Multi-excitation magnetic resonance (MR) image reconstruction

Abstract

Systems and methods are provided for estimating a navigation map for multi-shot phase navigation image reconstruction. Techniques described herein include acquiring a reference image of a subject, calculating a navigation map for a multi-shot MR image of the subject using the reference image, and reconstructing the multi-shot MR image by applying the navigation map to segments of the multi-shot MR image. The MR image and the reference image may be, for example, a T1 weighted image, a T2 weighted image, a liquid attenuation inversion recovery (FLAIR) image, a DWI image, a water separation image, a fat separation image, a fat suppression image, a phase contrast image, or a blood contrast image. The reference image may be of a different type than the MR image, and / or acquired from the same or different sequence as the MR image. The navigation map may correct for motion-induced phase differences of excitation variations in the multi-excitation MR image.

Description

Technical Field

[0001] The present disclosure generally relates to multi-shot magnetic resonance (MR) image reconstruction, to phase-navigated multi-shot MR image reconstruction based on regularized convolution kernel estimation, and to motion correction in multi-shot images using a reference image. Background Art

[0002] Magnetic resonance imaging (MRI) systems can be used to generate images of the interior of the human body. MRI systems can be used to detect magnetic resonance (MR) signals in response to an applied electromagnetic field. MRI techniques can include diffusion-weighted imaging (DWI), which uses water diffusion to study white matter activity in the human brain. DWI processing techniques can be used to address resolution issues with images produced using DWI. Multi-excitation images complicate image reconstruction because varying patient motion across different excitations can result in phase inconsistencies. Summary of the Invention

[0003] At least one aspect of the present disclosure relates to a method for compensating for motion in a multi-shot MR image, the image processing method comprising estimating phase changes between multiple excitations of acquired MR imaging data by: (a) acquiring a reference image of a subject; (b) for each of the multiple excitations, calculating a navigation map using (i) first data for the excitation and (ii) second data from the reference image; and (c) reconstructing the multi-shot MR image of the subject by applying the navigation map of (b) to the multiple excitations of the acquired MR imaging data.

[0004] In some implementations, multiple excitations of the acquired MR imaging data are acquired separately from a reference image. In some implementations, (b) includes minimizing a regularized difference between (i) a Fourier transform of second data of the reference image convolved with a navigation map and (ii) corresponding excitations in the multiple excitations. The Fourier transform can be a partial Fourier transform. In some implementations, the regularized difference is a difference over a limited region of k-space for each excitation. In some implementations, the method includes regularizing phase variations to allow a navigation map to be calculated from undersampled data. In some implementations, regularizing phase variations includes using a geometric regularizer, a null space regularizer, or a combination of a geometric regularizer and a null space regularizer.

[0005] In some implementations, the reference image is a diffusion-weighted image (DWI) having a b-value of zero, and each of the multiple shots is a corresponding DWI image having a b-value greater than zero. In some implementations, the multi-shot MR image is a T1-weighted image, a T2-weighted image, a fluid-attenuated inversion recovery (FLAIR) image, or a DWI image. In some implementations, a navigation map is used to select shots from the multiple shots that should be rejected and reacquired.

[0006] At least one other aspect of the present disclosure relates to a method for generating a motion phase map for multi-shot MR image reconstruction. The method may include generating a transformation of a reference image of a subject. The reference image may correspond to a set of signals captured from a magnetic resonance (MR) scan of the subject. The method may include generating an estimated convolution kernel based on the transformation of the set of signals and the reference image. The method may include generating a motion phase map for the set of signals based on the transformation of the estimated convolution kernel.

[0007] In some implementations, a reference image of the subject is captured using a first imaging process, and the set of signals is captured using a second imaging process. In some implementations, the set of signals corresponds to a T1-weighted image, a T2-weighted image, a fluid-attenuated inversion recovery (FLAIR) image, a DWI image, a hydrolysis image, a fat separation image, a fat-suppressed image, a phase contrast image, or a blood contrast image. In some implementations, the reference image includes a T1-weighted image, a T2-weighted image, a FLAIR image, a DWI image, a hydrolysis image, a fat separation image, a fat-suppressed image, a phase contrast image, or a blood contrast image. In some implementations, the method includes generating a navigation map including a motion phase map and an amplitude map based on a transformation of the estimated convolution kernel.

[0008] In some implementations, the method includes detecting motion exceeding a defined level in the DWI data based on the set of signals and the navigation map. In some implementations, generating an estimated convolution kernel includes executing a regularizer based on a set of regularization terms. In some implementations, the method includes generating one or more reconstructed multi-shot MR images based on the motion phase map.

[0009] At least one other aspect of the present disclosure relates to a system for generating a motion phase map for multi-excitation magnetic resonance image reconstruction. The system may include one or more processors coupled to a non-transitory memory. The system may generate a transformation of a reference image of a subject. The reference image may correspond to a set of signals captured from a magnetic resonance (MR) scan of the subject. The system may generate an estimated convolution kernel based on the transformation of the set of signals and the reference image. The system may generate a motion phase map based on the transformation of the estimated convolution kernel.

[0010] At least one other aspect of the present disclosure relates to an image processing method for compensating for motion in MR images. The image processing method may include estimating phase changes between segments of acquired imaging data by: (a) acquiring a reference image of a subject; (b) computing a navigation map for multi-shot MR images of the subject using the reference image; and (c) reconstructing the multi-shot MR images by applying the navigation map of (b) to the segments of the multi-shot MR images.

[0011] Yet another aspect of the present disclosure relates to a method for MRI reconstruction using a multi-shot MRI pulse sequence. The method comprises (a) obtaining a reference image; (b) acquiring a multi-shot interleaved MR image with excitation-variant phase modulation and an excitation-invariant phase modulation; (c) estimating a complex-valued excitation-specific navigator map corresponding to an element-by-element ratio map between (i) the multi-shot interleaved MR image with excitation-variant phase modulation and (ii) the reference image; (d) using the navigator map to select excitations that should be rejected or reacquired; (e) applying the navigator map to correct for motion or system-induced image phase of the multi-shot interleaved MR image; and (f) performing navigator image reconstruction of k-space data from all excitations of the multi-shot interleaved MR image using the estimated navigator map.

[0012] In some implementations, the reference image includes an image having a different contrast than the multi-excitation interleaved MR image, a combination of reference images, or an image scanned for calibration purposes. In some implementations, the reference image is well aligned with and overlaps the multi-excitation interleaved MR image. In some implementations, an acquisition having the same coil sensitivity profile as the reference image is performed. In some implementations, an excitation-invariant phase difference map is estimated between the reference image and the multi-excitation interleaved DWI, and the excitation-invariant phase difference map is applied by: (i) estimating an excitation-invariant image phase difference map of the multi-excitation interleaved MR image of (b); and (ii) applying the excitation-invariant phase difference map to the reference image or the multi-excitation interleaved MR image such that both data have the same excitation-invariant phase.

[0013] In some implementations, an excitation-invariant phase difference map is obtained via non-navigated reconstruction of k-space data from some or all of the excitations of multi-excitation imaging. In some implementations, (c) includes performing an image-domain regularized minimization technique to generate the navigation map. In some implementations, the regularized minimization technique is performed using k-space data or using data in a hybrid image-frequency domain. In some implementations, the hybrid image-frequency domain is defined as any intermittent domain following an incomplete multidimensional Fourier transform.

[0014] These and other aspects and implementations are discussed in detail below. The foregoing information and the following detailed description include illustrative examples of various aspects and implementations, and provide an overview or framework for understanding the nature and characteristics of the claimed aspects and implementations. The accompanying drawings provide an explanation and further understanding of various aspects and implementations, and are incorporated into and constitute a part of this specification. Various aspects can be combined, and it will be readily understood that the features described in the context of an aspect of the present disclosure can be combined with other aspects. Various aspects can be implemented in any convenient form. In a non-limiting example, by an appropriate computer program, it can be carried on an appropriate carrier medium (computer-readable medium), which can be a tangible carrier medium (e.g., a disk) or an intangible carrier medium (e.g., a communication signal). Various aspects can also be implemented using suitable equipment, which can take the form of a programmable computer that runs a computer program arranged to implement this aspect. As used in the specification and claims, the singular forms of "one," "an," and "the" include plural indicators unless the context clearly indicates otherwise. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The drawings are not intended to be drawn to scale. Like reference numbers and names in the various drawings indicate like elements. For clarity, not every component may be labeled in every drawing. In the drawings:

[0016] Figure 1A Example components of a magnetic resonance imaging system according to one or more implementations are shown;

[0017] Figure 1B An example system for estimating a navigation map for DWI image reconstruction according to one or more implementations is shown;

[0018] Figure 2 depicts an example data flow diagram of an example process for estimating a navigation map for multi-shot magnetic resonance (MR) image reconstruction according to one or more implementations;

[0019] Figure 3 depicts an example data flow diagram of another process for estimating a navigation map for multi-shot MR image reconstruction according to one or more implementations;

[0020] Figure 4 A flow chart depicting an example method of estimating a navigation map for multi-shot MR data using a reference image and performing multi-shot MR image reconstruction using the navigation map according to one or more implementations;

[0021] Figure 5 shows example reference images of a patient according to one or more implementations;

[0022] Figure 6 shows an example DWI image of a patient without motion correction according to one or more implementations;

[0023] Figure 7 shows an example DWI image of a patient for which image correction has been performed according to one or more implementations; and

[0024] Figure 8 is a block diagram of an example computing system suitable for use with the various arrangements described herein, according to one or more example implementations. DETAILED DESCRIPTION

[0025] The following is a detailed description of various concepts and implementations related to techniques, approaches, methods, devices, and systems for phase-guided multi-shot MR image reconstruction using estimated convolution kernels. The various concepts introduced above and discussed in detail below can be implemented in any of a variety of ways, as the concepts described are not limited to any particular implementation. Examples of specific implementations and applications are provided primarily for illustrative purposes. Although illustrative examples are presented for application to DWI imaging, the disclosed approaches are applicable to other multi-shot MR imaging modalities.

[0026] DWI imaging technology can be used to image microscopic diffusivity, which may be particularly useful for studying brain white matter and performing clinical assessments of diffusion within soft tissues. DWI imaging technology utilizes a combination of radio frequency (RF) pulses and gradient magnetic fields to intentionally twist and then twist the transverse magnetization spin phase, which results in motion-related phases for each spin. Incoherent motion (such as diffusion, etc.) manifests as spin phase incoherence, which results in the elimination of a single spin field and ultimately results in a reduction in (e.g., diffusion-weighted) electromagnetic signal levels. The signal loss caused by diffusion can be viewed as absolute image amplitude attenuation (scaling). This mechanism enables clinical assessment of diffusion and can be used for various diagnostic treatments.

