Blood motion imaging - a new technique to visualize 2D blood flow

In this study we present new signal processing algorithms for visualization of blood flow in ultrasound imaging systems. Color flow systems produce each image from packets of typically 5-15 pulses transmitted along each scan line in the image. The Doppler shift from moving blood is utilized to remove clutter noise, and for color coding of the blood velocity component along the ultrasound beam. Blood motion causes a corresponding movement in the speckle pattern of the received signal from pulse to pulse. In conventional color flow imaging, substantial temporal and spatial averaging is used in order to get reliable detection of blood vessels, and low variance in the velocity estimate. This averaging process suppresses the spatial speckle pattern in the signal amplitude. In our technique the speckle pattern from the moving blood cells is preserved and enhanced, enabling the user to visually track the blood motion from pulse to pulse. The speckle visualization is combined with conventional color flow color encoding. In addition to preserving the speckle pattern, several image frames per packet of pulse transmissions are computed. The perception of movement is further improved if the scatterers in a large spatial region are imaged almost simultaneously. This is obtained by increasing the time between pulse transmissions in the same beam direction, and using a technique called beam interleaving. After transmitting a pulse in a first direction, there is little available to acquire data in several other beam directions before transmitting the next pulse in the first direction. Visualization of the speckle pattern movement gives the user a correct perception of the blood flow direction and magnitude, and is also useful in separating true blood flow from wall motion artifacts.