Abstract:Function of spinal cord is crucial to nerve conduction pathway. Traumatic spinal cord injury often results in a vasculature disruption after primary insult and further leads to abnormal responses of the intact vessels in neighboring tissue during secondary injury. Therefore, the vasculature and blood supply play significant roles in evaluating the spinal cord function . Ultrasound localization microscopy (ULM) overcomes the shortcomings of extensively used angiography, such as computed tomography angiography (CTA) and magnetic resonance angiography (MRA), in terms of limited resolution, radiation and poor-portability, which meets the needs of comprehensive intraoperative examination and prognosis tracking. In this study, an L22-14vX probe with a transmission frequency of 15.625 MHz is utilized, yielding an imaging wavelength of 100 μm. The ULM is conducted based on ultrafast ultrasound technology with multiple tilted plane-wave illuminations. Robust principal component analysis (RPCA) based spatial-temporal clutter filtering method is used for separating the microbubble signals from tissue signals and high frequency noise. Through microbubble localization, trajectory tracking and mapping, subwavelength super-resolution ultrasound imaging is finally achieved. The whole process of microbubble localization and vessel reconstruction are monitored through measuring the time dependent microbubble detections and saturation. Saturation curve corresponds to the time dependent total area covered by microbubble detections on the image. Quantification analysis is carried out for evaluating the imaging results including resolution measurements based on the Fourier ring correlation (FRC) and full-width at half-maximum (FWHM). The in-vivo experimental results show that ULM can be used to obtain super-resolution vasculature imaging in rat spinal cord. The velocity distributed from 1 mm/s to 50 mm/s can be detected. Within the same vessel, the velocity of a point is inversely correlated with the distance from the point to the center of the vessel. The velocity in the center of the vessel is larger than that at the wall of the vessel. The larger vessels support higher flow in the center of the vessel. The FWHM results indicate that ultrafast Doppler displays vessels in diameters between 135 μm and 270 μm while ULM displays them in diameters between 28 μm and 35 μm. The FRC-based resolution evaluation shows that the ULM achieves a super resolution of 16 μm, much less than the imaging wavelength of 100 μm. Yet, long acquisition time is required to detect microbubbles in the smallest vessels, leading to long reconstruction of the microvasculature, which is still a problem worth studying . Compromise between saturation and acquisition time needs considering. Generally speaking, microbubbles are more likely to flow in large vessels, leading to relatively short reconstruction time of large vessels. When saturation curve almost converges, the imaging improvement with new vessels is not so significant that the detail sacrifice of some small microvessels can reduce acquisition time (i.e. most of microvasculature can still be gained when the saturation curve does not converge). Besides, the increase of microbubble concentration and advanced track identification and extraction may also accelerate the saturation rate of convergence with acquisition time decreasing. In conclusion, ULM can be used to obtain a super-resolution imaging of spinal cord microvasculature, giving a 10-fold improvement in resolution in comparison with ultrafast Doppler imaging. Relevant results can facilitate the super-resolution ULM imaging of spinal cord which may promote the function diagnosis, treatment intervention, disability prevention, and prognosis recovery of spinal cord injury.

Calibration‐free parallel transmission of the cervical, thoracic, and lumbar spinal cord at 7T

Optimized Slab Selective Parallel RF Excitation at Ultra High Field

RF shimming in the cervical spinal cord at 7 T

Reliable Off-Resonance Correction in High-Field Cardiac MRI Using Autonomous Cardiac B0 Segmentation with Dual-Modality Deep Neural Networks.

Self-Adaptive Ultrasound Scanning System for Imaging Human Spine

Head-and-neck multi-channel B1+ mapping and carotid arteries RF shimming using a parallel transmit head coil

Dynamic shimming in the cervical spinal cord for multi-echo gradient-echo imaging at 3 T

Non-contrast-enhanced Ultrafast Ultrasound Doppler Imaging of Spinal Cord Micro-Vessels

Automated slice‐specific z‐shimming for functional magnetic resonance imaging of the human spinal cord

Accelerated B1+ mapping and robust parallel transmit pulse design for heart and prostate imaging at 7 T

Unsupervised deep learning with convolutional neural networks for static parallel transmit design: A retrospective study

Deep learning‐based whole‐brain B1+‐mapping at 7T

Influence of scanning plane on Human Spinal Cord functional Magnetic Resonance echo planar imaging

Improving T2*-weighted human cortico-spinal acquisitions with a dedicated algorithm for region-wise shimming

RF shimming strategy for an open 60-channel RF transmit 7T head coil for routine use on the single transmit mode

Universal modes: Calibration‐free time‐interleaved acquisition of modes

Optimizing Transmit Field Inhomogeneity of Parallel RF Transmit Design in 7T MRI using Deep Learning

Towards non-invasive imaging through spinal-cord generated magnetic fields

Dynamic B0 field shimming for improving pseudo-continuous arterial spin labeling at 7 Tesla

Ultrafast Ultrasound Localization Microscopy Method for Spinal Cord Mircovasculature Imaging

Strategies and safety simulations for ultrasonic cervical spinal cord neuromodulation