[0027] Although DWI data can be acquired using a single-shot pulse sequence, such as single-shot echo planar imaging (EPI), this single-shot DWI technique is limited in spatial resolution, which introduces challenges when measuring detailed diffusion properties in fine structures that require high spatial resolution. To address these limitations, multi-shot techniques can be used to resolve phase changes caused by amplified excitation-to-excitation motion and produce sufficiently high-resolution DWI data. In contrast to single-shot, multi-shot DWI acquisition techniques involve multiple spatially encoded acquisitions, where the entire two-dimensional (2D) or 3D image is encoded and acquired using a single excitation and a single acquisition series.

[0028] All motion, regardless of its source and spin coherence, contributes to phase accumulation during DWI imaging. Coherent spin motion from various sources (e.g., patient movement such as breathing, startling, etc.) accumulates coherent spin phase, which distinguishes it from diffusion and does not cause signal cancellation. Therefore, the coherent phase remains unchanged during signal acquisition after the diffusion encoding module and appears as an additional phase component on complex MR images (e.g., motion phase maps, etc.). Such phase maps are motion-dependent and may cause inconsistent data for certain types of acquisitions.

[0029] Clinical DWI images can be noisy. For diffusion sensitivity, MR imaging systems can utilize high b-values ​​(a measure of diffusion weighting that quantitatively describes the amount of spin "wrapping" and "re-wrapping"), which significantly attenuates the signal. Three-dimensional (3D) imaging can also be used to improve the DWI signal-to-noise ratio (SNR), which generally requires multi-excitation acquisition. Multi-excitation acquisition techniques complicate the DWI image reconstruction process because the patient's motion across different excitation changes can cause phase inconsistencies. Phase navigation can be used to cope with this phase inconsistency during image reconstruction, enabling successful reconstruction of excitation-invariant diffusion-weighted images.

[0030] Model-based phase-navigated DWI reconstruction utilizes explicit knowledge of the motion phase map at each shot to establish a Fourier transform (FT) linear mapping between the acquired k-space data for each shot and a diffusion-weighted image that can be modulated by the shot-specific motion phase map. Phase maps are often estimated using part or all of the data for each shot (e.g., navigator data, etc.), particularly for in vivo scans.

[0031] The systems and methods described herein provide a motion phase map estimation technique that is independent of a particular navigation data acquisition strategy. Using arbitrary structural image data of the same subject from a DWI scan to be corrected, it is exploited that two aligned complex images can be transformed into each other using their ratio image (voxel-wise division of the two images, etc.), where the ratio image is referred to herein as a "navigation map". This image domain voxel-wise relationship can also be compactly captured in k-space as a convolution. The estimation of the navigation map can be formulated as a regularized minimization problem, so the resulting phase can be used as a motion phase map for phase navigation reconstruction. The amplitude map of the estimated navigation map can be used to detect excitations with excessive motion that should be rejected and reacquired.

[0032] Figure 1A An example MRI system that can be utilized in conjunction with the navigation map estimation techniques described herein is shown. Figure 1A, the MRI system 100 may include a computing device 104, a controller 106, a pulse sequence repository 108, a power management system 110, and a magnetic assembly 120. The MRI system 100 is illustrative and includes, but is not limited to, Figure 1A In addition to or in place of the components shown Figure 1A In addition to the components shown, the MRI system may have one or more other components of any suitable type. In addition, the implementation of the components for a particular MRI system may vary from that described herein. Examples of low-field MRI systems may include portable MRI systems that may have a field strength that may be less than or equal to 0.5 T, may be less than or equal to 0.2 T, may be in the range of 1 mT to 100 mT, may be in the range of 50 mT to 0.1 T, may be in the range of 40 mT to 80 mT, may be approximately 64 mT, etc., in non-limiting examples.

[0033] The magnetic assembly 120 may include a BO magnet 122, a shim 124, a radio frequency (RF) transmit and receive coil 126, and a gradient coil 128. The BO magnet 122 may be used to generate a main magnetic field BO. The BO magnet 122 may be a magnetic assembly of any suitable type or combination that can generate a desired main magnetic field BO. In some embodiments, the BO magnet 122 may be one or more permanent magnets, one or more electromagnets, one or more superconducting magnets, or a hybrid magnet including one or more permanent magnets and one or more electromagnets or one or more superconducting magnets. In some embodiments, the BO magnet 122 may be configured to generate a BO magnetic field having a field strength that may be less than or equal to 0.2 T or in the range of 50 mT to 0.1 T.

[0034] In some implementations, the B0 magnet 122 may include a first B0 magnet and a second B0 magnet, each of which may include permanent magnet blocks arranged in concentric rings around a common center. The first B0 magnet and the second B0 magnet may be arranged in a bi-planar configuration such that the imaging area is located between the first B0 magnet and the second B0 magnet. In some embodiments, the first B0 magnet and the second B0 magnet may each be coupled to and supported by a ferromagnetic yoke configured to capture and direct magnetic flux from the first B0 magnet and the second B0 magnet.

[0035] The gradient coils 128 can be arranged to provide a gradient field, and in a non-limiting example, can be arranged to generate gradients in three substantially orthogonal directions (X, Y, and Z) in the B0 field. The gradient coils 128 can be configured to encode the transmitted MR signals by systematically varying the B0 field (the B0 field generated by the B0 magnet 122 or the shim 124) to encode the spatial position of the received MR signals as a function of frequency or phase. In a non-limiting example, the gradient coils 128 can be configured to vary the frequency or phase as a linear function of the spatial position along a particular direction, but more complex spatial encoding profiles can also be provided by using nonlinear gradient coils. In some embodiments, in a non-limiting example, the gradient coils 128 can be implemented using a laminate (e.g., a printed circuit board, etc.).

[0036] MRI scans are performed by exciting and detecting emitted MR signals using transmit and receive coils, referred to herein as radio frequency (RF) coils, respectively. The transmit and receive coils may include separate coils for transmitting and receiving, multiple coils for transmitting or receiving, or the same coil for transmitting and receiving. Thus, the transmit / receive assembly may include one or more coils for transmitting, one or more coils for receiving, or one or more coils for transmitting and receiving. The transmit / receive coils may be referred to as Tx / Rx or Tx / Rx coils to generally refer to various configurations of transmit and receive magnetic components of an MRI system. These terms are used interchangeably herein. Figure 1A In FIG, the RF transmit and receive coils 126 may include one or more transmit coils that may be used to generate RF pulses to induce an oscillating magnetic field B1. The transmit coils may be configured to generate any type of suitable RF pulses.

[0037] The power management system 110 includes electronics that provide operating power to one or more components of the MRI system 100. In non-limiting examples, the power management system 110 may include one or more power supplies, energy storage devices, gradient power components, transmit coil components, or any other suitable power electronics required to provide appropriate operating power to power and operate the components of the MRI system 100. Figure 1A As shown, the power management system 110 may include a power system 112, (one or more) power components 114, a transmit / receive circuit system 116, and may optionally include a thermal management component 118 (e.g., cryogenic cooling equipment for superconducting magnets, water cooling equipment for electromagnets, etc.).

[0038] The power system 112 may include electronics that provide operating power to the magnetic assembly 120 of the MRI system 100. In a non-limiting example, the electronics of the power system 112 may provide operating power to one or more gradient coils (e.g., gradient coil 128 , etc.) to generate one or more gradient magnetic fields, thereby providing spatial encoding of MR signals. Additionally, the electronics of the power system 112 may provide operating power to one or more RF coils (e.g., RF transmit and receive coils 126 , etc.) to generate or receive one or more RF signals from a subject. In a non-limiting example, the power system 112 may include a power supply configured to provide power from a mains supply to the MRI system or an energy storage device. In some embodiments, the power supply may be an AC-to-DC power supply that converts AC power from the mains supply to DC power for use by the MRI system. In some embodiments, the energy storage device may be any one of a battery, a capacitor, a supercapacitor, a flywheel, or any other suitable energy storage device that can bidirectionally receive (e.g., store, etc.) power from the mains supply and supply power to the MRI system. Additionally, the power supply system 112 may include additional power electronics including, but not limited to, power converters, switches, buses, drivers, and any other suitable electronics for powering the MRI system.

[0039] The amplifier 114 may include one or more RF receive (Rx) preamplifiers that amplify MR signals detected by one or more RF receive coils (e.g., coil 126 ), one or more RF transmit (Tx) power components configured to provide power to one or more RF transmit coils (e.g., coil 126 ), one or more gradient power components configured to provide power to one or more gradient coils (e.g., gradient coil 128 ), and one or more shim power components configured to provide power to one or more shims (e.g., shim 124 ). In some implementations, the shim 124 may be implemented using a permanent magnet, an electromagnetic element (e.g., a coil), or a combination thereof. The transmit / receive circuitry 116 may be used to select whether the RF transmit coil or the RF receive coil is in operation.

[0040] like Figure 1AAs shown, the MRI system 100 may include a controller 106 (also referred to as a console), which may include control electronics for sending instructions to and receiving information from a power management system 110. The controller 106 may be configured to implement one or more pulse sequences, which determine instructions sent to the power management system 110 to operate the magnetic assembly 120 according to a desired sequence (e.g., parameters for operating the RF transmit and receive coils 126, parameters for operating the gradient coils 128, etc.). Additionally, the controller 106 may perform processing for estimating a navigation map for DWI reconstruction according to various techniques described herein. A pulse sequence may generally describe the order and timing of the operation of the RF transmit and receive coils 126 and the gradient coils 128 to acquire the resulting MR data. For example, the pulse sequence may indicate the order and duration of transmit pulses, gradient pulses, and acquisition times for the receive coils to acquire MR data.

[0041] The pulse sequence can be organized into a series of time periods. In a non-limiting example, the pulse sequence can include a preprogrammed number of pulse repetition time periods, and applying the pulse sequence can include operating the MRI system according to the parameters of the pulse sequence during the preprogrammed number of pulse repetition time periods. In each time period, the pulse sequence can include parameters for generating RF pulses (e.g., parameters for identifying transmit duration, waveform, amplitude, phase, etc.), parameters for generating gradient fields (e.g., parameters for identifying transmit duration, waveform, amplitude, phase, etc.), timing parameters for controlling when to generate RF or gradient pulses or when to configure (one or more) receiving coils to detect MR signals generated by the subject, and other functionality. As described herein, in some embodiments, the pulse sequence can include parameters for specifying one or more navigation RF pulses.

