Super-resolution ultrasound imaging

Abstract

The apparatus (106) comprises a processing pipeline (124) including a stationary structure motion compensator (204), a stationary structure remover (206), and a flow structure detector (208) configured to motion correct stationary structures in a series of ultrasound images including stationary and flowing structures with respect to subject motion. The stationary structure remover is configured to remove the stationary structures from the motion corrected series of ultrasound images to generate a flow image of the flowing structures. The flow structure detector is configured to detect maximum flow rates of the flow structures in the flow images over time to generate an image of the detected maximum flow rates. The images of the detected maximum flow rates are accumulated over time to generate a super-resolution ultrasound image.

Description

[Technical field]

[0001] The following relates generally to ultrasound, and more particularly to super-resolution ultrasound imaging. [Background technology]

[0002] Literature shows that the diffraction limit of conventional ultrasound imaging is approximately half the wavelength. Clinical ultrasound imaging applications use wavelengths between 200 microns (200 μm) and 1 millimeter (1 mm), which makes it impossible to image small structures less than 100 μm in diameter, such as microvasculature. Super-resolution ultrasound imaging produces images with a resolution that exceeds the diffraction limit of conventional ultrasound imaging, visualizing microvessels with diameters of 8 μm and greater.

[0003] Unfortunately, super-resolution ultrasound imaging is an invasive procedure that requires continuous intravenous administration of a microbubble-based contrast agent during the examination. In general, in super-resolution ultrasound imaging, the centers of individual microbubbles are tracked over time in ultrasound images. The density of microbubbles in the administered contrast agent is sparse, and because individual microbubbles are tracked, it may not be possible to distinguish individual microbubbles within a dense population. However, when the density is sparse, it may take several minutes (e.g., 7 minutes) for the microbubbles to disperse throughout the entire circulation.

[0004] During the test, the subject must remain motionless for several minutes with an accuracy of about 50 μm. Unfortunately, it is difficult for a person to remain motionless for several minutes. In addition to voluntary movements, the acquisition is also affected by involuntary movements such as breathing and heartbeat. Unlike conventional scans that are completed with a single breath-hold, it is not reasonable to ask the subject to hold their breath for a test that lasts several minutes to reduce the effects of breathing. To compensate for both voluntary and involuntary movements, the processing includes motion compensation. Long acquisition times also hinder real-time display.

[0005] The microbubbles in microbubble-based contrast agents are fragile and tend to burst when exposed to certain levels of acoustic pressure. To prevent bursting, acquisition sequences have been limited to a mechanical index (MI) (used as an indicator of cavitation bioeffects) of 0.05-0.20. The United States Food and Drug Administration (US FDA) limits the MI for diagnostic ultrasound imaging to 1.9.

[0006] Unfortunately, the low MI used in super-resolution ultrasound imaging reduces the transmitted energy, thus reducing the signal-to-noise ratio (SNR) and penetration depth.

[0007] For at least the above reasons, there is an unmet need for improved approaches to super-resolution ultrasound imaging. Summary of the Invention

[0008] Aspects of the present application address the above and others.

[0009] In one aspect, the apparatus comprises a processing pipeline. The processing pipeline comprises a stationary structure motion corrector configured to motion correct stationary structures in a sequence of ultrasound images including stationary structures and flowing structures with respect to subject motion. The processing pipeline further comprises a stationary structure remover configured to remove stationary structures from the sequence of motion corrected ultrasound images to generate flow images of the flowing structures. The processing pipeline further comprises a flow structure detector configured to detect maximum flow rates of flow structures in the flow images over time to generate images of the detected maximum flow rates. The images of the detected maximum flow rates are accumulated over time to generate high resolution ultrasound images, in particular super-resolution ultrasound images.

[0010] The term "super-resolution ultrasound imaging" herein refers to a process for generating ultrasound images having a resolution that exceeds the diffraction limit of the ultrasound waves used to generate the ultrasound images, i.e., a resolution high enough to distinguish structures smaller than half the wavelength of the ultrasound waves used to generate the ultrasound images. The term "super-resolution ultrasound image" herein refers to an ultrasound image generated by super-resolution ultrasound imaging, i.e., an ultrasound image having a resolution high enough to distinguish structures smaller than half the wavelength of the ultrasound waves used to generate the ultrasound images, e.g., structures smaller than 100 μm in size, e.g., smaller than 50 μm in size, e.g., smaller than 10 μm in size.

[0011] An apparatus may include one or more processors programmed or otherwise configured to implement a processing pipeline. The processors may include a central processing unit (CPU), a graphics processing unit (GPU), a microprocessor, and / or others.

[0012] In some embodiments, the device comprises an imaging system configured for ultrasound imaging, including, for example, super-resolution ultrasound imaging of at least the microvasculature of the subject. In some embodiments, the imaging system includes a probe and a console operably coupled to the probe. The probe may include a transducer array having one or more transducer elements, such as, for example, a one-dimensional (1D), matrix or row-column array, a linear, curvilinear or other shaped array, a fully-populated or sparse array, etc. The transducer elements may be configured to convert excitation electrical pulses into an ultrasound pressure field and convert received ultrasound pressure fields (echoes) into electrical (e.g., radio frequency (RF)) signals. The echoes are generated in response to the transmitted pressure field interacting with matter, such as red blood cells, tissue, etc.

