Multiple contrast biological material imaging using three- dimensional bssfp UTE MRI

Abstract

A method of imaging a biological material having a component that exhibits an ultra-short transverse relaxation time after excitement by electromagnetic energy is disclosed. More specifically, balanced steady state free precession ultra-short echo time (bSSFP UTE) magnetic resonance imaging (MRI) is used in combination with 3D center-out trajectory data collection and image subtraction techniques, to yield accurate high spatial resolution images of a material component having an ultra-short transverse relaxation time. In an embodiment, the 3D center-out trajectory can be a 3D rosette k-space trajectory. The disclosed methods can be used to image, among other things, biological materials such as without limitation, cortical and trabecular bones, lung parenchyma, tendons, and ligaments. In a particular example, the disclosed methods are used to image the myelin bilayer in brain white matter, such as for example, to detect, treat, or monitor brain lesions in multiple sclerosis patients.

Description

Attorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 MULTIPLE CONTRAST BIOLOGICAL MATERIAL IMAGING USING THREE- DIMENSIONAL BALANCED STEADY STATE FREE PRECESSION ULTRA-SHORT ECHO TIME MRI CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Application No.63 / 520,032, filed August 16, 2023, which is herein incorporated by reference in its entirety for all purposes. STATEMENT OF GOVERNMENT SUPPORT

[0002] The invention was made with government support under W81XWH-21-1-0399 awarded by the Defense Health Agency, Medical Research and Development Branch. The government has certain rights in the invention. FIELD

[0003] The present disclosure relates to ultra-short echo time (UTE) magnetic-resonance imaging (MRI) techniques, and more particularly to balanced steady state free precession ultra- short echo time (bSSFP UTE) MRI techniques for accurately imaging biological materials that exhibit an ultra-short transverse relaxation time. BACKGROUND

[0004] Depending on the surrounding chemical environment, certain biological materials such as, for example and without limitation, the myelin bilayer in cerebral white matter, cortical and trabecular bones, lung parenchyma, tendons, and ligaments, will exhibit a rapid decay in the transverse magnetization imparted to the protons of the biological material by the radio frequency (RF) energy pulses of a magnetic-resonance imaging (MRI) machine. Such a material may be referred to as having an ultra-short transverse relaxation time (uT2) because the complete decay in transverse magnetization can occur in approximately 1 millisecond or less. Unfortunately, conventional MRI sequences have relatively long echo times – on the order of milliseconds versus less that one millisecond – which renders conventional MRI sequences incapable of properly capturing the fast decaying signals from the protons of biological materials with ultra-short transverse relaxation times. SUMMARY

[0005] Disclosed herein are novel magnetic resonance imaging (MRI) techniques for imaging and analyzing materials that are difficult to accurately image via conventional MRIAttorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 sequences. Such materials may be, but are not necessarily limited to, biological materials that exhibit an ultra-short transverse relaxation time. Examples of imaging system and method embodiments according to the present disclosure employ balanced steady state free precession ultra-short echo time (bSSFP UTE) MRI in combination with 3D rosette k-space trajectory data acquisition and signal subtraction techniques to yield high quality and accurate images of one or more components of the biological material that have an ultra-short transverse relaxation time. For example, accurately imaging the myelin bilayer of a biological brain can be accomplished according to embodiments of the present disclosure by subjecting the brain to a dual-echo bSSFP UTE MRI sequence, collecting the generated MRI data associated with each echo using a 3D rosette k-space trajectory, and subsequently applying a subtraction technique to separate and extract the resulting signal produced by the myelin bilayer from the signals produced by common brain tissues and, particularly, by water. According to example embodiments of the present disclosure, once the signals resulting from the exposure of common brain tissues, water, etc., to bSSFP UTE MRI are subtracted from the overall received signals, what remains is primarily the signal(s) from the myelin bilayer. Thus, a high quality and accurate image of the myelin bilayer of a brain can be generated using the techniques of the present disclosure.

[0006] The data generated by bSSFP UTE MRI-based techniques according to the present disclosure is useable to produce multiple contrast images with high spatial resolution. Also, the bSSFP UTE MRI-based techniques according to the present disclosure require only a very short scan time, particularly in contrast to the scan time associated with conventional MRI techniques. For example, while a conventional MRI sequence may require a total scan time of between thirty minutes to one hour, a bSSFP UTE MRI sequence can be completed in as few as 2-3 minutes. In addition to reducing any patient discomfort or apprehension associated with the imaging process, the significantly shorter scan time also minimizes the likelihood of patient movement, which could have a detrimental effect on image quality.

[0007] Images generated according to method embodiments of the present disclosure can have a variety of applications. For example, and without limitation, the images can be used as a diagnostic tool, as a therapeutic tool, and / or as a monitoring tool. In one non-limiting example, images generated according to method embodiments of the present disclosure can be utilized by clinicians for the purposes of diagnosing, monitoring the progression of, and / or treating multiple sclerosis (MS).Attorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0

[0008] In various embodiments, a computer-implemented method is provided comprising: subjecting a material of interest to a dual-echo balanced steady state free precession ultra-short echo time magnetic resonance imaging (bSSFP UTE MRI) sequence to yield a first echo (TE1) component comprising proton density-weighted signals and ultra-short transverse relaxation time signals and a second echo (TE2) component comprising only proton density-weighted signals; acquiring, by one or more processors, k-space data generated by the dual-echo bSSFP UTE MRI sequence using a center-out k-space trajectory; reconstructing, by the one or more processors, first echo images and second echo images of the material of interest from the acquired k-space data; performing, by the one or more processors, a subtraction operation between the first echo component and the second echo component to remove the proton density-weighted signals from the first echo component and to produce a selectively derived image showing primarily one or more components of the material of interest that have an ultra- short transverse relaxation time; and displaying at least the selectively derived image on a display device for observation by a user.

[0009] In some embodiments, the material of interest is a biological material.

[0010] In some embodiments, the material of interest is a myelin bilayer of a biological brain.

[0011] In some embodiments, the first echo image is a proton density-weighted contrast and is useable for reference and comparison.

[0012] In some embodiments, an echo time associated with the first echo component is less than one millisecond, and an echo time associated with the second echo component is less than three milliseconds.

[0013] In some embodiments, a repetition time associated with the dual-echo bSSFP UTE MRI sequence is less than five milliseconds.

[0014] In some embodiments, a flip angle associated with the dual-echo bSSFP UTE MRI sequence is less than ten degrees.

[0015] In some embodiments, a total scan duration of the dual-echo bSSFP UTE MRI sequence is less than three minutes.

[0016] In some embodiments, the center-out k-space trajectory is a 3D rosette k-space trajectory.

[0017] In some embodiments where the center-out k-space trajectory is a 3D rosette k- space trajectory, the k-space data is sampled in a single acquisition with balanced readout gradients.Attorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0

[0018] In some embodiments, data acquisition using the 3D rosette k-space trajectory excludes crusher gradients.

[0019] In some embodiments, a nonuniform fast Fourier transform (NUFFT) is used to calculate a forward encoding transform of the acquired k-space data prior to performing the subtraction operation.

[0020] In some embodiments, the acquired k-space data is pre-processed prior to performing the subtraction operation.

[0021] In some embodiments, the reconstructed images are subjected to a post-processing procedure prior to performing the subtraction operation, and the post-processing procedure is selected from the group consisting of image registration, feature extraction, normalization, bias-field correction, and combinations thereof.

[0022] In some embodiments, the selectively derived image is used as one or more of a diagnostic tool for identifying a patient condition or conditions; a therapeutic tool for determining and / or directing drug or other treatment protocols; and a monitoring tool that can be employed to observe changes in disease progression over time.

[0023] In some embodiments, the selectively derived image is used for purposes of diagnosing, treating, and / or monitoring the progression of multiple sclerosis (MS) in a human brain.

[0024] In some embodiments, a system is provided that includes one or more processors, and one or more computer-readable media storing instructions which, when executed by the one or more processors, cause the system to perform any of the methods disclosed herein.

[0025] In some embodiments, a computer-program product is provided that is tangibly embodied in one or more non-transitory computer-readable media storing instructions which, when executed by one or more processors, cause a system to perform any of the methods disclosed herein.