[0042] Examples of pulse sequences include a zero echo time (ZTE) pulse sequence, a balanced steady-state free precession (bSSFP) pulse sequence, a gradient echo pulse sequence, an inversion recovery pulse sequence, a DWI pulse sequence, a spin echo pulse sequence (which includes a conventional spin echo (CSE) pulse sequence, a fast spin echo (FSE) pulse sequence, a turbo spin echo (TSE) pulse sequence, or any multi-spin echo pulse sequence such as a diffusion-weighted spin echo pulse sequence, an inversion recovery spin echo pulse sequence, an arterial spin labeling pulse sequence, etc.) and an Overhauser imaging pulse sequence, etc.

[0043] like Figure 1AAs shown, the controller 106 can communicate with the computing device 104, which can be programmed to process the received MR data. In a non-limiting example, the computing device 104 can use any suitable image reconstruction processing (including the execution of techniques involving estimation of navigation maps for DWI data as described herein) to process the received MR data to generate one or more MR images. Additionally or alternatively, the controller 106 can process the received MR data based on the techniques described herein to perform DWI reconstruction. The controller 106 can provide information about one or more pulse sequences to the computing device 104 for the computing device to process the data. In a non-limiting example, the controller 106 can provide information about one or more pulse sequences to the computing device 104, and the computing device 104 can estimate the navigation map and perform DWI reconstruction based at least in part on the provided information.

[0044] The computing device 104 can be any electronic device configured to process the acquired MR data and generate one or more images of the subject being imaged. The computing device 104 can include at least one processor and memory (e.g., processing circuitry). The memory can store processor-executable instructions that, when executed by the processor, cause the processor to perform one or more of the operations described herein. The processor can include a microprocessor, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), a graphics processing unit (GPU), a tensor processing unit (TPU), or the like, or a combination thereof. The memory can include, but is not limited to, electronic, optical, magnetic storage or transmission devices, or any other storage or transmission device capable of providing program instructions to the processor. The memory can also include a floppy disk, a CD-ROM, a DVD, a magnetic disk, a memory chip, an ASIC, an FPGA, a read-only memory (ROM), a random access memory (RAM), an electrically erasable programmable ROM (EEPROM), an erasable programmable ROM (EPROM), flash memory, optical media, or any other suitable memory from which the processor can read instructions. The instructions can include code generated from any suitable computer programming language. The computing device 104 may include a combination of Figure 8 1. The computing device 104 may be located in the same room as the MRI system 100 or coupled to the MRI system 100 via a wired or wireless connection.

[0045] In some implementations, the computing device 104 can be a stationary electronic device, such as a desktop computer, a server, a rack-mounted computer, or any other suitable stationary electronic device that can be configured to process MR data and generate one or more images from DWI signals. Alternatively, the computing device 104 can be a portable device, such as a smartphone, a personal digital assistant, a laptop computer, a tablet computer, or any other portable device that can be configured to process DWI data according to the techniques described herein. In some implementations, the computing device 104 can include multiple computing devices of any suitable type, as the aspects of the disclosure provided herein are not limited in this respect. In some implementations, operations described as being performed by the computing device 104 can alternatively be performed by the controller 106, or vice versa. In some implementations, certain operations can be performed by both the controller 106 and the computing device 104 via communication between the devices.

[0046] The MRI system 100 may include one or more external sensors 178. The one or more external sensors may assist in detecting one or more error sources (e.g., motion, noise) that degrade image quality. The controller 106 may be configured to receive information from the one or more external sensors 178. In some embodiments, the controller 106 of the MRI system 100 may be configured to control the operation of the one or more external sensors 178 and to collect information from the one or more external sensors 178. The data collected from the one or more external sensors 178 may be stored in a suitable computer memory and may be utilized to assist in various processing operations of the MRI system 100.

[0047] As described above, the techniques described herein can be used to estimate a motion phase map for DWI data that is independent of a particular navigation data acquisition strategy. Reference image data from the same subject whose DWI scan is to be corrected can be utilized to take advantage of the fact that two aligned complex images can be transformed into each other using their ratio image (e.g., voxel-by-voxel division of the two images, etc.). This image domain voxel-by-voxel relationship can also be compactly captured in k-space as a convolution. A regularized minimization technique can be used to estimate the navigation map, where the output phase is used as the motion phase map for phase navigation reconstruction. Additionally, the amplitude map of the estimated navigation map can be used to detect excitations with excessive motion that should be rejected and reacquired.

[0048] Figure 1B An example system 150 for estimating a navigation map for DWI image reconstruction according to one or more implementations is shown. In a non-limiting example, the system 150 can be used to combine Figure 4In some implementations, the system 150 forms part of an MRI system, such as in conjunction with a Figure 1A MRI system 100, etc. In some implementations, system 150 is external to the MRI system but communicates with the MRI system (or components thereof) to perform the operations described herein. Figure 4 Example method 400.

[0049] like Figure 1B As shown, an embodiment of the system 150 may include a controller 106 and a user interface 160. In some implementations, in conjunction with Figure 1B The functionality of the controller 106 described may be implemented on a computing device in communication with the controller 106 (e.g., in conjunction with Figure 1A The user interface 160 may be implemented on a computing device 104 described herein, for example. The user interface 160 may present or enable inspection of any of the DWI signals, reference images, and reconstructed DWI images described herein. In a non-limiting example, the user interface 160 may provide input related to the performance of the DWI image reconstruction technique by receiving input or configuration data related to the estimation of the navigation map, the performance of the MR scan, or the utilization of reference images for a particular subject.

[0050] The user interface 160 may allow a user to select the type of imaging to be performed by the MRI system (e.g., diffusion-weighted imaging, etc.), select a sampling density for the MR scan, or define any other type of parameter related to MR imaging or model training as described herein. In some implementations, the user interface 160 may display DWI signal data, a reference image, or reconstructed DWI data generated using the techniques described herein via a display in communication with the user interface 160. The user interface 160 may allow a user to initiate imaging by the MRI system, or to perform or coordinate any estimation technique described herein.

[0051] In a non-limiting example, the controller 106 may control various aspects of the example system 150 to combine Figure 4 In some implementations, the controller 106 may control the MRI system (e.g., in conjunction with Figure 1A Additionally or alternatively, Figure 1A The computing device 104 may perform some or all of the functionality of the controller 106. In such an implementation, the computing device 104 may communicate with the controller 106 to exchange information as needed to achieve a desired result.

[0052] The controller 106 may be implemented using software, hardware, or a combination thereof. The controller 106 may include at least one processor and memory (e.g., processing circuitry). The memory may store processor-executable instructions that, when executed by the processor, cause the processor to perform one or more of the operations described herein. The processor may include a microprocessor, an ASIC, an FPGA, a GPU, a TPU, or the like, or a combination thereof. The memory may include, but is not limited to, an electronic, optical, magnetic storage or transmission device, or any other storage or transmission device capable of providing program instructions to the processor. The memory may also include a floppy disk, a CD-ROM, a DVD, a magnetic disk, a memory chip, an ASIC, an FPGA, a ROM, a RAM, an EEPROM, an EPROM, a flash memory, an optical medium, or any other suitable memory from which the processor can read instructions. The instructions may include code generated from any suitable computer programming language. The controller 106 may include a processor that is configured to perform one or more of the operations described herein. Figure 8 Any or all components of the computer system 800 described herein may be used to perform any or all functions of the computer system 800.

[0053] The controller 106 can be configured to perform one or more functions described herein. The controller 106 can store or capture a reference image 152 of the subject. The reference image 152 can be an image of the subject that captures the same portion as a corresponding set of DWI signals 154 (e.g., constituting a multi-shot interleaved DWI scan of the same subject, etc.). The reference image 152 can be a zero-b-value DWI scan image, and can be a complex reference image that can be well aligned with the corresponding set of DWI signals 154. In a non-limiting example, at a low image resolution level, the reference image 152 can completely overlap (or partially overlap beyond a predetermined threshold) with the multi-shot DWI signals 154 to be reconstructed using the techniques described herein. In some implementations, the reference image can be an image with different contrasts (e.g., T1, T2, proton density, fluid-attenuated inversion recovery (FLAIR), etc.), a combination of reference images, or an image scanned for calibration purposes, etc. In some implementations, the reference image 152 can be optimized for subject contrast. A given base type of sequence can be optimized to produce different types of contrast. In a non-limiting example, contrast can be optimized by modifying the timing within the sequence (e.g., TE (time to echo), TR (repetition time), echo train length, or inversion time, etc.). Optimization can also be performed by adding or removing RF pulses or changing parameters within the sequence. Such techniques can be used to create different types of image contrast using sequences of the same basic kind. Non-limiting examples of sequences include T1-weighted, T2-weighted, FLAIR, water / fat separation, fat suppression, phase contrast, or blood contrast, etc. Contrast in MRI refers to the relative image intensity of different tissues (e.g., white matter, gray matter). Such intensity is controlled by the image sequence used (e.g., DWI, T1, T2, etc.) to visualize the different pathologies to be diagnosed.

[0054] A reference image can be used in conjunction with the techniques described herein to estimate a navigation map for the set of DWI signals 154. The set of DWI signals 154 can include excitations from a multi-shot interleaved DWI scan of the subject. The DWI signals 154 can collectively capture a region of the subject, which can also be depicted in the corresponding reference image 152. In some implementations, the DWI signals 154 can be acquired using the same coil sensitivity profile as the reference image 154. In a non-limiting example, this can be applied to a virtual coil sensitivity profile obtained from coil compression and pre-whitening. Multi-shot imaging refers to splitting the acquisition of MRI data (e.g., DWI data, etc.) into multiple readouts. The DWI signals 154 can include data in k-space (e.g., frequency domain, etc.), which can be reconstructed using the techniques described herein to obtain a spatial domain (or "image domain") image. When a single readout (single shot) does not fully cover a sufficient portion of the k-space data, the multi-shot DWI signals 154 can be used to image the subject.

[0055] A multi-excitation DWI image divides the image into segments, acquiring one segment at a time, and once for each excitation. This synthesis scheme of the DWI signal 154 may be susceptible to motion. In a non-limiting example, if the subject being imaged is in a different state for each excitation, the "enough" portion of the synthesis may be a mixture and may leave artifacts in the reconstructed image. Typical patient motion may include large-scale movement of the body part being imaged, physiological movement, or movement of distal parts of the body (in a non-limiting example, extending into the imaging field of view), etc. In addition to motion, other potential causes of phase changes across the excitations of the DWI signal 154 include changes in the scanner magnetic field or RF field (e.g., eddy currents, hysteresis, movement of metal objects in or near the scanner, heating of electronic devices or magnetic components, etc.). Although the techniques described herein provide examples in which the reconstructed multi-excitation MR images are DWI MR images, it should be understood that the techniques described herein are not limited to multi-excitation DWI images and can be implemented with any type of multi-excitation MR images.

