2D Neural Network Method and apparatus of 2D sample volume based spectral Doppler imaging for motion tracking by using Neural Networks

摘要

An ultrasound imaging apparatus according to an embodiment of the present invention may include an ultrasound transceiving unit, a data acquisition unit, and a motion tracking unit. The ultrasound transceiving unit may focus and transmit a wide beam ultrasound signal on one line of an ultrasound image, and may generate a 2D sample volume by receiving a received ultrasound signal reflected from an object. The data acquisition unit may store received data corresponding to the received ultrasound signal, and may provide IQ data corresponding to an inphase component and a quadrature component of the received data. The motion tracking unit may detect movement of an object and provide motion data based on a correlation value between a 1D sample volume corresponding to a previous frame generated according to IQ data and a 2D sample volume corresponding to the current frame. . The ultrasonic imaging apparatus according to the present invention uses all elements to focus on one line and transmits ultrasonic waves, and then generates a two-dimensional sample volume through multi-beam reception beamforming, and a one-dimensional sample volume corresponding to a previous frame and By detecting a correlation value between the 2D sample volumes corresponding to the current frame through a neural network, it is possible to grasp the movement of the object. In the ultrasound imaging apparatus according to the present invention, by detecting the motion of the object based on the correlation value between the one-dimensional sample volume and the two-dimensional sample volume, it is possible to reduce the negative effect of motions occurring during a spectral Doppler scan , It is possible to improve the sensitivity and image quality of an ultrasound image by tracking and compensating the motion that occurs during scanning.