[0026] The terms and expressions which have been employed are used as terms of description and not of limitation, and there is no intention in the use of such terms and expressions of excluding any equivalents of the features shown and described or portions thereof, but it is recognized that various modifications are possible within the scope of the invention claimed. Thus, it should be understood that although the present invention as claimed has been specifically disclosed by embodiments and optional features, modification and variation of the concepts herein disclosed may be resorted to by those skilled in the art, andAttorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 that such modifications and variations are considered to be within the scope of this invention as defined by the appended claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The present disclosure is described in conjunction with the appended figures:

[0028] FIG. 1 is a curve representing the signal of the long (weighted) T2 components versus off-resonance phase accumulation of a material immediately after exposure to the radio frequency (RF) energy pulses produced by balanced steady state free precession ultra-short echo time magnetic resonance imaging (bSSFP UTE MRI) at different longitudinal relaxation time-to-transverse relaxation time ratios (T1 / T2) and flip angles (FA), in accordance with various embodiments;

[0029] FIG. 2 illustrates multiple passband-focused signal curves representing the relationship between the ratio of bSSFP signal intensity to proton density (Sss / S0) and the longitudinal relaxation time-to-transverse relaxation time ratio (T1 / T2) of the long (weighted) T2components of a material immediately after exposure to the radio frequency (RF) energy pulses produced by bSSFP UTE MRI at various flip angles (FA), in accordance with various embodiments.

[0030] FIG. 3 graphically depicts the ultra-short transverse relaxation time (uT2) component signals versus flip angles (FA) of a material immediately after exposure to the radio frequency (RF) energy pulses produced by bSSFP UTE MRI in an example where the longitudinal relaxation time (T1,ut2) is 300 milliseconds, in accordance with various embodiments;

[0031] FIG.4 illustrates one example of a 3D rosette k-space trajectory for collecting MRI data, in accordance with various embodiments;

[0032] FIG. 5 is a bSSFP UTE MRI sequence diagram depicting the operation of one example of a 3D rosette trajectory having ten sequential acquisitions, in accordance with various embodiments;

[0033] FIG.6 is a block diagram of an exemplary system embodiment comprising an MRI system having a bSSFP UTE MRI machine and being communicatively coupled to a computing device for performing image processing and analysis operations on data acquired from the bSSFP UTE MRI machine, in accordance with various embodiments;Attorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0

[0034] FIG. 7 depicts reconstructed images showing different views of an ACR phantom device used to test the performance of a bSSFP UTE MRI method, in accordance with various embodiments;

[0035] FIG. 8A depicts reconstructed images of a healthy brain using data produced by a first echo of several dual-echo bSSFP UTE MRI sequences undertaken at multiple different first (short) echo times, in accordance with various embodiments;

[0036] FIG. 8B depicts reconstructed images of the brain shown in FIG. 8A using data produced by a second echo of several dual-echo bSSFP UTE MRI sequences at multiple different second (long) echo times, in accordance with various embodiments;

[0037] FIG. 8C depicts reconstructed images of the brain shown in FIGS. 8A-8B after applying a subtraction operation between each pair of first and second dual echoes, in accordance with various embodiments;

[0038] FIG. 9 graphically illustrates signal intensity versus first echo time for each of a gray matter, white matter and cerebral spinal fluid region of interest within the healthy brain of FIGS.8A-8C;

[0039] FIG. 10 depicts reconstructed first echo, second echo, and post-subtraction operation images of a brain of a test subject known to have multiple sclerosis using data produced by several dual-echo bSSFP UTE MRI sequences undertaken at multiple different first (short) echo times, in accordance with various embodiments;

[0040] FIG. 11 graphically illustrates signal intensity versus first echo time for each of a gray matter, white matter and cerebral spinal fluid region of interest within the multiple sclerosis-affected brain of FIG.10;

[0041] FIG. 12A compares examples of reconstructed brain images acquired from MRI data generated using different types of MRI sequences and associated with a first test subject known to have multiple sclerosis, in accordance with various embodiments;

[0042] FIG. 12B compares examples of reconstructed brain images acquired from MRI data generated using different types of MRI sequences and associated with a second test subject known to have multiple sclerosis, in accordance with various embodiments;

[0043] FIG.13A graphically illustrates various average values of the ratios of brain white matter and brain gray matter to cerebral spinal fluid, as well as average values of a ratio of lesions to cerebral spinal fluid, based on MRI data acquired from various types of MRI sequences for seven test subjects known to have multiple sclerosis, in accordance with various embodiments;Attorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0

[0044] FIG. 13B graphically illustrates the signal intensity of regions of interest within gray matter, white matter, and lesions of the brains of each of the seven test subjects known to have multiple sclerosis, based on images reconstructed from data generated by dual-echo bSSFP UTE MRI sequences, in conjunction with 3D rosette k-space trajectory data acquisition and image echo subtraction, in accordance with various embodiments;

[0045] FIG. 14 is a reconstructed image showing earphones worn by a test subject from data generated by a first echo of a dual-echo bSSFP UTE MRI sequence, in conjunction with 3D rosette k-space trajectory data acquisition, in accordance with various embodiments; and

[0046] FIG.15 is a flowchart representing one embodiment of a method of using dual-echo bSSFP UTE MRI with 3D rosette k-space trajectory data acquisition and image echo subtraction to produce a reconstructed image of a material having at least one component that exhibits an ultra-short transverse relaxation time.

[0047] In the appended figures, similar components and / or features can have the same reference label. Further, various components of the same type can be distinguished by following the reference label by a dash and a second label that distinguishes among the similar components. If only the first reference label is used in the specification, the description is applicable to any one of the similar components having the same first reference label irrespective of the second reference label. DETAILED DESCRIPTION Overview

[0048] Imaging a material component exhibiting an ultra-short transverse relaxation time (uT2) is challenging because the echo times (TEs) required by conventional MRI sequences can far exceed the decay time of signals generated by protons of the uT2material component after excitation by the radio frequency (RF) energy pulses of an MRI machine. In the case of biological tissue materials with uT2relaxation times such as bone, tendons, ligaments, lung parenchyma, and brain myelin, the materials may appear dark or may be invisible when scanned using conventional MRI sequences.

[0049] Ultra-short echo time (UTE) MRI sequences and zero echo time (ZTE) MRI sequences may be better suited to imaging materials having one or more components that exhibit an ultra-short transverse relaxation time. While both techniques may be useable for this purpose, readout gradients are applied prior to transmission of the RF pulses and are always on in the case of ZTE MRI sequences. Further, the central k-space of the data produced by a ZTEAttorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 MRI sequence is empty due to the dead time resulting from switching between RF pulse transmission and data reception, resulting in the need for additional scans or interpolation based on an over-sampling strategy to fill in the missing k-space. Consequently, embodiments of the present disclosure employ UTE MRI sequencing techniques, and more particularly, balanced steady state free precession ultra-short echo time (bSSFP UTE) MRI in combination with 3D center-out trajectory data collection and image subtraction techniques, to yield high quality and accurate images of material components having an ultra-short transverse relaxation time.

[0050] Most UTE MRI sequences require immediate application of the readout gradients after completion of RF pulse emission, as well as center-out k-space data acquisition in each sequence to achieve the shortest possible echo time. The most common k-space trajectory currently used in UTE MRI applications is a 3D radial center-out trajectory. However, and as explained in more detail below, the 3D radial center-out trajectory is limited by the sampling efficiency in the outer k-space of radial acquisitions. Therefore, embodiments according to the present disclosure can utilize a 3D k-space trajectory with a greater curvature per spoke than the 3D radial center-out trajectory, and more particularly, a 3D rosette k-space trajectory. The 3D rosette k-space trajectory has demonstrated an ability during experimentation to provide better image quality than the known 3D radial center-out trajectory when used with UTE MRI, such as bSSFP UTE MRI.

