The Research On Super-Resolution Ultrasound Imaging Methods

Ultrasound imaging has become one of the most widely used clinical imaging modalities because of its safety,portability,real-time display and low cost.However,ultrasound imaging is limited by the diffraction limit,with spatial resolution in the range of approximately 0.4-2 mm.In order to overcome the diffraction limit,inspired by super-resolution optical microscopy,the concept of super-resolution ultrasound(SR-US)imaging has been proposed and applied.At present,SR-US imaging is mainly divided into two parts,namely SR-US imaging based on single-molecule localization and SR-US imaging based on temporal fluctuation.Although SR-US imaging based on single-molecule localization and SR-US imaging based on temporal fluctuation have made significant breakthroughs in overcoming the diffraction limit of ultrasound imaging,there are still some problems in these two technologies,mainly as follows:In SR-US imaging based on single-molecule localization,the most commonly used single-molecule localization method is the centroid method.However,the centroid method has the disadvantage of low localization accuracy,compared to other the methods.Moreover,the distribution of microbubbles in each ultrasound image needs to be sparse enough so that the single-molecule localization method can accurately locate the position of each microbubble.Therefore,tens of thousands of ultrasound data need to be collected to accumulate enough microbubbles,resulting in low temporal resolution.In addition,the method needs to process the ultrasound images frame by frame,which greatly limits the computational efficiency.Compared with SR-US imaging based on single-molecule localization,SR-US imaging based on temporal fluctuation can reconstruct high-density overlapping microbubbles,which can reduce the number of image frames used in SR-US imaging and improve time resolution.The method processes the entire image sequence instead of frame by frame,which has higher computational efficiency.But,the existing SR-US imaging based on temporal fluctuation is aimed at real-domain signals,so it is impossible to calculate the cumulant of complex-domain signals.Moreover,the spatial resolution of the method is limited by the order of cumulant,which is lower than that of single-molecule localization method.In view of the above problems,the following research work has been carried out in this paper:Firstly,the SR-US imaging based on Gaussian fitting is established to improve the accuracy of individual microbubble localization.At the same time,plane wave(PW)imaging is used to speed up ultrasound data acquisition.The numerical simulation experimental results show that the proposed method can accurately locate the positions of microbubbles in PW ultrasound images,and the imaging speed is also improved.Secondly,the SR-US imaging based on compressed sensing(CS)is established to resolve the positions of high-density overlapping microbubbles.Moreover,the effects of different ultrasound imaging modalities(synthetic transmit aperture imaging and PW imaging)on CS reconstruction results are evaluated.The results of numerical simulation experiments and in vivo rat experiment show that the CS method can effectively resolve the positions of high-density overlapping microbubbles,and then distinguish the adjacent small blood vessel structures.Hence,it improves the temporal resolution while maintaining high spatial resolution.In addition,the numerical simulation results also show that the CS reconstruction results are not easily affected by the ultrasound imaging modalities.Therefore,fast SR-US imaging can be achieved by combining PW imaging with CS reconstruction method.Thirdly,the SR-US imaging based on high-order cumulant is established,and extended to the complex domain to improve SR-US imaging performance.The results of numerical simulation experiments and in vitro physical phantom experiment show that the proposed method can effectively distinguish the adjacent vascular structures,compared with the traditional SR-US imaging based on temporal fluctuation.Therefore,the method can greatly speed up the imaging speed on the basis of ensuring the spatial resolution.