[0013] In some embodiments, the console includes a transmit circuit configured to generate excitation electrical pulses that excite the transducer elements, and a receive circuit configured to receive and, optionally, condition and / or pre-process electrical signals, e.g., RF signals, generated by the transducer elements.

[0014] In some embodiments, the device includes a circuit, e.g., a processing unit, configured to implement processing of the received electrical signals, in particular the received RF signals, to generate a sequence of ultrasound images. In some embodiments, the device further includes a circuit, e.g., the same or another processing unit, configured to implement a processing pipeline for processing the sequence of ultrasound images to generate high-resolution ultrasound images, in particular super-resolution ultrasound images, as described herein. It will be understood that the generation of the sequence of ultrasound images from the received electrical signals and the generation of the high-resolution ultrasound images, in particular super-resolution ultrasound images, from the sequence of ultrasound images may be implemented as part of a single processing pipeline, which may be implemented by a single processor. Alternatively, different operations may be distributed among multiple processors. In some embodiments, the processor implementing the processing pipeline or a part of the processing pipeline is included in the console. In some embodiments, the processing pipeline or a part of the processing pipeline is part of a computing device separate from the console. In this embodiment, the RF signals stored in the memory and / or the beamformed images stored in the memory of the computing device or received from the console (e.g., via a suitable wired or wireless connection) may be loaded and processed in the processing pipeline to generate peak-detected images, as described herein.

[0015] In some embodiments, the processing pipeline is configured to generate a series of high-resolution images from the received electrical signals, e.g., RF signals, motion correct the series of high-resolution images for motion of stationary structures due to subject motion (voluntary and / or unconscious), remove stationary structures from the motion-corrected series of high-resolution images to generate flow images of flowing structures (e.g., red blood cells), detect positions of maximum flow rates of the flowing structures in the flow images, and accumulate images of the peak detection over time. The device may further comprise a display and may be configured to display the generated high-resolution ultrasound images. The display may be included in the console or in a separate computing device. In particular, the series of high-resolution images may be a series of super-resolution ultrasound images.

[0016] In some embodiments, the processing pipeline is configured to generate images of microvasculature with diameters of less than 100 μm, e.g., 2 μm or more. The generated high-resolution images are therefore also referred to as super-resolution images, as they may have a resolution high enough to discern structures with dimensions smaller than half the wavelength of the ultrasound emitted from the transducer array. The density of red blood cells in circulation is approximately 5 million / mm 3 Instead of tracking individual red blood cells, the approach tracks their entire stream with examination times of a few seconds, allowing real-time display, easing motion compensation requirements and subject motion constraints, and enabling MI up to the US FDA limits in diagnostic ultrasound imaging, thus without compromising SNR or depth penetration. Furthermore, the approach described herein is non-invasive, unlike microbubble-based approaches.

[0017] Thus, the flowing structure may consist of the flowing biological material of the subject's body under examination without adding any contrast agent or microbubbles. In particular, the flowing biological material may comprise one or more components of the subject's blood, in particular red blood cells. The stationary structure may comprise stationary biological material such as tissue of the subject's body. The flowing structure is a structure that flows relative to the stationary structure, i.e. the stationary structure is stationary relative to the flowing structure. However, it will be understood that the stationary structure may be moving during data acquisition, for example due to voluntary or involuntary movements of the subject's body.

[0018] In some embodiments, the device comprises a flow tracker configured to link detected maximum flow rates across multiple images of detected maximum flow rates to create a flow track. In some embodiments, the device comprises a velocity estimator configured to estimate velocity information based on the flow track. In some embodiments, the device comprises a vessel width estimator configured to estimate a vessel width based on the super-resolution ultrasound images. The flow tracker, velocity estimator, and / or vessel width estimator may be implemented by a processor implementing the processing pipeline, e.g., as part of the processing pipeline, or may be implemented separately therefrom. Alternatively, the flow tracker, velocity estimator, and / or vessel width estimator may be implemented by a separate processor.

[0019] In another aspect, a method, particularly a computer implemented method, comprises motion correcting stationary structures in a sequence of ultrasound images including stationary structures and flowing structures with respect to subject motion. The method further comprises removing the stationary structures from the motion corrected sequence of ultrasound images to generate a flow image of the flowing structure. The method further comprises detecting a maximum flow rate of the flow structure in the flow image over time to generate an image of the detected maximum flow rate. The method further comprises accumulating the images of the detected maximum flow rate over time to generate a high resolution ultrasound image, particularly a super resolution ultrasound image.