[0051] The use of a 3D rosette k-space data acquisition trajectory in conjunction with bSSFP UTE MRI sequencing and can also reduce the required scan time of an associated bSSFP UTE MRI machine. For example, the 3D rosette k-space trajectory has relatively benign artifacts upon violating the Nyquist criteria, which can allow for faster imaging by employing high under-sampling factors. Additionally, use of the 3D rosette k-space trajectory may reduce scan time while maintaining image quality because the rosette pattern of the 3D rosette k-space trajectory loops around the origin of k-space and can provide better coverage of the k-space with fewer samples. Further, because the 3D rosette k-space trajectory is a center out trajectory, data acquisition emanates from the center of k-space, where the most important spatial frequency information is commonly located. This means that important image details can be acquired early and quickly, which can allow for high-quality images to be obtained with shorter scan times. By employing a pattern that continuously spirals outward, the 3D rosette k-space trajectory can also minimize the dead time associated with scanning of the k-space using other k-space trajectories. The 3D rosette k-space trajectory data acquisition technique can also beAttorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 combined with different parallel imaging techniques that leverage the simultaneous acquisition of multiple lines of k-space using multiple coils, which can further reduce required scan time.

[0052] Of particular, but necessarily exclusive, interest according to the present disclosure, is the application of bSSFP UTE MRI with 3D rosette k-space trajectory data acquisition to the whole brain for purposes of producing images of the myelin sheath. The myelin sheath consists of trapped water (i.e., myelin water) and dry mass, in which the dry mass forms a lipid-protein bilayer that is composed of about 80% lipids and about 20% protein. The myelin sheath plays an important role in normal brain function and development by facilitating long range conduction of electrical impulses across the brain, as well as protecting nerve fibers from injury. Loss or damage of myelin in the brain is implicated in numerous neurological disorders. One such disorder is the neurodegenerative disease of multiple sclerosis (MS), which is associated with gross white matter damage and characteristic foci of demyelination.

[0053] Due to the specific chemical environment of the lipid-protein bilayer-bound protons, approximately 75% of the myelin lipid protons typically exhibit transverse relaxation time (T2) values well below 1 millisecond. Consequently, the signals generated by the myelin lipid protons after excitation by electromagnetic energy decay too quickly to be captured by conventional MRI techniques, which normally have echo times on the order of milliseconds or longer. Despite the minimal required echo time, bSSFP UTE MRI, on the other hand, is capable of capturing the signals emitted by excited myelin lipid protons even at transverse relaxation time values below 1 millisecond, which allows bSSFP UTE MRI to provide direct measurement of myelin.

[0054] Myelin images generated according to embodiments of the present disclosure can have a variety of applications. For example, and without limitation, images generated according to embodiments of the present disclosure can be used as a diagnostic tool by helping to identify a patient condition or conditions, as a therapeutic tool, such as by helping to determine and / or direct drug or other treatment protocols, and as a monitoring tool that can be employed to observe changes in disease progression over time. Also, because bSSFP UTE MRI produces no radiation, it can be safely used in long term monitoring applications.

[0055] In one non-limiting embodiment, images generated according to embodiments of the present disclosure can be utilized by clinicians for the purposes of diagnosing, treating, and / or monitoring the progression of multiple sclerosis (MS), as system and method embodiments of the present disclosure are able to accurately image the lesions that result from demyelination and are indicative of MS. In addition to helping a clinician diagnose and / or monitor theAttorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 progression of MS in a given patient, images generated according to embodiments of the present disclosure can also be used to help direct remyelination therapy.

[0056] Images reconstructed from UTE MRI sequence data typically have a gradient echo- like contrast because of unbalanced readout gradients and / or crusher gradients at the end of each acquisition. However, the combination of UTE MRI techniques with balanced readout gradients and no crusher gradients, results in a bSSFP contrast for components whose transverse relaxation time is longer than the repetition time (TR). Proton signals with an ultra- short transverse relaxation time (uT2) can also be captured using bSSFP UTE MRI.

[0057] According to embodiments of the present disclosure, high quality, accurate images of a biological material of interest, such as but not limited to brain myelin, can be produced by subjecting the material of interest to a dual-echo bSSFP UTE MRI sequence, acquiring the resulting data using a 3D rosette k-space trajectory, and subsequently applying echo subtraction to the images associated with the first and second echoes.

[0058] The dual-echo bSSFP UTE MRI data can be sampled in a single acquisition with balanced readout gradients, and the ultra-short transverse relaxation time proton signals are captured only by the first echo, which has an ultra-short echo time. This results in a first echo (TE1) set of proton density (PD) weighted images that also includes ultra-short transverse relaxation time component signals, and a second echo (TE2) set of PD weighted images without ultra-short transverse relaxation time component signals. Subtraction of the second echo set of images from the first echo set of images can be used to selectively derive images of primarily the biological material component(s) having an ultra-short transverse relaxation time (e.g., myelin).

[0059] When employing small flip angles (FA) of, for example, less than 10 degrees, bSSFP MRI also has the ability to produce proton density (PD) weighted contrast images. A PD weighted contrast image has an advantage in post-processing, such as during registration or segmentation, and also provides a more quantitative reference signal, because there is no longitudinal relaxation time (T1) or transverse relaxation time (T2) weighting. Further, bSSFP MRI sequences have short repetition times and, therefore, can support very high duty cycles, which results in fast and efficient imaging.Attorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 bSSFP MRI Theory

[0060] In the case of bSSFP MRI applied to a biological material, the steady state magnetization (^^^) of the biological material immediately after application of the radiofrequency (RF) energy pulses is defined by the following equation:where ^ is the flip angle, ^ is the off-resonance phase accumulation, magnetization (proportional to the proton density), ^ ൌ ^ି்ோ / ^^ି்ோ / െ 1^^1^ ^^^^^and

[0061] The value of ^^^is highly modulated by off-resonance phase accumulation, which can be separated into passband and stopband. This is illustrated in FIG. 1, which graphically represents the signal of long (weighted) transverse relaxation time components versus off- resonance phase of a material immediately after exposure to the RF energy pulses produced by bSSFP UTE MRI at various longitudinal relaxation time-to-transverse relaxation time ratios (T1 / T2) and flip angles (FA), in accordance with various embodiments.

[0062] With the application of RF energy pulses at alternate positive and negative flip angles, the passband is shifted and centered to an off-resonance phase accumulation range of [-2^ / 3, 2^ / 3]. Given that the width of the passband is proportional to 1 / repetition time (TR), a short repetition time when employing bSSFP MRI is desirable to expand the passband as well as to minimize banding artifacts that are centered around – ^ and ^. As may be further observed in FIG. 1, when the flip angle is large (e.g., 60 degrees), the stopband behaves like signal suppression, which shows up as dark bands in images produced from the bSSFP MRI data. However, when the flip angles are small (e.g., 6 degrees) the banding artifacts are significantly different, consisting instead of one dark band centered between two bright bands.

[0063] If only the signals at the center of the passband are considered, the steady state magnetization caused by the RF energy pulses of bSSFP MRI can be simplified to:

[0064] With a moderate to large flip angle (e.g., larger than 40 degrees), the steady state magnetization is the ratio of transverse relaxation time to longitudinal relaxation time-weightedAttorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 (T2 / T1-weighted). However, if the flip angle is small (e.g., less than 10 degrees), the signal contrast of steady state magnetization is no longer related to the ratio of T2 / T1. Rather, the remaining property that contributes to signal contrast at low flip angles is the proton density (^^), while the flip angle influences the overall signal intensity. This is represented in FIG. 2, which illustrates multiple passband-focused signal curves demonstrating the relationship between the ratio of bSSFP signal intensity to proton density (Sss / S0) and the longitudinal relaxation time-to-transverse relaxation time ratio (T1 / T2) of the long (weighted) T2components of a material immediately after exposure to the RF energy pulses produced by bSSFP UTE MRI at various flip angles (FAs).

[0065] Meanwhile, the signals produced by the components of the material having ultra- short transverse relaxation times (where transverse relaxation time is far less than repetition time) after application of the bSSFP UTE MRI RF energy pulses have a gradient echo-like contrast, as they experience a very different steady-state. While the material components that exhibit a longer transverse relaxation time (T2) have a remaining transverse magnetization by the end of the repetition time (TR) and thus are flipped back across the z-axis, the components of the material that exhibit an ultra-short transverse relaxation time (uT2) have effectively no transverse magnetization by the end of the repetition time.