[0056] The DWI signals 154 can be captured by any suitable MR imaging technique (e.g., via the MR system 100, etc.). In a non-limiting example, the DWI signals 154 can be captured using a low-field MR system. In some implementations, a set of DWI signals 154 can be generated using a pulse sequence (e.g., a diffusion-weighted steady-state free precession (DW-SSFP) sequence, etc.) that may be specifically designed for use or optimal performance in a low-field environment. Other types of MR imaging systems can also be used to capture the set of DWI signals 154. Each DWI signal in the set of DWI signals 154 can correspond to a single excitation of multi-excitation interleaved DWI. Each excitation of the set of DWI signals 154 can correspond to a high b-value image (as a non-limiting example, the b-value may be greater than the b-value of the reference image 152). Both the set of DWI signals 154 and the corresponding reference image 152 can be provided as input to a navigation map estimator 156, which can generate a corresponding navigation map for each excitation in the set of DWI signals 154. The reference image 152 may have a different phase than the DWI signal 154 .

[0057] The controller 106 may include a navigation map estimator 156 that may estimate a navigation map for each excitation in the set of DWI signals 154 using the reference image 152. The navigation map estimator 156 may be implemented in hardware, software, or a combination of hardware and software. The navigation map estimator 156 may use the technical solutions described herein to generate a navigation map 170 that may include a phase component and an amplitude component. The phase component may be extracted from the navigation map 28 and represented as a phase map 172. The phase map 172 and the DWI signals 154 may be provided as inputs to a navigation reconstruction solver 174 that may apply the estimated navigation map 170 to the DWI signals 154 to generate a reconstructed image 158 by solving Expression 5, which is described in further detail herein. The navigation map 170 may be similar to the navigation map 170 that is combined with the phase component 172, respectively. Figure 2 and Figure 3 The estimated navigation map 220 or the estimated low resolution map 320 of motion phase and magnitude described above can be compared to the phase map 225 ( Figure 2 ) and motion phase diagram 322( Figure 3 The navigation map 170 may be generated using the techniques described herein.

[0058] To estimate the navigation map 170, the navigation map estimator 156 can perform a regularized optimization to estimate the ratio image of the reference image and the corresponding excitation of the DWI signal 154 (e.g., voxel-by-voxel division of the two images, etc.). This image domain voxel-by-voxel relationship can also be compactly captured in k-space as a convolution. The estimation of the navigation map 170 is performed as a regularized minimization problem, and the phase of the result is used as the motion phase map 172 for phase navigation reconstruction. The amplitude map of the estimated navigation map can be used to detect excitations with excessive motion that should be rejected and reacquired. The estimation process of the navigation map 170 is further described herein. In some implementations, the DWI signal 154 itself can indicate excessive motion of the excitation to be rejected. In some implementations, data from external sensors 178 can indicate excessive motion (e.g., vibration sensors, accelerometers, gyroscopes, etc.). Excessive motion can be any amount of motion or acceleration that exceeds a predetermined or configured threshold.

[0059] A corresponding navigation map 170 can be estimated for each shot in the DWI signal 154. A navigation reconstruction solver 174 can apply the phase component of the navigation map 170 (phase map 172 or "motion phase map 172") to the DWI signal 154 to generate a reconstructed image 158. The navigation reconstruction solver 174 can include hardware, software, or a combination of hardware and software for optimizing Expression 5 as described in detail herein to generate the reconstructed image 158. Motion phase can be primarily a DWI concept. In the context of DWI, motion can be classified as global motion or micromotion, depending on the extent to which the subject's movement is imaged. Micromotion refers to movement that is sub-pixel or sub-voxel (e.g., movement that may not be large enough to shift into another pixel). Therefore, for micromotion, the amplitude of the pixel value does not change. Due to the physical properties of DWI, the phase of the pixel value does change. These phase changes cause data inconsistencies in multi-shot DWI, which can be corrected using the navigation map estimated by the navigation map estimator 156 using the techniques described herein.

[0060] Once the navigation map of the DWI signal 154 has been estimated, the navigation map estimator 156 can perform a navigated reconstruction to reconstruct an image from the DWI signal 154. In the context of DWI, navigated reconstruction refers to reconstructing the DWI image in a manner that removes or mitigates inconsistencies in the motion phase data so that these inconsistencies are less apparent in the reconstructed image. As described below, a navigated reconstruction method can be performed by solving Expression 5 using the motion phase map 172 of the corresponding navigation map 170 for the corresponding excitation of the DWI signal 154. The reconstruction technique performed by the navigated reconstruction solver 174 can be referred to as "self-navigation" because the technique described herein does not require reliance on external data (e.g., other instruments for measuring the motion phase map, etc.).

[0061] Once the navigation map 170 has been used to generate the reconstructed image 158, the reconstructed image 158 may be stored in a memory of the controller 106 or another computing device (e.g., computing device 104, etc.). In a non-limiting example, the reconstructed image 158 may then be presented on the user interface 160 or sent to another computing device. The image reconstruction techniques described herein may be self-navigating techniques because external data (e.g., from an external instrument that measures a motion phase map in some manner) may not be required for image reconstruction.

[0062] Figure 2An example data flow diagram 200 is depicted of an example process for estimating a navigation map for DWI image reconstruction according to one or more implementations. In a non-limiting example, the estimation process shown in diagram 200 can be performed by any computing device described herein (e.g., controller 106, computing device 104, or computer system 800, etc.). The estimation process shown in diagram 200 can be used to estimate a navigation map 220, which can be used to reconstruct a diffusion-weighted image or determine whether certain shots of a multi-shot interleaved diffusion-weighted image must be reacquired for proper reconstruction.

[0063] In order to estimate the navigation map 220 of the DWI encoded signal 210, both the reference image 205 and the DWI encoded signal 210, which may be well aligned (e.g., completely overlapped or partially overlapped beyond a predetermined threshold, etc.), can be provided as input to a regularized navigation map reconstruction solver 215 (e.g., which can be similar to and include any structure or functionality of the navigation map estimator 156). The regularized navigation map reconstruction solver 215 can perform an estimation process to generate an estimated navigation map 220, which can include a motion phase map 225 component and an amplitude map 230 component.

[0064] The reference image 205 can be an image of the subject that captures the same portion as the corresponding DWI encoded signal 210. The reference image 205 can be a zero b-value DWI scan image. The reference image 210 can be well aligned so that at a low image resolution level, the reference image 205 can completely overlap (or partially overlap beyond a predetermined threshold) with the multi-excitation DWI signal 210 to be reconstructed using the techniques described herein. In some implementations, an image registration process can be performed to align the reference image 205 with the corresponding DWI signal 210. The reconstructed version of the reference image and the DWI encoded signal 210 in the image domain (e.g., without any correction, such as Figure 6 An image registration process is performed between the reference image 205 and the image corresponding to the DWI encoded signal 210 (e.g., the DWI image shown in FIG. 2 ). The result of the image registration process can be a transformation (e.g., rotation, translation, etc.) of the reference image, which aligns the reference image 205 with the image corresponding to the DWI encoded signal 210. The transformation can be applied to the reference image 205 to create a well-aligned reference image, which can be provided as an input to the regularized navigation map estimation solver 215.

[0065] As described herein, the reference image 205 can be an image with different contrasts (e.g., T1, T2, proton density, fluid-attenuated inversion recovery (FLAIR), etc.), a combination of reference images, or an image scanned for calibration purposes. Contrast in MRI refers to the relative image intensity of different tissues (e.g., white matter, gray matter, etc.). Such intensity is controlled by the image sequence used (e.g., DWI, T1, T2, etc.) to visualize different pathologies to be diagnosed. Images from different MRI sequences can inherently have different phases. Therefore, the reference image 205, which can be obtained from a sequence different from the DWI signal 210, can have a phase different from that of the DWI signal 210. This phase difference can be a static phase difference (sometimes referred to as an excitation-invariant phase difference). In some implementations, a static phase or amplitude map can be applied to the reference image 205 (e.g., via element-by-element multiplication, etc.) to account for this excitation-invariant phase difference. The navigation map 220 generated by the navigation map reconstruction solver 215 can be used to correct for the excitation-variant phase difference between the reference image 205 and the DWI signal 210.

[0066] In order to estimate the navigation map 220 of the DWI signal 210, the navigation map reconstruction solver 215 can perform a regularized optimization process. The theory related to the optimization process for estimating the navigation map 220 can be provided as follows. First, a suitable MR imaging technique can be used to acquire the DWI signal 210 and a well-aligned reference image 205. For high b-value DWI scans, this can be achieved at little cost because DWI can, in practice, be part of a protocol with a structural imaging sequence, such as a zero b-value DWI scan. A zero b-value DWI scan provides a zero b-value image, which can be used as the reference image 205. Although this example describes the process in the context of a zero b-value image as the reference image 205, it should be understood that the reference image 205 can also be an image of the subject with different contrasts (e.g., T1, T2, proton density, FLAIR), a combination of reference images, or an image scanned for calibration purposes.

[0067] The estimation process can assume that the motion phase map is smooth and the overall motion during the DWI scan is negligible (e.g., significant motion that produces substantial misalignment with the reference image, etc.). The DWI image can be modeled as three components: a well-aligned structural image, multiplied by an excitation-invariant scaling map, and then multiplied again by an excitation-specific motion phase map. The excitation-invariant scaling map describes the tissue-related contrast changes between the reference image and the subject image. Mathematically, this multiplication relationship can be expressed as the following equation 1. In equation 1, for a discrete image of size N, at excitation s, the high b-value image x b,s(e.g., DWI signals 210, each of which may correspond to a corresponding excitation of a corresponding DWI image) equal to the zero b-value image x0 (e.g., reference image 205, etc.), the excitation-invariant absolute amplitude attenuation map a, and the motion phase map p s The product of .

[0068] x b,s =x0⊙ a⊙p s =x0⊙h s , (1)

[0069] In Equation 1 above, ⊙ represents element-wise multiplication; and represents the navigation graph 220 to be estimated. The following Expression 2 provides the navigation graph 220 estimation process performed by the navigation graph reconstruction solver 215. In Expression 2, Represents the image corresponding to the excitation x b,s The size is N s The (full or partial) k-space DWI signal 210.

