Phase shift variance imaging - a new technique for destructive microbubble imaging

The detection of microbubble contrast agents with ultrasound imaging techniques is the subject of ongoing research. Commonly, the nonlinear response of the agent is employed for detection. The performance of these techniques is, however, affected by nonlinear sound propagation. As an alternative, the change in echo response resulting from microbubble destruction can be employed to detect the agent. In this work, we propose a novel criterion for microbubble destruction detection that allows the rejection of tissue at a defined significance level even for highly echogenic structures in the presence of nonlinear propagation. Most clinical systems provide the hardware requirements for acquisitions consisting of multiple pulses transmitted at the same position, as used in Doppler imaging. Therefore, we develop a processing strategy that distinguishes contrast agent from other stationary or moving structures using these sequences. The proposed criterion is based on the variance of the phase shift of consecutive echoes in the sequence, which, in addition to tissue rejection, permits the distinction of motion from agent disruption. Phantom experiments are conducted to show the validity of the criterion and demonstrate the performance of the new method for contrast detection. Each detection series consists of 20 identical pulses at 9.5 MHz (4.7 MPa peak negative pressure) transmitted at a pulse repetition frequency of 5 kHz. The sequence is applied to phantoms under varied motion and flow conditions. As a first step toward molecular imaging, the technique is applied to microbubbles targeted to vascular endothelial growth factor receptor 2 (VEGFR2) in vitro. The results show a uniform rejection of the background signal while maintaining a contrast enhancement by more than 40 dB. The area under the receiver operating characteristics (ROC) curve is used as the performance metric for the separation of contrast agent and tissue signals, and values larger than 97% demonstrate that an excellent separation was achieved.