[0066] These characteristics of the material components having a longer transverse relaxation time (T2) and the material components having an ultra-short transverse relaxation time (uT2), are illustrated in FIG. 3, which graphically depicts the ultra-short transverse relaxation time (uT2) component signals versus flip angles (FA) of a material having a longitudinal relaxation time (T1, ut2) of 300 milliseconds, immediately after exposure to the RF energy pulses produced by bSSFP UTE MRI. As represented in FIG. 3, with such an echo- gradient-like contrast, the material components having an ultra-short transverse relaxation time (uT2) will produce a maximum signal value at the Ernst angle, which corresponds with small flip angles (e.g., less than 10 degrees) and short repetition times (e.g., 5 milliseconds). Data Acquisition

[0067] Two-dimensional (2D) methods for acquiring UTE MRI data are known. However, these 2D methods have limitations, including the presence of eddy currents and imperfect gradients that make it difficult to achieve an appropriate slice selection and a minimized echo time when used. As discussed above, three-dimensional (3D) methods for acquiring UTE MRI data are also known, and the 3D k-space trajectory most commonly employed with UTE MRIAttorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 applications is a 3D radial center-out trajectory. However, the 3D radial center-out trajectory has drawbacks of its own. For example, a conventional 3D radial center-out trajectory samples uniformly along the 1D radius of k-space – i.e., 25% of samples are taken at center k-space, and 25% of samples are taken at peripheral k-space. Such a sampling procedure may be inefficient at the large radii of 3D k-space. In this regard, it has been found, for example, that under circumstances where use of a 3D radial center-out trajectory results in significant (e.g., 80%) undersampling of k-space MRI data, images reconstructed from the acquired MRI data can exhibit signs of noise amplification and may suffer from a loss of structural detail. In addition to providing insufficient sampling at the large radii of a 3D k-space, it has also been determined that suboptimal imaging parameters, such as an insufficient number of spokes and the long repetition time required to match the performance of a 3D dual-echo rosette trajectory when used for UTE MRI, further contributes to the poor results of data acquisition using the 3D radial center-out trajectory.

[0068] In light of the above-described drawbacks associated with use of a 3D radial center- out trajectory for MRI k-space data acquisition, embodiments of the present disclosure preferably utilize novel 3D k-space trajectories with greater curvature per spoke to provide for a more efficient sampling strategy. For example, various embodiments may utilize a 3D rosette k-space trajectory to acquire bSSFP UTE MRI data. A 3D rosette k-space trajectory still allows for a desired center-out sampling pattern, but produces a greater sampling density in the outer k-space than a radial trajectory. In addition, the rosette k-space trajectory acquires MRI data in a more incoherent pattern than a radial trajectory, which is preferred.

[0069] Rosettes are non-Cartesian k-space trajectories with high design flexibility. The shape of a given 3D rosette k-space trajectory can vary, and ultimately depends highly on its defining parameters. A 3D rosette k-space trajectory will produce multiple crossings of the k- space origin, which lends itself well to multiple-echo data acquisition.

[0070] The 3D rosette k-space trajectory is defined by the following equations:^௭^^^ൌ ^^^^^ sin^^^sin^^^^^[6]where kmax is the maximum extent of the k-space,is the frequency of oscillation in the radial direction, ^ଶis the frequency of rotation in the angular direction, ^ determines the location in the z-axis, and ^ determines the initial phase in the angular direction.Attorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0

[0071] One example of a 3D rosette k-space trajectory pattern according to embodiments of the present disclosure is shown in FIG. 4, where ^^and ^ଶare set to be equal. As may be observed, the origin of the k-space is sampled at the beginning and the end of each repetition time, and a petal-like sampling pattern is formed having varied rotations in the kx-ky plane and varied extensions along the kz-axis. Echo time values are separately determined by the time required for two crossings of the k-space origin. The petals of the 3D rosette k-space trajectory pattern are distributed uniformly in the example of FIG. 4. However, with a golden angle acquisition or a randomized distribution, the incoherence of the 3D rosette k-space trajectory can be further increased, which may be advantageous relative to compressed sensing techniques, the performance of which are highly related to the incoherence of the k-space pattern.

[0072] One example of a bSSFP UTE MRI sequence diagram utilizing a 3D rosette k-space trajectory for acquiring the MRI data according to embodiments of the present disclosure is illustrated in Figure 5. In the example of FIG. 5, ten sequential acquisitions are depicted. Analog-to-digital converter (ADC) signals, the bSSFP UTE MRI RF energy pulses, the gradients (Gx, Gy, Gz in the unit of mT / m) in all directions, and the k-space trajectories of the first acquisition (^) and last acquisition (^) are also graphically illustrated.

[0073] Unlike a conventional bSSFP MRI sequence where the echo time (TE) is located in the center of the repetition time (TR), it can be observed in FIG. 5 that in the case of the illustrated bSSFP UTE MRI with a 3D rosette k-space trajectory sequence, the readout gradient begins at the k-space origin to produce a first echo (TE1) that is as short as possible. A second echo (TE2) is also acquired as the 3D rosette k-space trajectory pattern returns to the center of k-space at the end of the repetition time. It can be additionally observed that the gap between the end of acquisition to the next bSSFP UTE MRI RF energy pulse is kept as short as possible to minimize the repetition time so as to maintain a high duty cycle and reduce banding artifacts in the images subsequently generated from the acquired MRI data.

[0074] By utilizing a 3D rosette k-space trajectory for data acquisition, dual-echo images can be generated within a single acquisition, with a manual separation at the middle of each data readout. A more detailed discussion of the 3D rosette k-space trajectory, including the influence of the trajectory parameters, Nyquist criteria, and the specific gradient ramp-up of the 3D rosette k-space trajectory, can be obtained from the dissertation of Xin Shen titled Development and Applications of 3d Ultra-Short Echo Time MRI with Rosette K-Space Pattern, which was submitted to the faculty of the Weldon School of Biomedical EngineeringAttorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 at Purdue University in August 2022, and is incorporated by reference in this application for all purposes as if fully written herein. System

[0075] FIG. 6 is a block diagram of one exemplary embodiment of a system 100 for producing accurate, multiple contrast, high spatial resolution images of one or more components of a biological material that exhibit an ultra-short transverse relaxation time. While FIG. 6 depicts the system 100 (and a computing device 110 of the system 100) as including certain components, other exemplary system embodiments may involve more, fewer, or different components than are shown in FIG.6.

[0076] The exemplary embodiment of the system 100 of FIG. 6 is shown to include a magnetic resonance imaging (MRI) system 102 comprising a balanced steady state free precession ultra-short echo time magnetic resonance imaging (bSSFP UTE MRI) machine 104. The bSSFP UTE MRI machine 104 may include an MRI machine controller 106 that controls, among other operations, various parameters of the radio frequency (RF) energy pulses emitted by magnets of the bSSFP UTE MRI machine 104. The MRI machine controller 106 may be communicatively coupled to an MRI operator computer 108 or a similar system that allows an MRI machine operator to control the MRI bSSFP UTE MRI machine 104 through the MRI machine controller 106, to monitor the MRI operation, to communicate with a subject being imaged by the MRI machine controller bSSFP UTE MRI machine 104, etc.

[0077] The system 100 may further include a computing device 110 that is communicatively coupled to the MRI system 102 for receiving and processing dual-echo bSSFP UTE MRI data generated during an MRI sequence. In the exemplary embodiment shown in FIG. 6, the computing device 110 is communicatively coupled to the MRI system 102 by a network 112. When present, the network 112 may be, for example, a local area network (LAN), a wide-area network (WAN) such as the Internet, an institutional network, cellular or other wireless networks, etc. In other embodiments, it may be possible to transfer MRI data from the MRI system 102 to the computing device 110 using a transportable non- transitory computer-readable medium such as, for example, an electronic, optical, magnetic, or other data storage device, which can be read by the computing device 110.