[0070]

[0071] In the above expression 2, represents the partial Fourier transform that maps the reference image 205 to the acquired excitation k-space positions; the operator diag(·) arranges the image of size N into a diagonal matrix of size N×N; is a set of image vectors from which the coefficients The estimated value of the navigation chart 220 is calculated to be and R(g) includes (one or more) regularization terms using an adjustment factor λ ≥ 0. For brevity, the subscript s for B, g, R, and λ has been omitted. Expression 2 can be generalized to multi-coil acquisition scenarios, which involves replacing with their coil-specific counterparts The LS fit error is summed across coils.Multi-coil acquisition serves as data augmentation for the estimation since g is invariant across coils.

[0072] Expression (2) emphasizes the image domain role of the variables and, in a non-limiting example, may be helpful to incorporate prior knowledge of the image domain into the design of the regularizer. In the context of k-space data, the following Expression 3 utilizes the identity F H F=I indicates the convolution between x0 and the estimated navigation map Bg in k-space, thereby enabling the incorporation of prior knowledge of k-space.

[0073]

[0074] In a non-limiting example based on Expression 3, assume that B is a set of linear phase unit vectors such that FB arranges the elements of g to the center of k-space, where the design of g effectively estimates h s The technique described in this paper is independent of any particular navigation data acquisition strategy. Using the center of k-space for estimation may be more robust because it is likely to be the area with the highest signal amplitude and has a greater correlation with the convolution kernel to be estimated. Once g is found, ＊ The solution of , which can be zero-padded to size N, can then be inverse Fourier transformed to obtain the navigation map 220 estimate

[0075] For this k-space design example, the regularizer can be selected based on a number of different criteria. Although certain regularizers are described herein, it should be understood that any suitable regularizer can be used to achieve the desired results. The k-space of is compact (e.g., most of its energy is concentrated near the center of the k-space), which is similar to the properties of the Fourier transform of general structural images. Therefore, the regularizer can be used in conjunction with these techniques. in is a diagonal matrix with non-negative elements, so that the energy promoting g is concentrated to the center of k space. D can be constructed according to the following equation 4.

[0076]

[0077] In Equation 4 above, crop(·) extracts a number of size N from the operands. g flip(·) flips its (k-space) operand on all axes, and takes the magnitude element-wise. The value of λ can be chosen to be proportional to the largest singular value of the LS term matrix. This value can vary across applications.

[0078] Once the navigation map 220 has been estimated using the techniques described herein above, the motion phase map 225 can be used to perform phase-navigated DWI reconstruction, which can be modeled using Expression 5 below.

[0079]

[0080] In the above expression 5, The phase component of the estimated navigation map 220 (e.g., is an estimate of the motion phase map 225, which can be used to solve for the excitation-invariant diffusion-weighted image x b (For example, reconstructing image 158, etc.) In Expression 5, y sis the acquired signal for each excitation (eg, the corresponding DWI signal 210 , etc.), E s is the corresponding MR signal encoding, and P is used to regularize the reconstructed image using the weighting coefficient η. In addition, the amplitude component of the estimated navigation map 220 is an amplitude map 230, which can be used to reject excitations with excessive motion. In a non-limiting example, DWI signals 210 that indicate excessive motion (e.g., an amplitude greater than a predetermined threshold, etc.) in the estimated amplitude map 230 can be discarded or otherwise marked as exhibiting excessive motion. Excitations exhibiting excessive motion can be reacquired and then reprocessed using the techniques described herein.

[0081] Figure 3 An example data flow diagram 300 is depicted for a process for estimating a navigation map for DWI image reconstruction based on k-space data according to one or more implementations. In a non-limiting example, the estimation process shown in diagram 300 can be performed by any computing device described herein (e.g., the controller 106, the computing device 104, or the computer system 800, etc.). As described herein, the estimation process shown in diagram 300 can be used to estimate a navigation map 320, which can be used to reconstruct a diffusion-weighted image or determine whether certain shots of a multi-shot interleaved diffusion-weighted image must be reacquired for proper reconstruction.

[0082] The well-aligned structural image 302 and the DWI-encoded signal 310 (e.g., including motion phase modulation of excitation variations that may be corrected, etc.) can be used in conjunction with a regularization term 312 (e.g., the term described in conjunction with Expression 4, etc.) to estimate a convolution kernel 316 in k-space. The convolution kernel 316 can then be applied to an inverse Fourier transform 318 to generate a low-resolution map 320 of estimated motion phase and amplitude (e.g., the estimated navigation map 220, etc.).

[0083] As described herein, the well-aligned structural image 302 can be any image domain reference image of the subject (e.g., reference image 205, reference image 152, etc.) that completely overlaps (or partially overlaps beyond a predetermined threshold) with the DWI encoded signal 310. An optional static phase / amplitude map 304 can be applied to the well-aligned structural image 302 using an element-by-element multiplication process 306. The well-aligned structural image 302, including data in the image domain, can be Fourier transformed 308 (e.g., a fast Fourier transform (FFT), other Fourier transform processing, etc.) to generate k-space data for the well-aligned structural image 302 in the frequency domain. The DWI encoded signal 310 can also include k-space data for multiple excitations of the DWI image that may be corrected. The DWI encoded signal 310 can be suitable DWI data, such as high b-value image data in the frequency domain (e.g., k-space data, etc.). As described herein, the DWI encoded signal 310 includes motion phase modulation of the excitation variations. The DWI encoded signal 310 can correspond to a DWI image of the same subject as the well-aligned structural image 302.

[0084] The regularized convolution kernel estimation problem solver 314 can be executed by a suitable computer system (e.g., the controller 106, the computing device 104, the computer system 800, etc.) to estimate the convolution kernel 316. The convolution kernel 316 can be a corresponding low-resolution map 320 of the estimated motion phase and amplitude represented in the frequency domain (e.g., k-space, etc.). As described herein, the regularized convolution kernel estimation problem solver 314 can use the frequency domain version of the well-aligned structural image 302 and the DWI encoded signal 310 to estimate the convolution kernel 316. As described herein, in a non-limiting example, the regularized convolution kernel estimation problem solver 314 can use Expression 3 to estimate the convolution kernel 316. The regularization term 312 can be any regularization term described herein and can be used in conjunction with the regularizer selected to optimize the convolution kernel 316.

[0085] Once the convolution kernel 316 has been estimated using the techniques described herein, an inverse Fourier transform 318 (e.g., with zero padding as needed, etc.) can be applied to the convolution kernel 316 to generate an estimated motion phase and magnitude map 320 (e.g., navigation map 220, etc.). The motion phase portion (motion phase map 322) of the estimated motion phase and magnitude map 320 can be used in Expression 5 to reconstruct the DWI encoded signal 310 to generate a corrected DWI image. The magnitude portion (magnitude map 324) of the estimated motion phase and magnitude map 320 can be used to reject certain excitations of the DWI signal 310 that have excessive motion (e.g., excitations that indicate motion above a predetermined threshold).

[0086] Figure 4A flow chart depicts an example method 400 for estimating a navigation map (e.g., navigation map 220) for DWI data using a reference image and reconstructing a DWI image using the navigation map according to one or more implementations. The method 400 may be performed using any suitable computing system (e.g., Figure 1A The controller 106, the computing device 104, Figure 8 800). It will be appreciated that certain steps of method 400 may be performed in parallel (e.g., concurrently) or sequentially while still achieving useful results. As described herein, method 400 may be iteratively performed to generate a navigation map for each shot of a multi-shot interleaved diffusion-weighted image.

[0087] Method 400 may include action 405, in which a reference image of the subject (e.g., reference image 152, reference image 205, well-aligned structural image 302, etc.) may be obtained. In a non-limiting example, the reference image may be obtained by capturing a structural image of a region of the subject for which a DWI scan is to be performed. In a non-limiting example, the reference image may be an image of the subject that captures the same portion as a corresponding set of DWI signals (e.g., a multi-excitation interleaved DWI scan of the same subject, etc.). In some implementations, the reference image may be a zero-b-value DWI scan image and may be a complex reference image that may be well-aligned with the corresponding set of DWI signals. In a non-limiting example, a DWI scan may, in practice, include a scheme with a structural imaging sequence, such as a zero-b-value DWI scan. A zero-b-value DWI scan provides a zero-b-value image that may be used as a reference image. The reference image may also be an image of the subject with different contrasts (e.g., T1, T2, proton density, FLAIR, etc.), a combination of reference images, or an image scanned for calibration purposes. In some implementations, the reference image can be obtained (eg, received, retrieved, etc.) from an external computing system or database.

[0088] Method 400 may include act 410, in which a multi-excitation DWI image of the subject is captured. The multi-excitation image divides the DWI image portion into segments, one segment is acquired at a time, and once for each excitation. Such multi-excitation DWI images are susceptible to motion. In a non-limiting example, if the subject being imaged is in a different state for each excitation, the synthesized "enough" portion is a mixture and may leave artifacts in the reconstructed image. Typical patient motion may include large-scale movement of the body part being imaged, physiological movement, or movement of distal parts of the body (in a non-limiting example, extending into the imaging field of view), etc. In addition to motion, other potential causes of phase changes across excitations of DWI images include changes in the scanner magnetic field or RF field (e.g., eddy currents, hysteresis, movement of metal objects in or near the scanner, heating of electronic devices or magnetic components, etc.). The techniques described herein can be used to correct phase changes across excitations.

[0089] Multi-shot DWI images can be captured using any suitable MR imaging technique (e.g., via MR system 100, etc.). In a non-limiting example, a low-field MR system can be used to capture the DWI images. In some implementations, DWI images can be captured or acquired using pulse sequences (e.g., diffusion-weighted steady-state free precession (DW-SSFP) sequences), which may be specifically designed for use or optimal performance in low-field environments. Other types of MR imaging systems can also be used to capture multi-shot DWI images. Each shot of a multi-shot DWI image can include a high b-value image (e.g., a b-value that is greater than a b-value of a reference image, etc.). In a non-limiting example, a high b-value can be a b-value in the range of 100 to 2000, or any b-value greater than zero. The DWI images can be acquired separately from the reference images, or during the same process as the reference images. In some implementations, the multi-shot DWI images can be obtained (e.g., received, retrieved, etc.) from an external computing system or database. In some implementations, the DWI images can be acquired using the same coil sensitivity profile as that used to acquire the reference images.

[0090] At low image resolution levels, the reference image may completely overlap (or partially overlap beyond a predetermined threshold) with the multi-shot DWI image to be corrected using the techniques described herein. In some implementations, an image registration process may be performed to align the reference image 205 with the corresponding DWI signal 210. A reconstructed version of the reference image and the DWI encoded signal 210 in the image domain (e.g., without any correction, such as Figure 6An image registration process is performed between the reference image 205 and the image corresponding to the DWI encoded signal 210 (e.g., the DWI image shown in FIG. 2 ). The result of the image registration process can be a transformation (e.g., rotation, translation, etc.) of the reference image, which aligns the reference image 205 with the image corresponding to the DWI encoded signal 210. The transformation can be applied to the reference image 205 to create a well-aligned reference image, which can be provided as an input to the regularized navigation map estimation solver 215.

