3D echocardiography fusion to improve image quality, leading to improved disease diagnosis

Abstract

3D echocardiography is a unique method of visualizing the morphology of heart. In 3D echocardiography multiple images of the subject are acquired by the probe are not of very good quality. Thus, ordinary image processing techniques do not provide satisfactory results. So we devised a method for improving image quality. Image fusion is a good method used for improving the quality of ultrasound images. This improvement in image quality leads to an improved cardiac functional analysis as the fused multi-view images show the whole picture of the heart. Four techniques have been proposed for the image fusion including maximum, averaging, Wavelet image fusion and principal component analysis (PCA). In Averaging, various data sets are taken and respective frames of all data sets are averaged. In image maximization, maximum values of all data sets are picked as most of the information lies in brighter area so that the resultant fused image contains the maximum information. Wavelet image fusion is done by taking wavelet transform of the two input images which are then combined using fusion rule.PCA is used to reduce the number of variables of the data set while retaining most of original variability in the data. Results of different techniques have been compared using the parameters like Signal to noise ratio (SNR), contrast to noise ratio (CNR), Contrast and field of view(FOV).