[0078] As shown, this exemplary embodiment of the computing device 110 includes a processor 114 communicatively coupled to a memory 116 by a bus 118. The processor 114 can include one processor or multiple processors. Non-limiting examples of the processor 114Attorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 include a Field-Programmable Gate Array (FPGA), an application specific integrated circuit (ASIC), a microprocessor, or any combination of these. Instructions 120 may be stored in the memory 116. The instructions are executable by the processor 114 for causing the processor to perform various operations. In some examples, the instructions 120 can include processor specific instructions generated by a compiler or an interpreter from code written in any suitable computer-programming language, such as C, C++, C#, or Java.

[0079] The memory 116 can include one memory device or multiple memory devices. The memory 116 can be non-volatile and may include any type of memory device that retains stored information when powered off. Non-limiting examples of the memory 116 include electrically erasable and programmable read-only memory (EEPROM), flash memory, or any other type of non-volatile memory. At least some of the memory device includes a non-transitory computer-readable medium from which the processor 114 can read instructions 120. A non- transitory computer-readable medium can include electronic, optical, magnetic, or other storage devices capable of providing the processor 114 with the instructions 120 or other program code. Non-limiting examples of a non-transitory computer-readable medium include magnetic disk(s), memory chip(s), ROM, random-access memory (RAM), an ASIC, a configured processor, optical storage, or any other medium from which a computer processor can read the instructions 120.

[0080] By executing the instructions 120, the computing device 110 can process the data received from the MRI system 102 in the manner described above, and as described in more detail below. The computing device 110 may be provided with various applications / programs / modules for performing desired processing operations. For example, embodiments of the computing device 110 may include a data acquisition module 122 that allows the processor 114 to perform a data acquisition operation on the received MRI data, which may be a 3D rosette k-space trajectory data acquisition operation. Likewise, exemplary embodiments of the computing device 110 may include a pre-processing module 124 that allows the processor 114 to perform pre-processing operations on the MRI data prior to image reconstruction, and an image reconstruction module 126 that allows the processor 114 to construct images of the sequenced material from the acquired MRI data. Embodiments of the computing device 110 may also include a subtraction module 128 that allows the processor 114 to perform a subtraction operation between a first echo component and a second echo component produced during a dual-echo MRI sequence of the bSSFP UTE MRI machine 104, to remove proton density-weighted signals from the first echo component and to produce aAttorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 selectively derived image showing primarily one or more components of the scanned material that exhibit an ultra-short transverse relaxation time. Embodiments of the computing device 110 may additionally include a post-processing module 130 that allows the processor 114 to perform post-processing operations on the reconstructed images, such as but not limited to, operations comprising image registration, feature extraction, normalization, bias-field correction, and combinations thereof.

[0081] Embodiments of the computing device 110 can further include an input device 132 or devices that may be communicatively coupled to the processor 114 through the bus 118, and allow a user to interact with and control the computing device 110. One or more display devices 134 may also be communicatively coupled to the processor 114 through the bus 118, such that reconstructed images and other information can be presented to a user, such as an MRI operator or a clinician, for purposes not limited to image analysis and resulting diagnostic, therapeutic, and / or monitoring determinations. Acquired MRI data, reconstructed images, application or system data, etc., may also be stored in a data storage device 136 that is associated with the computing device 102 and may be communicatively coupled to the processor 114 via the bus 118. Similar or other data may instead be, or may additionally be, stored in a locally or remotely located database 138, which may also be communicatively coupled to the processor 114 via the bus 118. Resolution Verification

[0082] Prior to scanning any test subjects, an exemplary embodiment of a bSSFP UTE MRI protocol was evaluated to determine if images of acceptable spatial resolution could be produced. In this regard, FIG. 7 illustrates reconstructed images of several different views of an ACR phantom test device based on data generated by corresponding bSSFP UTE MRI sequences. The reconstructed images clearly show the intrinsic structure of the phantom test device, without visible artifacts. In addition, it can be observed that the bSSFP UTE MRI sequence was able to separate holes in the phantom test device having a 1.0 millimeter diameter and spacing (bottom row, pointed by red arrows), which supports the disclosure that bSSFP UTE MRI can be used to produce images having an isotropic spatial resolution of less than 1.0 millimeter. Image Reconstruction and Post-Processing

[0083] The image reconstruction process according to embodiments of the present disclosure can be accomplished using a variety of techniques. For example, a Fourier transformAttorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 method such as a fast Fourier transform (FFT) method, or another method, may initially be used to decompose and resolve (decode) the acquired encoded k-space MRI data, whereafter a reconstruction technique such as, but not necessarily limited to, a gridding technique, can be used for image reconstruction.

[0084] Similarly, post-processing procedures may vary. For example, the post-processing steps applied to reconstructed images according to embodiments of the present disclosure may include, without limitation, one or more of subtraction, registration, feature extraction, normalization, bias-field correction, segmentation, and noise reduction. The post-processing technique(s) employed may depend to at least some degree on the particular nature of the imaged material. Experiment

[0085] An in vivo experiment was conducted to evaluate the effectiveness of using embodiments of the above-described bSSFP MRI with 3D rosette k-space trajectory data acquisition and image subtraction method for imaging the myelin sheath of the human brain. During the experiment, the brains of six healthy subjects and eight subjects known to have multiple sclerosis (MS) were subjected to dual-echo bSSFP UTE MRI sequences. The MRI system utilized to perform the scans was a Skyra whole-body 3T MRI scanner from Siemens in Erlangen, Germany. Additionally, a vendor-supplied 64-channel receiver head coil was used for all subjects. The operational parameters associated with the dual-echo bSSFP UTE MRI sequences of this experiment comprised: a field of view (FOV) of 240 x 240 x 240 mm3, a matrix size of 256 x 256 x 256, a spatial resolution of 0.94 x 0.94 x 0.94 mm3, a readout dwell time of 5 microseconds (with the usage of duplicated data), a flip angle of 6 degrees, a first echo time (TE1) of 40 microseconds, a second echo time (TE2) of 2.2 milliseconds, a repetition time (TR) of 2.4 milliseconds, a readout duration of 2.16 milliseconds, and a RF energy pulse duration of 10 microsecond.

[0086] Data generated by the dual-echo bSSFP UTE MRI sequences was acquired using a 3D rosette k-space trajectory, as previously described. In this example, 71,820 petals were acquired for one 3D rosette dual-echo acquisition operation. This resulted in a k-space coverage of 40% compared to the Nyquist criteria. The total scan duration of each dual-echo bSSFP UTE MRI sequence was 2.9 minutes (i.e., 71,820 petals x 2.4 milliseconds).

[0087] Image reconstruction and post-processing steps were performed in this example using MATLAB (MathWorks, USA). A nonuniform fast Fourier transform (NUFFT) was usedAttorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 to calculate a forward encoding transform of the acquired k-space data. A regular gridding method applying a density compensated adjoint NUFFT was used for image reconstruction. Phase information was extracted for coil combination after a coil-by-coil reconstruction. Other techniques and software may be used in other examples.

[0088] The post-processing procedures performed in this example included image registration to T1-weighted anatomy images of the human brain, brain tissue extraction, and bias-field correction. Other examples may include other post-processing procedures. In this example, the identified post-processing procedures were performed using the FSL (i.e., FMRIB Software Library) software package. Additionally, the reconstructed images associated with the second echoes were normalized using a cerebral spinal fluid (CSF) signal as reference to compensate for minor k-space misalignment from eddy currents and / or system delays. One effect of such a misalignment is overall intensity changes due to center k-space errors, which are corrected by this normalization. A subtraction operation by which the second echo images were subtracted from the first echo images (i.e., TE1-TE2) was performed to selectively produce images showing primarily brain components that exhibit ultra-short transverse relaxation times (e.g., myelin).

[0089] For purposes of further understanding the signal decay in the sequences, three of the healthy subjects and one of the MS subjects also underwent additional dual-echo bSSFP UTE MRI scans with extended repetition time at various echo times. In this example, the extended repetition time (TR) was 3 milliseconds, at first echo times (TE1s) of 40, 100, 200, 400, and 600 microseconds, and second echo times (TE2s) of 2.2, 2.26, 2.36, 2.56, and 2.76 milliseconds.