[0020] In yet another aspect, a computer program includes instructions that, when executed by a computer, cause the computer to motion-correct stationary structures in a series of ultrasound images including stationary and flowing structures with respect to subject motion, remove stationary structures from the series of motion-corrected ultrasound images to generate flow images of the flowing structures, detect maximum flow rates of the flowing structures in the flow images over time to generate images of the detected maximum flow rates, and accumulate the images of the detected maximum flow rates over time to generate high-resolution ultrasound images, in particular super-resolution ultrasound images. The computer program may be implemented as a computer-readable storage medium having instructions stored thereon, or as a data signal encoding the instructions.

[0021] According to another aspect, disclosed herein are embodiments of a data processing system configured to perform the operations of the methods described herein. In particular, the data processing system may store program code adapted to cause the data processing system to perform the steps of the methods described herein when executed by the data processing system. The data processing system may be embodied as a single computer or as a distributed system including multiple computers, e.g., a client-server system, a cloud-based system, etc.

[0022] Those skilled in the art will appreciate still further aspects of the present application upon reading and understanding the accompanying description. [Brief description of the drawings]

[0023] The present application is illustrated, without limitation, by the accompanying drawings in which like reference numerals indicate similar elements and in which:

[0024] [Figure 1] 1 diagrammatically illustrates an exemplary imaging system according to an embodiment of the present disclosure; [Diagram 2] 2 illustrates a schematic diagram of an exemplary processing pipeline for the system of FIG. 1, according to one embodiment described herein. [Diagram 3]3 diagrammatically illustrates a variation of the processing pipeline of FIG. 2 including a flow tracker, according to one embodiment described herein. [Figure 4] 2 diagrammatically illustrates another variation of the processing pipeline of FIG. 1 including a flow tracker and a velocity estimator, according to an embodiment described herein. [Diagram 5] 2 illustrates a schematic diagram of a variation of the system of FIG. 1 including a vessel width estimator, according to an embodiment described herein. [Figure 6] 1 is a portion of a super-resolution ultrasound image with indicators superimposed thereon that identify locations for vessel width estimation. [Figure 7] 7 is a graph of the vessel width estimated by the vessel width estimator of FIG. 5 for the vessel shown in FIG. 6. [Figure 8] 1 is an exemplary method according to an embodiment of the present disclosure. [Figure 9] 2 shows the accumulation of a super-resolution ultrasound image produced by the system of FIG. 1 after t1 seconds. [Figure 10] 3 shows the accumulation of a super-resolution ultrasound image generated by the system of FIG. 1 after ti seconds. [Figure 11] 2 shows the accumulation of a super-resolution ultrasound image produced by the system of FIG. 1 after tn seconds. [Figure 12] 2 is a super-resolution ultrasound image produced by the system of FIG. 1. [Figure 13] A computed tomography image of the same tissue as that of FIG. 12 is shown. [Figure 14] 13 shows a fusion of the super-resolution ultrasound image of FIG. 12 and the computed tomography image of FIG. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0025] Described below is a non-invasive super-resolution ultrasound imaging approach that mitigates one or more of the above-mentioned shortcomings of microbubble-based super-resolution ultrasound imaging. In general, the approach includes a processing pipeline configured to process ultrasound images to create images of only / mainly flow structures, such as red blood cells, displaying an accumulation of peaks of detected flow structures over time in the image, and generate a super-resolution ultrasound image.

[0026] 1 illustrates an exemplary imaging system 102 configured for ultrasound imaging, including, for example, super-resolution ultrasound imaging of at least a subject's microvasculature. The imaging system 102 includes a probe 104 and a console 106, which interface with each other via appropriate complementary hardware (such as electromechanical connectors 108 and 110, and cable 112, as shown) and / or a wireless interface (not shown).

[0027] The probe 104 has a transducer array 114 that includes one or more transducer elements 116. The transducer array 114 may include a one-dimensional (1-D), matrix or row-column array, a linear, curvilinear or other shaped array, a fully-populated or sparse array, etc. The transducer elements 116 are configured to convert excitation electrical pulses into an ultrasonic pressure field and convert received ultrasonic pressure fields (echoes) into electrical (e.g., radio frequency (RF)) signals. The echoes are generated in response to the transmitted pressure field interacting with matter, e.g., red blood cells, tissue, etc.

[0028] The console 106 has a transmit circuit (TX) 118 configured to generate excitation electrical pulses that excite the transducer elements 116, and a receive circuit (RX) 120 configured to receive RF signals generated by the transducer elements 116.

[0029] In one embodiment, RX 120 (or other circuitry) is also configured to condition or pre-process the RF signal, e.g., amplify, digitize, etc. In the illustrated embodiment, a switch (SW) 122 is configured to switch between TX 118 and RX 120 for transmit and receive operations. In an alternative embodiment, a separate switch is used.

