Abstract
OBJECTIVE: Harmonic motion imaging (HMI) is an acoustic radiation force-based elasticity imaging technique, which can be used to monitor changes in tissue mechanical properties caused by focused ultrasound (FUS)-induced thermal ablation. In conventional HMI, the amplitude-modulated FUS sequence and imaging pulse are transmitted simultaneously. With this method, the high-amplitude FUS signal must be separated from the imaging data for tissue displacement estimation. Frequency domain notch filtering and bandpass filtering were previously used to filter FUS interference from imaging data. However, FUS interference becomes more pronounced at the increased intensities and durations necessary for thermal ablation, rendering frequency domain filtering challenging. In this study, three methods for FUS interference filtering during HMI-guided FUS ablation (HMIgFUS) were compared. The methods were notch filtering; interleaved HMIgFUS, with interleaved FUS and imaging pulses; and FUS-net, a convolutional neural network-based U-Net autoencoder developed by our group for FUS interference filtering in radiofrequency data. FUS-net was applied here for the first time for the purpose of ablation displacement monitoring. METHODS: The three filtering methods were tested during 20 HMIgFUS acquisitions in an ex vivo canine liver using a range of peak positive pressures from 11 to 18 MPa and durations ranging from 60 to 180 seconds. The B-mode mean squared errors (MSEs) and displacement amplitude contrast-to-noise ratios (CNRs) of the three methods were calculated and compared. RESULTS: The interleaved method for HMIgFUS was found to be significantly robust in avoiding FUS interference in all tested cases for FUS ablation monitoring, especially cases with high FUS pressure and long durations, as opposed to traditional notch filtering and FUS-net filtering. CNRs obtained from displacement amplitude maps in the interleaved data set were significantly higher in all cases than those obtained from the notch filtered and FUS-net data sets. There was not a significant trend in displacement CNR between the FUS-net and notch filtered data sets. However, B-mode MSE was found to be significantly higher when comparing the FUS-net and interleaved data sets as opposed to the notch filtered and interleaved data sets, suggesting further potential of FUS-net as an FUS interference filtering method. CONCLUSION: These findings indicate the robustness of interleaved HMIgFUS in avoiding FUS interference during HMIgFUS monitoring and the advantages, limitations and future potential of FUS-net and notch filtering.