[0090] FIGS.8A-8C illustrate reconstructed images generated from dual-echo bSSFP UTE MRI sequence data associated with a selected one of the three healthy subjects who underwent the additional dual-echo bSSFP UTE MRI scans. Particularly, FIG. 8A depicts right sagittal, coronal, and axial first echo (TE1) images of the brain of the selected subject generated from dual-echo bSSFP UTE MRI data acquired using a 3D rosette k-space trajectory and subjected to the reconstruction and post-processing steps described above. As indicated therein, the five sets of first echo (TE1) images shown in FIG.8A correspond to separate dual-echo bSSFP UTE MRI sequences having first echo times (TE1s) of 40, 100, 200, 400, and 600 microseconds, respectively. Each of the individual dual-echo bSSFP UTE MRI sequences used to generate the first echo (TE1) data from which the images of FIG.8A were reconstructed was performed with an extended repetition time (TR) of 3 milliseconds.Attorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0

[0091] FIG. 8B depicts right sagittal, coronal, and axial second echo (TE2) images of the brain of the selected subject generated from dual-echo bSSFP UTE MRI data acquired using a 3D rosette k-space trajectory and subjected to the reconstruction and post-processing steps described above. As indicated therein, the five sets of images shown in FIG.8B correspond to separate dual-echo bSSFP UTE MRI sequences having second echo times (TE2s) of 2.2, 2.26, 2.36, 2.56, and 2.76 milliseconds, respectively. Each of the individual dual-echo bSSFP UTE MRI sequences used to generate the second echo (TE2) data from which the images of FIG.6A were reconstructed was performed with an extended repetition time (TR) of 3 milliseconds.

[0092] FIG. 8C depicts the right sagittal, coronal, and axial images of the brain of the selected subject shown in FIG. 8A and 8B after a subtraction operation has been performed. More specifically, each of the five sets of images shown in FIG. 8C were generated by subtracting the second echo images associated with a given first and second echo time (e.g., TE1= 100 microseconds, TE2= 2.26 milliseconds) from the first echo images associated with the same first and second echo time (in this example, TE1= 100 microseconds, TE2= 2.26 milliseconds). After the subtraction operation, the resulting images of FIG. 8C represent primarily brain components that exhibit ultra-short transverse relaxation times (e.g., myelin).

[0093] For purposes of comparison, other types of MRI sequences were also performed on most of the subjects. In this example, these other sequences included: a Magnetization- Prepared Rapid Acquisition Gradient Echo (MPRAGE) sequence having a repetition time (TR) of 2.3 seconds, an echo time (TE) of 2.98 milliseconds, and an inversion time (TI) of 0.9s; a T2 Weighted Fluid Attenuated Inversion Recovery (T2-FLAIR) sequence with a repetition time (TR) of 5 seconds, an echo time (TE) of 388 milliseconds, and an inversion time (TI) of 1.6 seconds); and a proton density-weighted (PD-weighted) sequence having a repetition time (TR) of 3 seconds, and an echo time (TE) of 11 milliseconds.

[0094] FIG.9 graphically illustrates proton signal intensity versus first echo time for each of a gray matter, white matter, and cerebral spinal fluid region of interest within the healthy brain of FIGS. 8A-8C as a result of exposure to the RF energy pulses produced during execution of a bSSFP UTE MRI protocol according to an example of the present disclosure. More specifically, FIG. 9 presents decay curves associated with the proton signals produced by each of the aforesaid gray matter, white matter, and cerebral spinal fluid regions of interest during the first echo time of a dual-echo bSSFP UTE MRI sequence. As may be observed, the cerebral spinal fluid region of interest produces highest signal intensity and experiences almost no signal drop over the course of the first echo time. In contrast, the white matter region ofAttorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 interest produces the lowest signal intensity and experiences the largest signal drop over the course of the first echo time. Mono-exponential fitting indicates that the ultra-short transverse echo time (uT2) components of the imaged brain were responsible for about 20% of the signal intensity in the white matter at a first echo time (TE1) of 40 microseconds, which has an average ultra-short transverse echo time (uT2) value of approximately 500 microseconds.

[0095] FIG. 10 illustrates reconstructed images generated from dual-echo bSSFP UTE MRI sequence data associated with the aforementioned MS subject who underwent the additional dual-echo bSSFP UTE MRI scans. Particularly, FIG. 10 depicts coronal first echo (TE1) images, second echo (TE2) images, and post-subtraction operation images of the brain of the MS subject who underwent the additional dual-echo bSSFP UTE MRI scans. The images were generated from dual-echo bSSFP UTE MRI data acquired using a 3D rosette k-space trajectory and subjected to the reconstruction and post-processing steps described above.

[0096] As indicated in FIG. 10, the five first echo (TE1) images correspond to separate dual-echo bSSFP UTE MRI sequences having first echo times (TE1s) of 40, 100, 200, 400, and 600 microseconds, respectively. The five second echo (TE2) images correspond to separate dual-echo bSSFP UTE MRI sequences having second echo times (TE2s) of 2.2, 2.26, 2.36, 2.56, and 2.76 milliseconds, respectively. Each of the individual dual-echo bSSFP UTE MRI sequences used to generate the first echo (TE1) data and the second echo (TE2) data from which the images of FIG.10 was reconstructed was performed with an extended repetition time (TR) of 3 milliseconds. Each of the five post-subtraction operation images shown in FIG. 10 was generated by subtracting the second echo images associated with a given first and second echo time (e.g., TE1 = 100 microseconds, TE2 = 2.26 milliseconds) from the first echo images associated with the same first and second echo time (in this example, TE1 = 100 microseconds, TE2 = 2.26 milliseconds). The post-subtraction operation images of FIG.10 represent primarily brain components that exhibit ultra-short transverse relaxation times (e.g., myelin).

[0097] As can be observed in FIG.10, a lesion is identifiable and is indicated by arrows in all the first echo (TE1) images and the second echo (TE2) images. The lesion is also clearly identifiable and indicated by arrows in the post-subtraction operation images having shorter first echo times (TE1s) between 40 to 400 microseconds. The lesion is not identifiable in thepost-subtraction operation image corresponding to the longer first echo time (TE1) of 600microseconds, as the signal intensity from the ultra-short transverse relaxation time (uT2)components is almost fully decayed at that point.Attorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0

[0098] FIG.11 graphically illustrates proton signal intensity versus first echo time for each of a gray matter, white matter, and cerebral spinal fluid region of interest within the MS-affected brain of FIG.10 as a result of exposure to the RF energy pulses produced during execution of a bSSFP UTE MRI protocol according to an example of the present disclosure. More specifically, FIG.11 presents decay curves associated with the proton signals produced by each of the aforesaid gray matter, white matter, and cerebral spinal fluid regions of interest during the first echo time of a dual-echo bSSFP UTE MRI sequence. As may be observed in FIG.11, the cerebral spinal fluid region of interest produces highest signal intensity and experiences only a minimal signal drop over the course of the first echo time. In contrast, the white matter region of interest produces the lowest signal intensity and experiences the largest signal drop over the course of the first echo time. Mono-exponential fitting indicates that the ultra-short transverse echo time (uT2) components contribute about 22% of the signal intensity in the white matter at an average ultra-short transverse echo time (uT2) value of approximately 400 to 500 microseconds.

[0099] FIGS.12A-12B show reconstructed brain images for two of the seven subjects with multiple sclerosis (MS). The reconstructed brain images were generated from the different types of MRI sequence data identified above. More specifically, each of FIG. 12A and 12B shows a right sagittal, coronal, and axial image generated from, in a top-to-bottom order of appearance, a T2-FLAIR MRI sequence, a MPRAGE MRI sequence, a conventional PD MRI sequence, a dual-echo bSSFP UTE MRI first echo sequence with 3D rosette k-space trajectory data acquisition, a dual-echo bSSFP UTE MRI second echo sequence with 3D rosette k-space trajectory data acquisition, and a dual-echo bSSFP UTE MRI sequence with 3D rosette k-space trajectory after application of a subtraction operation on the first echo and second echo images to produce resulting images that represent primarily brain components that exhibit ultra-short transverse relaxation times. All of the images in FIGS. 12A-12B were registered into the MPRAGE image atlas.