[0030] In one embodiment, the TX118 and RX120 are controlled to simultaneously acquire all image lines in each emission. For example, a subset (i.e., one or a subgroup) of elements can be excited to simultaneously generate a pressure field that emits a focused beam together, and echoes can be received using all of the elements. This can be repeated for multiple different subsets, with each emission / reception providing data to generate a low-resolution image, and a high-resolution image can be generated by combining the low-resolution images. An example of a suitable sequence is described in Jensen et al., “Synthetic aperture ultrasound imaging,” Ultrasonics, vol. 44, pp. e5-el5, 2006. Another example of using plane waves is mentioned in Tanter et al., “Ultrafast imaging in biomedical ultrasound, IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, 2014, 61, 1, pp. 102-119.

[0031] In another embodiment, the TX118 and RX120 are controlled to sequentially acquire one image line at a time. For example, a subset (i.e., one or a subgroup) of elements can be excited to simultaneously generate a pressure field that emits a focused beam together, and the subset can then receive a line of echoes in response. This is repeated for multiple different subsets to sequentially acquire multiple image lines that form an image. It should be understood that the above three sequences are non-limiting and other acquisition sequences are contemplated herein, including known sequences for 2-D, 3-D and / or 4-D imaging.

[0032] The console 106 further includes a processing pipeline 124. The processing pipeline 124 may include one or more processors (e.g., a central processing unit (CPU), a graphics processing unit (GPU), a microprocessor, etc.) configured to execute computer-readable instructions encoded or embedded in a computer-readable storage medium, such as a memory 126, to perform the operations described herein. In general, the processing pipeline 124 is configured to process the RF signals to create a set of images and combine the images to generate a super-resolution ultrasound image.

[0033] As described in more detail below, in one example, the processing pipeline 124 generates a series of super-resolution images using the RF signal, motion-corrects the series of super-resolution images for movement of stationary structures due to the subject's movement (voluntary and / or unconscious), removes stationary structures from the motion-corrected series of super-resolution images to generate flow images of flowing structures (e.g., red blood cells), detects the location of maximum flow rate of the flowing structures in the flow images, and accumulates images of the peak detection over time.

[0034] In one example, this may include generating images of microvasculature less than 100 μm in diameter, for example 2 μm in diameter or greater. The density of red blood cells in circulation is approximately 5 million / mm 3Rather than tracking individual red blood cells, the present approach tracks their entire stream over a few seconds of examination time, allowing real-time display, easing motion compensation requirements and subject motion constraints, and enabling MI up to the US Food and Drug Administration (US FDA) limits for diagnostic ultrasound imaging, thus without compromising SNR or depth penetration. Furthermore, the approach described herein is non-invasive, unlike microbubble-based approaches.

[0035] The console 106 further includes a scan converter 128 and a display 130. The scan converter 128 is configured to scan convert each image for display, for example, by transforming the image into the coordinate system of the display 130. A (2D and / or 3D) super-resolution ultrasound image is then constructed over time by adding the most recent processed image to the currently displayed image. The images may be further processed, for example, smoothed with a Gaussian kernel or other kernels, and / or processed in other ways to account for motion estimation inaccuracies, sub-sample registration, and / or detection sparsity.

[0036] The console 106 further includes a user interface 132 including one or more input devices (e.g., buttons, touch pad, touch screen, etc.) and one or more output devices (e.g., a display screen, speakers, etc.). The console 106 further includes a controller 134 configured to control one or more of the transmit circuitry 118, the receive circuitry 120, the switch 122, the processing pipeline 124, the scan converter 128, the display 130, and / or the user interface 132.

[0037] 2 diagrammatically illustrates a non-limiting example of the processing pipeline 124. The illustrated processing pipeline 124 receives as input RF signals from the receiver circuitry 120 and outputs images that are combined to provide a super-resolution ultrasound image. In this example, the acquisition includes both stationary structures (e.g., organs, blood vessels, etc.) and flowing structures (e.g., red blood cells), and the output image is an image of the detected peaks of the flowing structures.

[0038] The illustrated processing pipeline 124 includes a beamformer 202 configured to beamform the RF signals and output a sequence of images.

[0039] Non-limiting examples of suitable beamforming are described in Stuart et al., “Real-time volumetric synthetic aperture software beamforming of row-column probe data,” IEEE Trans. Ultrason., Ferroelec., Freq. Contr., April 8, 2021. Other beamforming approaches are contemplated herein.

[0040] The illustrated processing pipeline 124 further includes a stationary structure motion corrector 204 configured to correct the sequence of images for motion of stationary structures, for example due to unconscious and / or spontaneous movements of the subject. An example includes rotating and translating the sequence of images to align the images with the stationary structures. A suitable approach to estimating motion of stationary structures from envelope data using speckle correlation is described in Trahey et al., “Angle independent ultrasonic detection of blood flow,” IEEE Trans. Biomed. Eng., vol. BME-34, no. 12, pp. 965-967, 1987.

[0041] In this approach, for local motion estimation, each image is divided into multiple overlapping sub-regions, one of which is identified as the reference image. An example of a suitable sub-region size is 1 × 1 square millimeter (mm 2 ), 10×10mm 2 , more than this, less than this, non-square, etc. An example of a suitable reference image is the first B-mode image, an intermediate B-mode image, a final B-mode image, or another B-mode image in the sequence of B-mode images. In another example, multiple reference images are utilized.

