Feature Detection from Echocardiography Images Using Local Phase Information

Ultrasound images are characterized by their special speckle appearance, low contrast, and low signalto-noise ratio. It is always challenging to extract important clinical information from these images. An important step before formal analysis is to transform the image to significant features of interest. Intensity based methods do not perform particularly well on ultrasound images. However, it has been previously shown that these images respond well to local phase-based methods which are theoretically intensity-invariant and thus suitable for ultrasound images. We extend the previous local phase-based method to detect features using the local phase computed from monogenic signal which is an isotropic extension of the analytic signal. We apply our method of multiscale feature-asymmetry measurement and local phase-gradient computation to cardiac ultrasound (echocardiography) images for the detection of endocardial, epicardial and myocardial centerline.