[0100] A review of the T2-FLAIR brain images for each of the two subjects with MS represented in FIGS. 12A-12B reveals some degree of visible white matter lesions, which are also indicated by red arrows. The same lesions are also visible and indicated by red arrows in each of the dual-echo bSSFP UTE MRI first echo images and the dual-echo bSSFP UTE MRI second echo images, and by white arrows in the dual-echo bSSFP UTE MRI images resulting from application of a subtraction operation on the first echo and second echo images. This verifies that dual-echo bSSFP UTE MRI with 3D rosette k-space trajectory data acquisitionAttorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 can be used to accurately image brain tissues that exhibit ultra-short transverse relaxation times. It may also be observed in FIGS.12A-12B that the contrast of the images generated using data obtained from the dual-echo bSSFP UTE MRI first echo and second echo images is substantially similar to the contrast of the images generated using data obtained from the conventional PD MRI sequence. This further verifies the viability of dual-echo bSSFP UTE MRI with 3D rosette k-space trajectory data acquisition for imaging myelin or other brain tissues that exhibit ultra-short transverse relaxation times.

[0101] For purposes of experimentation, the aforementioned different regions of interest located within the white matter (WM), gray matter (GM), and cerebral spinal fluid (CSF) of the brains of the subjects were selected, and masks of the lesions were generated by a clinician. The average ratios of WM / CSF, GM / CSF, and lesions / CSF were reported and compared across the different MRI sequences – e.g., the first echoes and the second echoes from the bSSFP UTE MRI sequences, the T2-FLAIR MRI sequences, the MPRAGE MRI sequences, and the conventional PD-weighted MRI sequences.

[0102] FIG. 13A graphically summarizes the average ratios of WM / CSF, GM / CSF, and lesions / CSF as determined from data generated by multiple bSSFP UTE MRI sequences on seven test subjects known to have multiple sclerosis. Because the fluid was suppressed by inversion recovery, the ratios are higher in the T2-FLAIR MRI sequences than in the other sequences. As expected, the various ratios are close to but less than one for each of the PD- weighted sequences, as well as each of the first echo and second echo bSSFP UTE MRI sequences.

[0103] FIG. 13B graphically summarizes the proton signal intensity exhibited by regions of interest within gray matter, white matter, and lesions of the brains of each of the seven test subjects, based on images reconstructed from data generated by dual-echo bSSFP UTE MRI sequences, in conjunction with 3D rosette k-space trajectory data acquisition and image echo subtraction, according to various embodiments. The bSSFP UTE subtraction signal intensity from the regions of interest in the gray matter, white matter, and lesions is quantified in FIG. 13B, and is expected to result primarily from components within the regions of interest that exhibit ultra-short transverse relaxation times. The subtraction signal intensity in the white matter (WM) is indicated to be about 0.05, which is significantly higher than in the gray matter (GM) (around 0.03, P<0.001), and in the lesions (around 0.03-0.04, P<0.001) for each of the seven test subjects.Attorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0

[0104] As further evidence that the use of a dual-echo bSSFP UTE MRI sequence, in conjunction with 3D rosette k-space trajectory data acquisition and image echo subtraction according to various embodiments can result in accurate, high spatial resolution reconstructed images of materials that exhibit ultra-short transverse relaxation times. FIG. 14 depicts a reconstructed image of earphones worn by a test subject during a dual-echo bSSFP UTE MRI sequence. Notably, the material from which the earphones of constructed exhibits an ultra-short transverse relaxation time after excitation by MRI RF energy.

[0105] FIG. 15 is a flowchart representing one embodiment of a computer-implemented method for using dual-echo bSSFP UTE MRI in combination with 3D rosette k-space trajectory data acquisition and image echo subtraction, to produce high quality and accurate images of materials (e.g., biological materials) having at least one component of interest that exhibits an ultra-short transverse relaxation time. As represented in FIG. 15, at block 200, a material of interest is subjected to a dual-echo bSSFP UTE MRI sequence to yield a first echo (TE1) component comprising proton density-weighted signals and ultra-short transverse relaxation time signals and a second echo (TE2) component comprising only proton density-weighted signals. At block 202, k-space data generated by the dual-echo bSSFP UTE MRI sequence is acquired using a 3D rosette k-space trajectory. At block 204, first echo images and second echo images of the material of interest are reconstructed from the acquired k-space data, and at block 206, a subtraction operation is performed between the first echo component and the second echo component to remove the proton density-weighted signals from the first echo component and to produce a selectively derived image showing primarily one or more components of the material of interest that have an ultra-short transverse relaxation time. Additional Considerations

[0106] Some embodiments of the present disclosure include a system including one or more processors. In some embodiments, the system includes one or more computer-readable media storing instructions which, when executed by the one or more processors, cause the system to perform part or all of one or more methods and / or part or all of one or more processes disclosed herein. Some embodiments of the present disclosure include a computer-program product tangibly embodied in One or more non-transitory computer-readable media storing instructions which, when executed by one or more processors, cause a system to perform part or all of one or more methods and / or part or all of one or more processes disclosed herein.Attorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0

[0107] The terms and expressions which have been employed are used as terms of description and not of limitation, and there is no intention in the use of such terms and expressions of excluding any equivalents of the features shown and described or portions thereof, but it is recognized that various modifications are possible within the scope of the invention claimed. Thus, it should be understood that although the present invention as claimed has been specifically disclosed by embodiments and optional features, modification and variation of the concepts herein disclosed may be resorted to by those skilled in the art, and that such modifications and variations are considered to be within the scope of this invention as defined by the appended claims.

[0108] The ensuing description provides preferred exemplary embodiments only, and is not intended to limit the scope, applicability or configuration of the disclosure. Rather, the ensuing description of the preferred exemplary embodiments will provide those skilled in the art with an enabling description for implementing various embodiments. It is understood that various changes may be made in the function and arrangement of elements without departing from the spirit and scope as set forth in the appended claims.

[0109] Specific details are given in the following description to provide a thorough understanding of the embodiments. However, it will be understood that the embodiments may be practiced without these specific details. For example, circuits, systems, networks, processes, and other components may be shown as components in block diagram form in order not to obscure the embodiments in unnecessary detail. In other instances, well-known circuits, processes, algorithms, structures, and techniques may be shown without unnecessary detail in order to avoid obscuring the embodiments.

Claims

Attorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 CLAIMS What is claimed is:

1. A computer-implemented method comprising: subjecting a material of interest to a dual-echo balanced steady state free precession ultra-short echo time magnetic resonance imaging (bSSFP UTE MRI) sequence to yield a first echo (TE1) component comprising proton density-weighted signals and ultra-short transverse relaxation time signals and a second echo (TE2) component comprising only proton density- weighted signals; acquiring, by one or more processors, k-space data generated by the dual-echo bSSFP UTE MRI sequence using a center-out k-space trajectory; reconstructing, by the one or more processors, a first echo image and a second echo image of the material of interest from the acquired k-space data; performing, by the one or more processors, a subtraction operation between the first echo component and the second echo component to remove the proton density-weighted signals from the first echo component and to produce a selectively derived image showing primarily one or more components of the material of interest that have an ultra-short transverse relaxation time; and displaying at least the selectively derived image on a display device for observation by a user.

2. The computer-implemented method of claim 1, wherein the material of interest is a biological material.

3. The computer-implemented method of claim 2, wherein the biological material is a myelin sheath of a brain.

4. The computer-implemented method of claim 1, wherein the first echo image is a proton density-weighted contrast and is used for reference and comparison.

5. The computer-implemented method of claim 1, wherein an echo time associated with the first echo component is less than one millisecond, and an echo time associated with the second echo component is less than three milliseconds.Attorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 6. The computer-implemented method of claim 1, wherein a repetition time associated with the dual-echo bSSFP UTE MRI sequence is less than five milliseconds.

7. The computer-implemented method of claim 1, wherein a flip angle associated with the dual-echo bSSFP UTE MRI sequence is less than ten degrees.