[0091] Method 400 may include act 415, in which a navigation map is generated for each shot of the DWI image. A corresponding navigation map may be generated for each shot of the multi-shot DWI image acquired in act 410. The navigation map may be used for phase navigation correction of the DWI image to correct for motion phase variations in each shot of the DWI image. In a non-limiting example, the navigation map may be combined with Figure 2 and Figure 3 In some embodiments, before estimating the navigation map for each shot, a shot-invariant phase difference map is estimated between the reference image and the multi-shot interleaved DWI image. This phase difference map can then be applied to the reference image or the multi-shot DWI image so that both data have the same shot-invariant phase.

[0092] In addition to the phase differences caused by the motion of the excitation variation, the DWI image to be corrected may contain additional phases that may remain static across excitations. In a non-limiting example, uncompensated DWI encoding gradient moments or eddy current fields may generate (spatially varying) static phase differences relative to the structural reference image used for estimation. Static phase maps can be obtained using a variety of different approaches (e.g., non-navigation reconstruction, gridding, machine learning models, etc.). The static phase difference map can then be used to modulate the structural reference image before the navigation map estimation, allowing the proposed method to focus on estimating phase differences caused by phase corruption caused by time-varying motion in particular. Similar to the inclusion of an excitation-invariant static phase map for improved robustness, instead of using a structural image, in some implementations, the estimation can start from a synthetic high b-value image, which in a non-limiting example is obtained from a deep learning reconstruction method. The reference image can also be synthesized from a combination of other previously acquired images.

[0093] An excitation-invariant phase difference map (sometimes referred to herein as a static phase and amplitude map) is estimated via non-navigation reconstruction of k-space data from some or all of the excitations of the multi-excitation DWI image. Non-navigation reconstruction can be any reconstruction technique that does not involve navigation reconstruction, such as gridding or machine learning techniques (e.g., trained convolutional neural network models, etc.). Once estimated, the excitation-invariant phase difference map can be applied to the reference image or multi-excitation DWI (e.g., by element-wise multiplication or convolution, etc.) so that the two data have the same excitation-invariant phase. Once having the same excitation-invariant phase, the DWI image and the reference image can be used to estimate the navigation map.

[0094] As described herein, the navigation map can be estimated by minimizing the difference between the data of a reference image convolved with the navigation map and the corresponding excitations in a plurality of excitations. This approach is described herein in conjunction with Expression 2 and Expression 3, wherein a regularizer and a regularization term are used for phase variations between images, which allows the navigation map to be calculated based on undersampled data (e.g., captured from a low-field MR system, etc.). Any suitable regularizer can be utilized in the regularization process, including regularizing the phase variations using a geometric regularizer, a null space regularizer, or a combination of a geometric regularizer and a null space regularizer. As described herein, a regularizer can be executed according to a set of regularization terms to estimate the navigation map. As described herein, Expression 2 can be used to implement an image domain regularization minimization technique for generating a navigation map, and Expression 3 can be used to implement a frequency domain regularization minimization technique for generating a navigation map. In some implementations, if the reference image may be in the image domain, the reference image can be transformed into the frequency domain (e.g., using a Fourier transform, etc.) before performing the regularization minimization technique described herein.

[0095] In some implementations, the output of the regularized minimization process includes generating an estimated convolution kernel that represents an estimated navigation map for a particular excitation in the frequency domain. To generate the navigation map, the estimated convolution kernel can be transformed into the image domain (e.g., by performing an inverse Fourier transform, etc.). A motion phase map can be extracted from the navigation map as a phase component, and the motion phase map is used to correct for phase motion of excitation variations in multi-excitation interleaved DWI images. A corresponding amplitude map can be extracted from the navigation map as an amplitude component. The amplitude component can be used to reject excitations that include excessive motion, and the motion phase map can be used to correct for phase motion of excitation variations in each excitation of the acquired DWI images.

[0096] In some implementations, the navigation map can be estimated in a hybrid image-frequency domain. The hybrid image-frequency domain can include any intermittent domain after an incomplete multidimensional Fourier transform. The images described herein can be multidimensional (e.g., 2D, 3D images). Multidimensional Fourier transform processes (such as those described herein) are separable processes. In a non-limiting example, for a 3D spatial domain (with Cartesian directions x, y, and z), a multidimensional Fourier transform can be performed by performing a single-dimensional Fourier transform sequentially (or in parallel) along each direction. An incomplete multidimensional Fourier transform involves not completing the FT along all directions (e.g., leaving at least one dimension untransformed). To implement a hybrid approach, the estimation process described in conjunction with Expression 3 can be utilized, but using an incomplete Fourier transform F as part of the Fourier transform identity.

[0097] Method 400 may include act 420, in which a determination may be made as to whether any excitations should be rejected based on the navigation map. As described herein, the corresponding amplitude portion of the navigation map estimated for each excitation may be used to reject certain excitations of the DWI image that have excessive motion. In a non-limiting example, if the amplitude map indicates an amplitude value or set of amplitude values ​​that exceeds a predetermined threshold (e.g., defined via a configuration file, user input, etc.), the corresponding excitation may be marked as rejected. Any excitation marked as rejected may be retrieved by performing act 410 of method 400. Once all excitations of the DWI image are determined not to indicate excessive motion, method 400 may proceed to act 425.

[0098] Method 400 may include action 425, in which an image reconstruction may be performed using a navigation map. To this end, a navigation map may be applied to each corresponding excitation of a multi-excitation DWI image to correct for motion or system-induced image phase. Applying the navigation map may be performed by element-by-element multiplication or by convolution. The navigation map may be used to perform navigation image reconstruction of k-space data for all excitations from multi-excitation interleaved DWI using an estimated navigation map. Navigation reconstruction may be performed using Expression 5 described herein, which may be iteratively solved using an iterative algorithm such as the fast iterative shrinkage thresholding algorithm FISTA. Other suitable algorithms may also be used to generate a reconstructed DWI image based on Expression 5. In some implementations, a motion phase map of the navigation map (instead of the entire navigation map) may be used for reconstruction. The reconstructed DWI image may then be stored, provided for display, provided to another computing system, or subjected to further processing operations.

[0099] The techniques described herein can be used for any type of image encoding and can therefore be utilized with any type of 2D, 3D or other type of multi-dimensional imaging processing. As described herein, a hybrid data layout can be utilized, for example, in a 3D acquisition that is regularly sampled along a readout dimension, an inverse Fourier transform can be applied along the readout dimension. As described herein, the resulting 2D k-space data can then be used to estimate a navigation map. Therefore, due to the use of a separate reference image, these approaches are independent of both the navigation acquisition strategy and the image encoding. In addition, the techniques described herein can inherently handle contrast differences between the reference image and the DWI image that may be corrected. The source of the structural reference image can be imaging data, where motion sensitivity can be minimized.

[0100] These approaches described herein can be further exploited with severely undersampled data (e.g., sampling a single excitation out of approximately 500 excitations, etc.) to estimate a navigation map. The use of a separate reference image can also be used for global motion correction. Processing efficiency can be improved by generating the navigation map in parallel with the multi-excitation DWI image acquisition process because the estimation process does not need to know the complete high b-value k-space data of the complete DWI image. Performing the estimation process in parallel can shorten the waiting time between the end of data acquisition and the start of model-based high b-value image reconstruction using a motion phase map. In addition, corrections to image acquisition can be calculated and applied during acquisition, including updating the image acquisition field of view or direction to match the overall motion of the subject or determining whether to reject or reacquire DWI image excitations (e.g., segments, etc.).

[0101] The estimation of the motion phase map for each segment of the DWI data can be improved by utilizing an image consistency metric. This metric can be a combination of the consistency between the correction segments and between the segments and a reference image (all of which image the same portion of the subject). Note that while the various examples described herein provide techniques for multi-shot 3D spin echo DWI sequences, the same techniques can be applied to correct any multi-shot 2D or 3D imaging sequence using a reference image.

[0102] Figure 5 A non-limiting example reference image of a patient according to one or more implementations is shown. The reference image is a structural image of the patient's brain. The reference image can be a three-dimensional image, which can be shown here as multiple two-dimensional slices. The reference image can be used with the techniques described herein to correct phase differences in motion-induced excitation changes in DWI images. Figure 6 Shown is the absence of motion correction Figure 5 Example DWI images of a patient. A DWI image is a 3D image that can be shown as including Figure 5The DWI images are blurred and lack some details compared with the reference images due to the phase difference of the excitation changes. Figure 7 An example DWI image of a patient for which image correction has been performed according to one or more implementations is shown. Compared to an uncorrected DWI image, Figure 7 The corrected DWI image shows improved clarity. Figure 7 Corrected DWI images.

[0103] Figure 6 and Figure 7 The b-value is 900 s / mm with and without phase navigation correction. 2 A 3D multi-shot spin echo DWI sequence. This sequence consists of a spin echo DWI contrast preparation segment followed by a spin echo series of 40 echoes. The sequence can be acquired in 518 shots. Figure 5 The reference image shown in was acquired using the same sequence, using 106 excitations without utilizing DWI contrast encoding gradients. Although the same type of sequence may be utilized in this example, it may not be necessary to use the same type of sequence for the reference image.

[0104] Figure 8 1 , or various other example systems and devices described in this disclosure.

[0105] The computing system 800 includes a bus 802 or other communication component for communicating information and a processor 804 coupled to the bus 802 for processing information. The computing system 800 also includes a main memory 806, such as RAM or other dynamic storage device, coupled to the bus 802 for storing information and instructions to be executed by the processor 804. The main memory 806 can also be used to store location information, temporary variables, or other intermediate information during execution of instructions by the processor 804. The computing system 800 can also include a ROM 808 or other static storage device coupled to the bus 802 for storing static information and instructions used by the processor 804. A storage device 810, such as a solid-state device, magnetic disk, or optical disk, is coupled to the bus 802 for persistent storage of information and instructions.

[0106] The computing system 800 may be coupled via bus 802 to a display 814, such as a liquid crystal display or an active matrix display, for displaying information to a user. An input device 812, such as a keyboard including alphanumeric and other keys, may be coupled to bus 802 for communicating information and command selections to the processor 804. In another implementation, the input device 812 comprises a touch screen display. The input device 812 may include any type of biometric sensor or cursor control, such as a mouse, trackball, or cursor direction keys, for communicating direction information and command selections to the processor 804 and for controlling cursor movement on the display 814.

