain higher accuracy. Farneback methods typically used to search for patterns of common movement objects, such as the movement of people and cars, but on the system is to see the pattern of heart cavity movement. Based on the experimental results of this study, conducted comparative measurements of the distance between the walls of the heart using this system and the manual way. Error obtained: 4,09% for the left ventricle, 4,37% for the aorta, and 3,59% for the left atrium.