[0042] Another suitable approach to determining both axial and lateral components using lateral vibrations is described in U.S. Patent No. 6,859,659 to Jensen, entitled "Estimation of vector velocity," issued November 9, 2001, which is incorporated herein by reference in its entirety. Another approach using directional beamforming is described in U.S. Patent No. 6,725,076 to Jensen, entitled "Vector velocity estimation using directional beam forming and cross correlation," issued January 25, 2002, which is incorporated herein by reference in its entirety.

[0043] Motion estimation can alternatively be performed on 3D data as described in U.S. Patent Application Publication No. 2016 / 0206285, entitled "3-d flow estimation using row-column addressed transducer arrays," filed Jan. 19, 2015, by Christensen et al., which is incorporated herein by reference in its entirety. In these approaches, motion is determined relative to a reference image across the entire acquisition. All images are then co-registered to the reference frame and aligned via interpolation, e.g., spline interpolation, to obtain a series of images aligned to the same spatial location over time.

[0044] Each subregion in the reference image is cross-correlated with the corresponding subregion in the other image to estimate axial and lateral motion. The estimated motion of each subregion is assigned to the center (and / or other location) of the corresponding subregion, and the collection of motion estimates for an image provides a discrete motion field through that image. The estimated displacements vary in space and time. By using interpolation such as splines on the motion field, the motion at any point in any image (i.e., in space and / or time) can be estimated. All images are then corrected with this motion to align their content with the reference frame before the stationary structure remover 206.

[0045] The illustrated processing pipeline 124 further comprises a stationary structure remover 206 configured to remove stationary structures from the motion corrected image to generate an image of only or mainly flowing structures, e.g. red blood cells. In general, this can be achieved by subtracting stationary structures from the image, leaving the flowing structures.

[0046] Suitable approaches for removing stationary structures include singular value decomposition (SVD), filtering, principal component analysis (PCA), and / or other approaches. Examples of suitable approaches are described in Demene et al., “Spatiotemporal clutter filtering of ultrafast ultrasound data highly increases Doppler and fUltrasound sensitivity,” IEEE Trans. Med. Imag., vol. 34, no. 11, pp. 2271-2285, 2015.

[0047] In one approach, a Casarotti matrix can be formed from the data with a size of NzNx×Nt, where Nz is the number of axial samples, Nx is the number of horizontal lines, and Nt is the number of time samples. The number of calculations to be processed is 0{{NzNx}Nt 2}. The image can be divided into overlapping patches of pixels (e.g., 180x180 pixels) where the borders of the pixels overlap adjacent patches. The singular values ​​of these patches are then calculated. Singular values ​​that represent stationary structures can be set to zero. Singular values ​​that represent most of the noise can also be removed.

[0048] An image of the flowing structure can then be reconstructed from the remaining non-zero singular values. An example of a suitable reconstructor is described in Baranger et al., “Adaptive spatiotemporal SVD clutter filtering for ultrafast Doppler imaging using similarity of spatial singular vectors,” IEEE Trans. Med. Imag., vol. 37, no. 7, pp. 1574-1586, July 2018. In another example, stationary tissue structures are removed using other echo canceling approaches using filters such as those described in Torp, “Clutter rejection filters in color flow imaging: A theoretical approach,” IEEE Trans. Ultrason., Ferroelec., Freq. Contr., vol. 44, pp. 417-424, 1997, and / or other similar methods such as those described in Yu et al., “Eigen-based clutter filter design for ultrasound color flow imaging: A review,” IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, 57(5), 5456258, 2010.

[0049] The illustrated processing pipeline 124 further comprises a flow structure detector 208. The flow structure detector 208 is configured to detect flow structures (e.g., red blood cells) in the flow image. In one example, the flow structure detector 210 detects the location of a local peak in each of the patches. There are many red blood cells, and the complex sum of all scatterers within the vessel has a peak location within the vessel. Unlike the microbubble approach, which requires detection of individual microbubbles, individual red blood cells cannot and do not need to be detected.

[0050] For this approach, the envelope data can be log-compressed and normalized to the standard deviation of all patches in the first image. Peaks above a threshold of minus thirty decibels (-30 dB) from the maximum value in the first (or other) image are identified as the location of the flow of interest (i.e., red blood cells). Subsample locations of the peaks can be found from an interpolation, such as a polynomial interpolation, around the peak location. Those detected outside a predefined window (e.g., 100x100 pixels) around the central region of the patch are discarded, and only those detected inside the window are kept. This can be performed for every patch and every frame in the series of images.

[0051] Again, the output of the processing pipeline 124 is a series of images of maximum flow rates detected, which in this example correspond to flowing red blood cells. As described herein, the scan converter 128 scan converts each image for display, and the accumulation of the individual images over time produces a super-resolution ultrasound image on the display.