[0107] In some implementations, the computing system 800 may include a communication adapter 816, such as a network adapter. The communication adapter 816 may be coupled to the bus 802 and may be configured to enable communication with a computing or communication network or other computing systems. In various illustrative implementations, the communication adapter 816 may be used to implement any type of network configuration, such as wired (e.g., via Ethernet), wireless (e.g., via Wi-Fi, Bluetooth), pre-configured satellite (e.g., via GPS), ad-hoc, LAN, and WAN.

[0108] According to various implementations, the processes of the illustrative implementations described herein can be implemented by the computing system 800 in response to the processor 804 executing an implementation of the instructions contained in the main memory 806. Such instructions can be read into the main memory 806 from another computer-readable medium, such as the storage device 810. Execution of the implementation of the instructions contained in the main memory 806 causes the computing system 800 to perform the illustrative processes described herein. One or more processors in a multi-processing implementation can also be employed to execute the instructions contained in the main memory 806. In alternative implementations, hard-wired circuitry can be used in place of or in combination with software instructions to implement the illustrative implementations. Thus, the implementations are not limited to any specific combination of hardware circuitry and software.

[0109] Various potential non-limiting embodiments and aspects of the present disclosure include the following:

[0110] Embodiment AA: An image processing method for compensating for motion in a multi-excitation magnetic resonance image, i.e., a multi-excitation MR image, the image processing method comprising estimating phase changes between multiple excitations of acquired MR imaging data by: (a) acquiring a reference image of a subject; (b) for each of the multiple excitations, calculating a navigation map using (i) first data of the excitation and (ii) second data of the reference image; and (c) reconstructing the multi-excitation MR image of the subject by applying the navigation map of (b) to the multiple excitations of the acquired MR imaging data.

[0111] Embodiment AB: Any embodiment disclosed herein (eg, any of Embodiments AA through AI), wherein the plurality of shots of acquired MR imaging data are acquired separately from the reference image.

[0112] Embodiments AC: Any embodiment disclosed in this specification (e.g., any of Embodiments AA to AI), wherein (b) comprises minimizing a regularized difference between (i) a Fourier transform of the second data of the reference image convolved with the navigation map and (ii) corresponding excitations of the plurality of excitations.

[0113] Embodiment AD: Any embodiment disclosed in this specification (e.g., any embodiment of Embodiments AA to AI), wherein (b) comprises minimizing a regularized difference between (i) a partial Fourier transform of the second data of the reference image convolved with the navigation map and (ii) corresponding excitations of the plurality of excitations.

[0114] Embodiments AE: Any embodiment disclosed in this specification (eg, any of embodiments AA to AI), further comprising: regularizing the phase variation to allow computation of a navigation map from undersampled data.

[0115] Embodiments AF: Any embodiment disclosed in this specification (e.g., embodiments AE), wherein regularizing the phase variation comprises using a geometric regularizer, a nullspace regularizer, or a combination of a geometric regularizer and a nullspace regularizer.

[0116] Embodiment AG: Any embodiment disclosed in this specification (e.g., any embodiment of Embodiment AA to AI), wherein the reference image is a diffusion-weighted image (DWI) having a b-value of zero, and each of the multiple excitations is a corresponding DWI image having a b-value greater than zero.

[0117] Embodiments AH: Any embodiment disclosed in this specification (e.g., any embodiment of Embodiments AA to AI), wherein the multi-excitation MR image is a T1-weighted image, a T2-weighted image, a fluid-attenuated inversion recovery (FLAIR) image, a DWI image, a water separation image, a fat separation image, a fat-suppressed image, a phase contrast image, or a blood contrast image.

[0118] Embodiment AI: Any of the embodiments disclosed in this specification (eg, any of Embodiments AA through AH), wherein the navigation map is used to select a shot from the plurality of shots that should be rejected and reacquired.

[0119] Embodiment BA: A method for generating a motion phase map for multi-excitation magnetic resonance image reconstruction, comprising: generating, by one or more processors, a transformation of a reference image of a subject, the reference image corresponding to a set of signals captured from a magnetic resonance scan, i.e., an MR scan, of the subject; generating, by the one or more processors, an estimated convolution kernel based on the set of signals and the transformation of the reference image; and generating, by the one or more processors, a motion phase map of the set of signals based on the transformation of the estimated convolution kernel.

[0120] Embodiment BB: Any embodiment disclosed in this specification, wherein the reference image of the subject is captured using a first imaging process, and the set of signals is captured using a second imaging process.

[0121] Embodiment BC: Any embodiment disclosed in this specification (e.g., any embodiment of Embodiments BA to BH), wherein the set of signals corresponds to a T1-weighted image, a T2-weighted image, a fluid-attenuated inversion recovery (FLAIR) image, a DWI image, a water separation image, a fat separation image, a fat-suppressed image, a phase contrast image, or a blood contrast image.

[0122] Embodiment BD: Any embodiment disclosed herein (eg, any of Embodiments BA to BH), wherein the reference image comprises a T1-weighted image, a T2-weighted image, or a fluid-attenuated inversion recovery (FLAIR) image.

[0123] Embodiment BE: Any embodiment disclosed in this specification (for example, any embodiment in Embodiments BA to BH), further includes: generating a navigation map including the motion phase map and amplitude map by the one or more processors based on the transformation of the estimated convolution kernel.

[0124] Embodiments BF: Any embodiment disclosed herein (eg, embodiment BE), further comprising: detecting, by the one or more processors, motion exceeding a defined level in the DWI data based on the set of signals and the navigation map.

[0125] Embodiment BG: Any embodiment disclosed in this specification (e.g., any of Embodiments BA to BH), wherein generating the estimated convolution kernel includes executing a regularizer by one or more processors according to a set of regularization terms.

[0126] Embodiments BH: Any embodiment disclosed in this specification, further comprising: generating, by one or more processors, one or more reconstructed multi-shot MR images based on the motion phase map.

[0127] Embodiment CA: A system for generating a motion phase map for multi-excitation magnetic resonance image reconstruction, comprising: one or more processors configured to: generate a transformation of a reference image of a subject, the reference image corresponding to a set of signals captured from a magnetic resonance (MR) scan of the subject; generate an estimated convolution kernel based on the set of signals and the transformation of the reference image; and generate a motion phase map based on the transformation of the estimated convolution kernel.

[0128] Embodiment DA: An image processing method for compensating for motion in a magnetic resonance (MR) image, the image processing method comprising estimating phase changes between segments of acquired imaging data by: (a) acquiring a reference image of a subject; (b) calculating a navigation map for a multi-excitation MR image of the subject using the reference image; and (c) reconstructing the multi-excitation MR image by applying the navigation map of (b) to segments of the multi-excitation MR image.

[0129] Embodiment EA: A method for performing MRI reconstruction using a multi-shot magnetic resonance imaging (MRI) pulse sequence, comprising (a) obtaining a reference image; (b) acquiring a multi-shot interleaved MR image with excitation-varied phase modulation and an excitation-invariant phase modulation; (c) estimating a complex-valued excitation-specific navigator map corresponding to an element-by-element ratio map between (i) the multi-shot interleaved MR image with excitation-varied phase modulation and (ii) the reference image; (d) using the navigator map to select excitations that should be rejected or reacquired; (e) applying the navigator map to correct the multi-shot interleaved MR image for motion or system-induced image phase; and (f) performing navigator image reconstruction of k-space data from all excitations of the multi-shot interleaved MR image using the estimated navigator map.

[0130] Embodiment EB: Any embodiment disclosed herein (e.g., any of Embodiments EA to EI), wherein the reference image comprises an image having a different contrast than the multi-excitation interleaved MR image, a combination of reference images, or an image scanned for calibration purposes.

[0131] Embodiment EC: Any embodiment disclosed herein (eg, any of Embodiments EA to EI), wherein the reference image is well aligned and overlapped with the multi-shot interleaved MR image.

[0132] Embodiment ED: Any embodiment disclosed herein (eg, any of embodiments EA through EI), wherein the acquisition has the same coil sensitivity distribution as the reference image.

[0133] Embodiment EE: Any embodiment disclosed in this specification (for example, any embodiment of Embodiments EA to EI), wherein an excitation-invariant phase difference map is estimated between a reference image and a multi-excitation interleaved MR image, and the excitation-invariant phase difference map is applied by performing the following operations, the operations comprising: (i) estimating an excitation-invariant image phase difference map of the multi-excitation interleaved MR image of (b); and (ii) applying the excitation-invariant phase difference map to the reference image or the multi-excitation interleaved MR image so that the two data have the same excitation-invariant phase.

[0134] Embodiment EF: Any embodiment disclosed herein (eg, embodiment EE), wherein the shot-invariant phase difference map is obtained via non-navigated reconstruction of k-space data of some or all shots from multi-shot imaging.

[0135] Embodiment EG: Any embodiment disclosed in this specification (e.g., any of embodiments EA to EI), wherein (c) comprises performing an image-domain regularized minimization technique in generating the navigation graph.

[0136] Embodiments EH: Any embodiment disclosed in this specification (eg, Embodiments EG) wherein the minimization technique is performed using k-space data or using data in the hybrid image-frequency domain.

[0137] Embodiment EI: Any embodiment disclosed in this specification (eg, Embodiment EH), wherein the mixed image-frequency domain is defined as any intermittent domain following an incomplete multidimensional Fourier transform.

[0138] The implementations described herein are described with reference to the accompanying drawings. The accompanying drawings illustrate certain details of specific implementations that implement the systems, methods, and programs described herein. The use of the accompanying drawings to describe the implementations should not be construed as imposing any limitations on the disclosure that may be present in the accompanying drawings.

[0139] It is to be understood that claim elements herein are not to be construed under the provisions of 35 USC § 112(f) unless the element is expressly qualified using the phrase "means for."

[0140] As used herein, the term "circuit" may include hardware configured to perform the functions described herein. In some implementations, each corresponding "circuit" may include a machine-readable medium for configuring the hardware to perform the functions described herein. A circuit may be embodied as one or more circuit system components, including but not limited to processing circuit systems, network interfaces, peripheral devices, input devices, output devices, sensors, etc. In some implementations, a circuit may take the form of one or more analog circuits, electronic circuits (e.g., integrated circuits (ICs), discrete circuits, system-on-chip (SOC) circuits), telecommunications circuits, hybrid circuits, and any other type of "circuit". In this regard, a "circuit" may include any type of component used to perform or facilitate the operations described herein. In non-limiting examples, a circuit as described herein may include one or more transistors, logic gates (e.g., NAND, AND, NOR, OR, XOR, NOT, XNOR), resistors, multiplexers, registers, capacitors, inductors, diodes, and wiring, etc.