8. The computer-implemented method of claim 1, wherein a total scan duration of the dual-echo bSSFP UTE MRI sequence is less than three minutes.

9. The computer-implemented method of claim 1, wherein the center-out k-space trajectory is a 3D rosette k-space trajectory.

10. The computer-implemented method of claim 9, wherein the k-space data is sampled in a single acquisition with balanced readout gradients.

11. The computer-implemented method of claim 9, wherein data acquisition using the 3D rosette k-space trajectory excludes crusher gradients.

12. The computer-implemented method of claim 1, further comprising using a nonuniform fast Fourier transform (NUFFT) to calculate a forward encoding transform of the acquired k- space data prior to performing the subtraction operation.

13. The computer-implemented method of claim 1, further comprising pre-processing the acquired k-space data prior to performing the subtraction operation.

14. The computer-implemented method of claim 1, further comprising performing a post- processing procedure on the reconstructed images prior to performing the subtraction operation, the post-processing procedure selected from the group consisting of image registration, feature extraction, normalization, bias-field correction, and combinations thereof.

15. The computer-implemented method of claim 1, wherein the selectively derived image is used as one or more of: a diagnostic tool for identifying a patient condition or conditions;Attorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 a therapeutic tool for determining and / or directing drug or other treatment protocols; and a monitoring tool that can be employed to observe changes in disease progression over time.

16. The computer-implemented method of claim 13, wherein the selectively derived image is used for purposes of diagnosing, treating, and / or monitoring a progression of multiple sclerosis (MS) in a human brain.

17. A system comprising: one or more processors; and one or more computer-readable media storing instructions which, when executed by the one or more processors, cause the system to perform operations comprising: subjecting a material of interest to a dual-echo balanced steady state free precession ultra-short echo time magnetic resonance imaging (bSSFP UTE MRI) sequence to yield a first echo (TE1) component comprising proton density-weighted signals and ultra-short transverse relaxation time signals and a second echo (TE2) component comprising only proton density-weighted signals; acquiring, by one or more processors, k-space data generated by the dual-echo bSSFP UTE MRI sequence using a center-out k-space trajectory; reconstructing, by the one or more processors, a first echo image and a second echo image of the material of interest from the acquired k-space data; performing, by the one or more processors, a subtraction operation between the first echo component and the second echo component to remove the proton density-weighted signals from the first echo component and to produce a selectively derived image showing primarily one or more components of the material of interest that have an ultra-short transverse relaxation time; and displaying at least the selectively derived image on a display device for observation by a user.

18. The system of claim 17, wherein the material of interest is a biological material.

19. The system of claim 18, wherein the biological material is a myelin sheath of a brain.Attorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 20. The system of claim 17, wherein the first echo image is a proton density-weighted contrast and is usable for reference and comparison.

21. The system of claim 17, wherein an echo time associated with the first echo component is less than one millisecond, and an echo time associated with the second echo component is less than three milliseconds.

22. The system of claim 17, wherein a repetition time associated with the dual-echo bSSFP UTE MRI sequence is less than five milliseconds.

23. The system of claim 17, wherein a flip angle associated with the dual-echo bSSFP UTE MRI sequence is less than ten degrees.

24. The system of claim 17, wherein a total scan duration of the dual-echo bSSFP UTE MRI sequence is less than three minutes.

25. The system of claim 17, wherein the center-out k-space trajectory is a 3D rosette k- space trajectory.

26. The system of claim 25, wherein the k-space data is sampleable in a single acquisition with balanced readout gradients.

27. The system of claim 26, wherein data acquisition using the 3D rosette k-space trajectory excludes crusher gradients.

28. The system of claim 17, wherein the operations further comprise using a nonuniform fast Fourier transform (NUFFT) to calculate a forward encoding transform of the acquired k- space data prior to performing the subtraction operation. The system of claim 17, wherein the operations further comprise pre-processing the acquired k-space data prior to performing the subtraction operation.Attorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 30. The system of claim 17, wherein the operations further comprise performing a post- processing procedure on the reconstructed images prior to performing the subtraction operation, the post-processing procedure selected from the group consisting of image registration, feature extraction, normalization, bias-field correction, and combinations thereof.

31. The system of claim 17, wherein the selectively derived image is usable as one or more of: a diagnostic tool for identifying a patient condition or conditions; a therapeutic tool for determining and / or directing drug or other treatment protocols; and a monitoring tool that can be employed to observe changes in disease progression over time.

32. The system of claim 31, wherein the selectively derived image is usable for purposes of diagnosing, treating, and / or monitoring a progression of multiple sclerosis (MS) in a human brain.

33. One or more non-transitory computer-readable media storing instructions which, when executed by one or more processors, cause a system to perform operations comprising: subjecting a material of interest to a dual-echo balanced steady state free precession ultra-short echo time magnetic resonance imaging (bSSFP UTE MRI) sequence to yield a first echo (TE1) component comprising proton density-weighted signals and ultra-short transverse relaxation time signals and a second echo (TE2) component comprising only proton density- weighted signals; acquiring, by one or more processors, k-space data generated by the dual-echo bSSFP UTE MRI sequence using a center-out k-space trajectory; reconstructing, by the one or more processors, a first echo image and a second echo image of the material of interest from the acquired k-space data; performing, by the one or more processors, a subtraction operation between the first echo component and the second echo component to remove the proton density-weighted signals from the first echo component and to produce a selectively derived image showing primarily one or more components of the material of interest that have an ultra-short transverse relaxation time; andAttorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 displaying at least the selectively derived image on a display device for observation by a user.

34. The one or more non-transitory computer-readable media of claim 33, wherein the material of interest is a biological material.

35. The one or more non-transitory computer-readable media of claim 34, wherein the biological material is a myelin sheath of a brain.

36. The one or more non-transitory computer-readable media of claim 33, wherein the first echo image is a proton density-weighted contrast and is useable for reference and comparison.

37. The one or more non-transitory computer-readable media of claim 33, wherein an echo time associated with the first echo component is less than one millisecond, and an echo time associated with the second echo component is less than three milliseconds.

38. The one or more non-transitory computer-readable media of claim 33, wherein a repetition time associated with the dual-echo bSSFP UTE MRI sequence is less than five milliseconds.

39. The one or more non-transitory computer-readable media of claim 33, wherein a flip angle associated with the dual-echo bSSFP UTE MRI sequence is less than ten degrees.

40. The one or more non-transitory computer-readable media of claim 33, wherein a total scan duration of the dual-echo bSSFP UTE MRI sequence is less than three minutes.

41. The one or more non-transitory computer-readable media of claim 33, wherein the center-out k-space trajectory is a 3D rosette k-space trajectory.

42. The one or more non-transitory computer-readable media of claim 41, wherein the k- space data is sampleable in a single acquisition with balanced readout gradients.Attorney Docket No.: 081906-1458966-252810WO UCSF Docket No.: SF-2023-168-2-PCT-0 43. The one or more non-transitory computer-readable media of claim 41, wherein data acquisition using the 3D rosette k-space trajectory excludes crusher gradients.

44. The one or more non-transitory computer-readable media of claim 33, wherein the operations further comprise using a nonuniform fast Fourier transform (NUFFT) to calculate a forward encoding transform of the acquired k-space data prior to performing the subtraction operation.

45. The one or more non-transitory computer-readable media of claim 33, wherein the operations further comprise pre-processing the acquired k-space data prior to performing the subtraction operation.

46. The one or more non-transitory computer-readable media of claim 33, wherein the operations further comprise performing a post-processing procedure on the reconstructed images prior to performing the subtraction operation, the post-processing procedure selected from the group consisting of image registration, feature extraction, normalization, bias-field correction, and combinations thereof.

47. The one or more non-transitory computer-readable media of claim 33, wherein the selectively derived image is usable as one or more of: a diagnostic tool for identifying a patient condition or conditions; a therapeutic tool for determining and / or directing drug or other treatment protocols; a monitoring tool that can be employed to observe changes in disease progression over time.

48. The one or more non-transitory computer-readable media of claim 47, wherein the selectively derived image is usable for purposes of diagnosing, treating, and / or monitoring a progression of multiple sclerosis (MS) in a human brain.

Images