[0052] Next, a modified example will be described.

[0053] In one variation, the processing pipeline 124 is part of a different computing device. In this embodiment, the RF signals stored in memory and / or the beamformed images stored in memory can be loaded and processed by the processing pipeline 124 to generate the peak detection images described herein. Such a device can include a scan converter and a display for constructing and visually displaying a super-resolution ultrasound image. Alternatively or additionally, the peak detection images can be communicated to the system 102, another ultrasound system, and / or another computing device for constructing and visually displaying a super-resolution ultrasound image.

[0054] In another variation, the processing pipeline 124 further includes a flow tracker 302, as shown, for example, in FIG. 3. The flow tracker 302 is configured to generate tracks linking flow structures from frame to frame. Suitable approaches include, but are not limited to, nearest neighbor, multi-frame data structure, dynamic programming, combinatorial, multi hypothesis, explicit motion model (e.g., Kalman filtering), learning-based, and / or other technique approaches. Color coding and / or other image processing can be used to visually indicate flow information such as flow direction, volumetric flow rate, and / or derived quantities such as pressure gradient, resistance index, turbulence and / or perfusion. One example of flow tracking is described in U.S. Patent Application Serial No. 16 / 929,398, filed July 15, 2020, by Jensen et al., entitled "Ultrasound Super Resolution Imaging," which is incorporated herein by reference in its entirety.

[0055] In another variation, the processing pipeline 124 further comprises a flow tracker 302 and a velocity estimator 402, for example as shown in FIG. 4. The velocity estimator 402 is configured to process the track position to estimate a velocity, such as an average velocity, a peak velocity, etc. As a non-limiting example, the velocity estimator 402 can be configured to determine a time derivative of the track position, which results in both axial and lateral velocities. Color coding and / or other image processing can be used to visually indicate flow velocity in addition to or in lieu of flow direction and / or other flow information. An example of estimating velocity from flow tracking is described in U.S. Patent Application Serial No. 16 / 929,398, filed July 15, 2020, by Jensen et al., entitled "Ultrasound Super Resolution Imaging."

[0056] In another variation, the system 102 includes a vessel width estimator 502, for example as shown in Fig. 5. The vessel width estimator 502 is configured to estimate the width of a vessel based on a super-resolution ultrasound image. In one example, the user indicates, via the UI 132 and / or other methods, a location on the image where the vessel width is to be estimated, and the width is estimated and displayed. Examples of this are shown in Figs. 6 and 7 and will now be described.

[0057] FIG 6 illustrates a portion of a super-resolution ultrasound image in which a user has positioned an indicator 602 over several vessels of interest, including vessel 604, vessel 606, and vessel 608. The illustrated indicator 602 is a line; however, other shapes are contemplated herein. FIG 7 illustrates a graph 702 of detection density (first axis 704) as a function of spatial distance (second axis 706) over the indicator 602. In this example, estimated widths 708, 710, 712 represent the −3 dB widths of vessels 604, 606, 608, respectively.

[0058] In another variation, other components of blood (eg, white blood cells, platelets, plasma, etc.) are tracked using the approaches described herein.

[0059] FIG. 8 illustrates an exemplary method according to an embodiment of the present disclosure.

[0060] The order of the operations below is for illustration purposes only and is not limiting. As such, one or more operations may be performed in a different order, including but not limited to simultaneously. Additionally, one or more operations may be omitted and / or one or more other operations may be added.

[0061] At 802, ultrasound data is acquired as described herein and / or otherwise.

[0062] At 804, the ultrasound data is beamformed to generate an image as described herein and / or otherwise.

[0063] It should be understood that operations 802 and 804 may be omitted.For example, in another example, a previously acquired and stored image is retrieved from memory.

[0064] At 806, stationary structures are corrected for subject motion in the image as described herein and / or otherwise.

[0065] At 808, stationary structures are removed from the image generating a flow image as described herein and / or otherwise.

[0066] At 810, locations of peaks of flow structures are detected within the flow image as described herein and / or otherwise.

[0067] At 812, a super-resolution ultrasound image is generated and displayed by summing the images of the maximum detected flow rates as described herein and / or otherwise.

[0068] Optionally, flow information such as direction, volumetric flow rate, and / or derived quantities such as pressure gradient, resistance index, turbulence and / or perfusion can be estimated and visualized as described herein and / or otherwise.

[0069] Optionally, velocity information such as average velocity, peak velocity, etc. can be estimated and visualized as described herein and / or otherwise.

[0070] In one embodiment, the above is implemented for 2D imaging.

[0071] Alternatively or additionally, the above may be implemented for 3D ultrasound imaging, eg, to obtain a full volume, eg, using a matrix probe and / or row-column array for volumetric imaging.

[0072] The above may be implemented by computer readable instructions encoded or embodied on memory 126 (i.e., a computer readable storage medium excluding transitory media), which when executed by a computer processor, cause the processor to perform the operations described herein. Additionally or alternatively, at least one of the computer readable instructions is transmitted by a signal, carrier wave, or other transitory medium (not a computer readable storage medium).

