Temporal compounding of cardiac ultrasound data: Improving image quality and clinical measurement repeatability

Echocardiography provides a powerful and versatile tool for assessing cardiac morphology and function. However, cardiac ultrasound suffers from speckle as well as static and dynamic noise. Over the last three decades, a number of studies have attempted to address the challenging problem of speckle/noise suppression in cardiac ultrasound data. No single method has managed to provide a widely accepted solution. Temporal compounding is a noise suppression method that utilises spatial averaging of temporally aligned cardiac B-mode data. Reliable temporal alignment is vital for effective temporal compounding. In this study we introduce a novel, accurate and robust technique for the temporal alignment of cardiac cycles with variable temporal characteristics and examine the effect of temporal compounding in four clinical measurements performed on routine echocardiographic examinations. Results from 32 patients demonstrate speckle/noise suppression, shadowing reduction, anatomical structure enhancement and improvement in measurement repeatability with no significant or systematic bias introduced. Temporally compound data may be able to provide a good alternative to B-mode data in clinical measurements as well as a first step to further post-processing of cardiac ultrasound data.