[0141] A "circuit" may also include one or more processors communicatively coupled to one or more memories or memory devices. In this regard, the one or more processors may execute instructions stored in the memory, or may execute instructions that are otherwise accessible to the one or more processors. In some implementations, the one or more processors may be embodied in various ways. The one or more processors may be constructed in a manner sufficient to perform at least the operations described herein. In some implementations, the one or more processors may be shared by multiple circuits (e.g., circuit A and circuit B may include or otherwise share the same processor, which, in some example implementations, may execute instructions stored or otherwise accessed via different areas of memory). Alternatively or additionally, the one or more processors may be constructed to perform or otherwise perform certain operations independently of one or more coprocessors.

[0142] In other example embodiments, two or more processors can be coupled via a bus to enable independent, parallel, pipeline or multi-threaded instruction execution. Each processor can be implemented as one or more general-purpose processors, ASICs, FPGAs, GPUs, TPUs, digital signal processors (DSPs) or other suitable electronic data processing components configured to execute instructions provided by the memory. One or more processors can take the form of a single-core processor, a multi-core processor (e.g., a dual-core processor, a triple-core processor, or a quad-core processor), a microprocessor, etc. In some implementations, one or more processors can be outside the device, and in a non-limiting example, one or more processors can be remote processors (e.g., cloud-based processors). Alternatively or additionally, one or more processors can be inside or local to the device. In this regard, a given circuit or its components can be arranged locally (e.g., as part of a local server, a local computing system) or remotely (e.g., as part of a remote server such as a cloud-based server). To this end, a "circuit" as described herein can include components distributed across one or more locations.

[0143] An exemplary system for implementing the entire system or a portion of the implementation may include a general-purpose computing device in the form of a computer, which includes a processing unit, a system memory, and a system bus for coupling various system components including the system memory to the processing unit. Each memory device may include a non-transitory volatile storage medium, a non-volatile memory medium, a non-transitory storage medium (e.g., one or more volatile or non-volatile memories), etc. In some implementations, the non-volatile medium may take the form of a ROM, a flash memory (e.g., a flash memory such as NAND, 3D NAND, NOR, 3D NOR, etc.), an EEPROM, an MRAM, a magnetic storage unit, a hard disk, an optical disk, etc. In other implementations, the volatile storage medium may take the form of a RAM, a TRAM, a ZRAM, etc. Combinations of the above are also included within the scope of machine-readable media. In this regard, in a non-limiting example, machine-executable instructions include instructions and data that cause a general-purpose computer, a special-purpose computer, or a special-purpose processing machine to perform certain functions or groups of functions. According to example implementations described herein, each respective memory device may be operable to maintain or otherwise store information related to operations performed by one or more associated circuits, including processor instructions and associated data (e.g., database components, object code components, script components).

[0144] It should also be noted that, as used herein, the term "input device" may include any type of input device, including but not limited to a keyboard, keypad, mouse, joystick, or other input device that performs similar functions. In contrast, as used herein, the term "output device" may include any type of output device, including but not limited to a computer monitor, printer, fax machine, or other output device that performs similar functions.

[0145] It should be noted that although the figures herein may illustrate a specific order and composition of method steps, it should be understood that the order of these steps may be different from that depicted. In a non-limiting example, two or more steps may be performed simultaneously or partially simultaneously. In addition, some method steps performed as discrete steps may be combined, steps performed as combined steps may be separated into discrete steps, the sequence of certain processes may be reversed or otherwise changed, and the nature or quantity of discrete processes may be altered or varied. Depending on alternative implementations, the order or sequence of any element or device may be varied or replaced. Therefore, all such variations are intended to be included within the scope of the present disclosure as defined in the appended claims. Such variations will depend on the selected machine-readable medium and hardware system as well as designer choice. It should be understood that all such variations are within the scope of the present disclosure. Similarly, software and web implementations of the present disclosure may be accomplished using standard programming techniques with rule-based logic and other logic to complete various database search steps, correlation steps, comparison steps, and decision steps.

[0146] Although this specification contains many specific implementation details, these should not be interpreted as limitations on the scope of any invention or the content that may be claimed, but rather as descriptions of features specific to a particular implementation of the systems and methods described herein. Certain features described in this specification in the context of separate implementations may also be implemented in combination in a single implementation. Conversely, various features described in the context of a single implementation may also be implemented in multiple implementations, either individually or in any suitable sub-combination. Furthermore, although features may be described above as functioning in certain combinations and even initially claimed as such, in some cases, one or more features from the claimed combination may be removed from the combination, and the claimed combination may involve a sub-combination or a variation of the sub-combination.

[0147] In some cases, multitasking and parallel processing can be advantageous. Furthermore, the separation of various system components in the above implementations should not be understood as requiring such separation in all implementations, and it should be understood that the described program components and systems can generally be integrated together in a single software product or packaged into multiple software products.

[0148] Now that some exemplary implementations and implementations have been described, it will be apparent that the foregoing, which has been presented by way of example, is illustrative and not restrictive. In particular, although many of the examples presented herein relate to specific combinations of method actions or system elements, these actions and elements can be combined in other ways to accomplish the same goals. Actions, elements, and features discussed in conjunction with only one implementation are not intended to exclude similar roles in other implementations.

[0149] The phraseology and terminology used herein are for descriptive purposes and should not be construed as limiting. The use of "including," "comprising," "having," "involving," "characterized by," "characterized by," and variations thereof herein are intended to encompass the items listed thereafter, their equivalents and additional items, and alternative implementations consisting exclusively of the items listed thereafter. In one implementation, the systems and methods described herein consist of one, various combinations of more than one, or all of the described elements, actions, or components.

[0150] Any reference to an implementation or element or action of a system or method mentioned herein in the singular may also include implementations that include a plurality of such elements, and any plural reference to any implementation or element or action herein may also include implementations that include only a single element. Reference in the singular or plural form is not intended to limit the presently disclosed systems or methods, their components, actions, or elements to a single or plural configuration. Reference to any action or element based on any information, action, or element may include an implementation in which the action or element is based, at least in part, on any information, action, or element.

[0151] Any implementation disclosed herein may be combined with any other implementation, and references to "implementations," "some implementations," "alternative implementations," "various implementations," or "an implementation," etc., are not necessarily mutually exclusive and are intended to indicate that a particular feature, structure, or characteristic described in connection with an implementation may be included in at least one implementation. Such terms, as used herein, do not necessarily all refer to the same implementation. Any implementation may be combined, inclusively or exclusively, with any other implementation in any manner consistent with the aspects and implementations disclosed herein.

[0152] References to "or" may be construed as inclusive such that any term described using "or" may refer to any of a single, more than one, and all of the described terms.

[0153] Where a technical feature in the drawings, detailed description or any claims is followed by a reference numeral, the reference numeral is included solely for the purpose of increasing the intelligibility of the drawings, detailed description and claims. Therefore, neither the reference numeral nor its absence will have any limiting effect on the scope of any claim element.

[0154] The foregoing description of implementations is presented for purposes of illustration and description. The foregoing description is not intended to be exhaustive or to limit the present disclosure to the precise form disclosed, and in view of the above teachings, variations and changes are possible or can be obtained from the present disclosure. The implementations are selected and described to explain the principles of the present disclosure and its practical application, so that those skilled in the art can utilize various implementations and have various variations suitable for the specific use intended. Other substitutions, modifications, changes and omissions may be made in the design, operating conditions and implementation of the implementation without departing from the scope of the present disclosure as expressed in the appended claims.

Claims

1. An image processing method for compensating for motion in a multi-shot magnetic resonance image, i.e., a multi-shot MR image, the image processing method comprising estimating phase changes between multiple shots of acquired MR imaging data by: (a) Obtain a reference image of the subject; (b) for each shot in the plurality of shots, computing a navigation map using (i) first data for the shot and (ii) second data from the reference image; and (c) reconstructing a multi-shot MR image of the subject by applying the navigation map of (b) to the multiple shots of the acquired MR imaging data.

2. The method according to claim 1, wherein The plurality of shots of acquired MR imaging data are acquired separately from the reference image.

3. The method according to claim 1, wherein (b) comprises minimizing a regularized difference between (i) a partial Fourier transform of the second data of the reference image convolved with the navigation map and (ii) corresponding shots of the plurality of shots.

4. The method according to claim 3, further comprising: The phase variations are regularized to allow computation of a navigation map from undersampled data.

5. The method according to claim 4, wherein Regularizing the phase variation includes using a geometry regularizer.

6. The method according to claim 1, wherein The reference image is a diffusion weighted image (DWI) having a b-value of zero, and each of the plurality of shots is a corresponding DWI image having a b-value greater than zero.

7. The method according to claim 1, wherein The multi-excitation MR image is a T1-weighted image, a T2-weighted image, a fluid-attenuated inversion recovery image (FLAIR image), a DWI image, a water separation image, a fat separation image, a fat-suppressed image, a phase contrast image, or a blood contrast image.

8. The method according to claim 1, wherein The navigation map is used to select a challenge from the plurality of challenges that should be rejected and reacquired.

9. A method for generating a motion phase map for multi-shot magnetic resonance image reconstruction, comprising: generating, by one or more processors, a transform of a reference image of a subject, the reference image corresponding to a set of signals captured from a magnetic resonance scan (MR scan) of the subject; generating, by the one or more processors, an estimated convolution kernel based on the set of signals and the transform of the reference image; as well as A motion phase map of the set of signals is generated by the one or more processors based on a transformation of the estimated convolution kernel.

10. The method according to claim 9, wherein: The reference image of the subject is captured using a first imaging process, and the set of signals is captured using a second imaging process.

11. The method according to claim 9, wherein The reference images include T1-weighted images, T2-weighted images, fluid-attenuated inversion recovery images (FLAIR images), DWI images, water separation images, fat separation images, fat suppression images, phase contrast images, or blood contrast images.

12. The method according to claim 9, further comprising: A navigation map comprising the motion phase map and magnitude map is generated by the one or more processors based on the transformation of the estimated convolution kernel.

13. The method according to claim 12, further comprising: Motion exceeding a defined level is detected in the DWI data by the one or more processors based on the set of signals and the navigation map.

14. The method according to claim 9, further comprising: One or more reconstructed multi-shot MR images are generated by the one or more processors based on the motion phase map.

15. A system for generating a motion phase map for multi-shot magnetic resonance image reconstruction, comprising: One or more processors configured to: generating a transform of a reference image of a subject, the reference image corresponding to a set of signals captured from a magnetic resonance scan (MR scan) of the subject; generating an estimated convolution kernel based on the set of signals and the transform of the reference image; as well as A motion phase map is generated according to the transformation of the estimated convolution kernel.

Images