[0073] Figures 9-12 show the time evolution of super-resolution ultrasound images at t1, ti, and tn seconds. In this example, larger blood vessels are depicted in Figure 9, while smaller vessels with diameters of 100-200 μm are barely visible. This gradually improves over time as information from more images is added, as shown in Figures 10 and 11. In Figure 10, blood vessels with a diameter of 100 μm are identifiable. In Figure 12, it is also possible to distinguish blood vessels with diameters less than 100 μm, for example 28 μm, as well as paired arteries and veins.

[0074] Figures 12-14 are a comparison of the approach described herein to images generated from contrast-based micro-computed tomography (CT) scans. Both super-resolution ultrasound and CT images are of the kidney. Figure 12 is a super-resolution ultrasound image. The image clearly depicts a long, straight medullary rectus, approximately 20 μm in diameter, and a pair of arcuate arteries and veins at the border of the cerebral cortex and medulla.

[0075] Figure 13 shows a CT image of the same tissue after post-processing with a maximum intensity projection (MIP) algorithm, which projects only the voxels with the highest attenuation values ​​to produce an image. Figure 14 shows the fusion of the super-resolution ultrasound image of Figure 12 with the CT image of Figure 13.

[0076] FIG. 14 shows good correspondence between structures in super-resolution ultrasound and micro-CT images.

[0077] At least some embodiments and / or aspects disclosed herein may be summarized as follows.

[0078] Embodiment 1: An apparatus comprising a processing pipeline including: a stationary structure motion corrector configured to motion correct stationary structures in a series of ultrasound images including stationary structures and flow structures with respect to subject motion; a stationary structure remover configured to generate a flow image of the flow structure by removing stationary structures from the motion corrected series of ultrasound images; and a flow structure detector configured to detect a maximum flow rate of a flow structure in the flow image over time to generate an image of the detected maximum flow rate, the image of the detected maximum flow rate being accumulated over time to generate a super-resolution ultrasound image.

[0079] Embodiment 2: An apparatus as described in embodiment 1, wherein the flow structures include red blood cells, the detected peaks correspond to the flow of red blood cells, and the high resolution ultrasound image visually indicates the vasculature.

[0080] Embodiment 3: An apparatus described in any of embodiments 1 to 2, wherein the stationary structure motion compensator aligns stationary structures across a series of ultrasound images to compensate for subject motion.

[0081] Embodiment 4: An apparatus as described in embodiment 3, wherein the stationary structure motion employs a motion field to align stationary structures in time, space, or both time and space.

[0082] Embodiment 5: An apparatus described in any of embodiments 3 to 4, wherein the subject movement includes voluntary subject movement, involuntary subject movement, or both voluntary and involuntary subject movement.

[0083] Embodiment 6: An apparatus according to any of embodiments 3 to 5, wherein the stationary structure remover subtracts stationary structures from the series of motion-compensated ultrasound images to generate a flow image.

[0084] Embodiment 7: An apparatus as described in any one of embodiments 1 to 6, wherein the processing pipeline is configured to generate a high-resolution ultrasound image in 1 second to 10 seconds.

[0085] Embodiment 8: An apparatus as described in any of embodiments 1 to 7, further comprising a flow tracker configured to link detected maximum flow rates across multiple images of the detected maximum flow rates to create a track of the flow.

[0086] Embodiment 9: An apparatus as described in embodiment 8, wherein the flow tracker is configured to determine flow information based on the flow track.

[0087] Embodiment 10: An apparatus according to any of embodiments 8 to 9, further comprising a velocity estimator configured to estimate velocity information based on the flow track.

[0088] Embodiment 11: An apparatus described in any of embodiments 1 to 9, wherein the high resolution ultrasound image is a 2D image or a 3D image.

[0089] Embodiment 12: A method comprising: motion correcting stationary structures in a series of ultrasound images including stationary structures and flowing structures with respect to subject movement; generating a flow image of the flowing structure by removing the stationary structures from the motion-corrected series of ultrasound images; detecting a maximum flow rate of the flow structure in the flow image over time to generate an image of the detected maximum flow rate; and accumulating the images of the detected maximum flow rate over time to generate a super-resolution ultrasound image.

[0090] Embodiment 13: The method of embodiment 12, wherein the flow structures include red blood cells, the detected peaks correspond to red blood cell flow, and the super-resolution ultrasound image visually indicates microvasculature.

[0091] Embodiment 14: The method of any of embodiments 12 to 13, wherein the motion compensation includes aligning stationary structures across the series of ultrasound images to compensate for subject motion.

[0092] Embodiment 15: The method of embodiment 14, wherein aligning the stationary structures includes applying a motion field to align the stationary structures in time and space.

[0093] Embodiment 16: The method of any of embodiments 14 to 15, wherein removing stationary structures comprises subtracting stationary structures from the series of motion-compensated ultrasound images.

[0094] Embodiment 17: The method of any one of embodiments 15 to 16, further comprising generating a high-resolution ultrasound image in 1 second to 10 seconds.

[0095] Embodiment 18: The method of any of embodiments 12 to 17, further comprising estimating flow information based on the detected maximum flow rate.

[0096] Embodiment 19: The method of embodiment 18, further comprising estimating velocity information based on a detected maximum flow rate.

[0097] Embodiment 20: A computer-readable storage medium storing instructions which, when executed by a computer, cause the computer to: motion correct stationary structures in a series of ultrasound images including stationary structures and flowing structures with respect to subject movement; generate a flow image of the flowing structure by removing the stationary structures from the motion-corrected series of ultrasound images; detect a maximum flow rate of the flow structure in the flow image over time to generate an image of the detected maximum flow rate; and accumulate the images of the detected maximum flow rate over time to generate a high-resolution ultrasound image.

[0098] This application has been described with reference to various embodiments. Modifications and variations will occur to others upon reading this application. It is intended that the invention be construed as including all such modifications and variations insofar as they come within the scope of the appended claims or the equivalents thereof.

Claims

1. a stationary structure motion corrector (204) configured to motion correct stationary structures in a sequence of ultrasound images, the stationary structures including moving structures, with respect to a subject's motion; a stationary structure remover (206) configured to remove the stationary structures from the sequence of motion-corrected ultrasound images to generate a flow image of the flowing structures; a flow structure detector (208) configured to detect maximum flow rates of the flow structures in the flow image over time to generate an image of the detected maximum flow rates, the image of the detected maximum flow rates being accumulated over time to generate a super-resolution ultrasound image; a processing pipeline (124) including: Apparatus (106).

2. the flow structures include red blood cells, the detected peaks correspond to the flow of the red blood cells, and the super-resolution ultrasound image visually indicates a vasculature.

2. The apparatus of claim 1.

3. the stationary structure motion corrector aligns the stationary structures across the series of ultrasound images to compensate for motion of the subject.

3. Apparatus according to any one of claims 1 to 2.

4. the stationary structure motion compensator employs a motion field to align the stationary structures temporally, spatially, or temporally and spatially; 4. The apparatus of claim 3.

5. The subject movement includes voluntary subject movement, involuntary subject movement, or both voluntary and involuntary subject movement.

5. Apparatus according to any one of claims 3 to 4.

6. the stationary structure remover subtracts the stationary structures from the series of motion compensated ultrasound images to generate the flow image.

6. Apparatus according to any one of claims 3 to 5.

7. the processing pipeline is configured to generate the super-resolution ultrasound image in 1 second to 10 seconds.

7. Apparatus according to any one of claims 1 to 6.

8. a flow tracker (302) configured to link the detected maximum flow rates across multiple images of the detected maximum flow rates to create a track of the flow.

8. Apparatus according to any one of claims 1 to 7.

9. the flow tracker is configured to determine flow rate information based on a flow track; 9. The apparatus of claim 8.

10. a velocity estimator (402) configured to estimate velocity information based on the flow track; 10. Apparatus according to any one of claims 8 to 9.

11. The super-resolution ultrasound image is a 2D image or a 3D image.

10. Apparatus according to any one of claims 1 to 9.

12. a transducer array having one or more transducer elements for emitting ultrasonic waves; the super-resolution ultrasound image has a resolution high enough to distinguish structures with dimensions smaller than half the wavelength of the ultrasound emitted by the transducer array; 10. Apparatus according to any one of the preceding claims.

13. motion correcting stationary structures in a series of ultrasound images, the stationary structures including moving structures, for motion of a subject; removing the stationary structures from the series of motion-corrected ultrasound images to generate a flow image of the flowing structures; detecting a maximum flow rate of the flow structure in the flow image over time to generate an image of the detected maximum flow rate; accumulating images of the detected maximum flow rates over time to generate a super-resolution ultrasound image; Equipped with method.

14. the flow structures include red blood cells, the detected peaks correspond to the flow of the red blood cells, and the super-resolution ultrasound image visually indicates microvasculature. The method of claim 13.

15. and correcting the motion includes aligning the stationary structures across the series of ultrasound images to compensate for motion of the subject.

15. The method according to any one of claims 13 to 14.

16. the super-resolution ultrasound images have a resolution high enough to distinguish structures with dimensions smaller than half the wavelength of the ultrasound used to generate the series of ultrasound images; 16. The method according to any one of claims 13 to 15.

17. A computer-readable storage medium storing instructions that, when executed by the computer, cause the computer to: motion correcting stationary structures in a series of ultrasound images, the stationary structures including moving structures, for motion of a subject; removing the stationary structures from the series of motion-corrected ultrasound images to generate a flow image of the flowing structures; detecting a maximum flow rate of the flow structure in the flow image over time to generate an image of the detected maximum flow rate; accumulating images of the detected maximum flow rates over time to generate a super-resolution ultrasound image; A storage medium that performs the above.

Images