WO2017062882A1 - Échantillonnage sélectif pour évaluer des fréquences spatiales structurales avec des mécanismes de contraste spécifiques - Google Patents

Échantillonnage sélectif pour évaluer des fréquences spatiales structurales avec des mécanismes de contraste spécifiques Download PDF

Info

Publication number
WO2017062882A1
WO2017062882A1 PCT/US2016/056147 US2016056147W WO2017062882A1 WO 2017062882 A1 WO2017062882 A1 WO 2017062882A1 US 2016056147 W US2016056147 W US 2016056147W WO 2017062882 A1 WO2017062882 A1 WO 2017062882A1
Authority
WO
WIPO (PCT)
Prior art keywords
value
gradient
values
para
voi
Prior art date
Application number
PCT/US2016/056147
Other languages
English (en)
Inventor
David R. Chase
Timothy W. James
Kristin E. JAMES
Original Assignee
bioProtonics LLC
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from US15/167,828 external-priority patent/US9664759B2/en
Application filed by bioProtonics LLC filed Critical bioProtonics LLC
Priority to EP16854502.8A priority Critical patent/EP3359034A4/fr
Priority to AU2016334250A priority patent/AU2016334250B2/en
Priority to CA3000765A priority patent/CA3000765C/fr
Priority to JP2018517540A priority patent/JP6906507B2/ja
Priority to CN201680071304.4A priority patent/CN108366753B/zh
Publication of WO2017062882A1 publication Critical patent/WO2017062882A1/fr

Links

Classifications

    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/0033Features or image-related aspects of imaging apparatus classified in A61B5/00, e.g. for MRI, optical tomography or impedance tomography apparatus; arrangements of imaging apparatus in a room
    • A61B5/0036Features or image-related aspects of imaging apparatus classified in A61B5/00, e.g. for MRI, optical tomography or impedance tomography apparatus; arrangements of imaging apparatus in a room including treatment, e.g., using an implantable medical device, ablating, ventilating
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/05Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves 
    • A61B5/055Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves  involving electronic [EMR] or nuclear [NMR] magnetic resonance, e.g. magnetic resonance imaging
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/48Other medical applications
    • A61B5/4842Monitoring progression or stage of a disease
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R33/00Arrangements or instruments for measuring magnetic variables
    • G01R33/20Arrangements or instruments for measuring magnetic variables involving magnetic resonance
    • G01R33/44Arrangements or instruments for measuring magnetic variables involving magnetic resonance using nuclear magnetic resonance [NMR]
    • G01R33/48NMR imaging systems
    • G01R33/4818MR characterised by data acquisition along a specific k-space trajectory or by the temporal order of k-space coverage, e.g. centric or segmented coverage of k-space
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R33/00Arrangements or instruments for measuring magnetic variables
    • G01R33/20Arrangements or instruments for measuring magnetic variables involving magnetic resonance
    • G01R33/44Arrangements or instruments for measuring magnetic variables involving magnetic resonance using nuclear magnetic resonance [NMR]
    • G01R33/48NMR imaging systems
    • G01R33/483NMR imaging systems with selection of signals or spectra from particular regions of the volume, e.g. in vivo spectroscopy
    • G01R33/4833NMR imaging systems with selection of signals or spectra from particular regions of the volume, e.g. in vivo spectroscopy using spatially selective excitation of the volume of interest, e.g. selecting non-orthogonal or inclined slices
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R33/00Arrangements or instruments for measuring magnetic variables
    • G01R33/20Arrangements or instruments for measuring magnetic variables involving magnetic resonance
    • G01R33/44Arrangements or instruments for measuring magnetic variables involving magnetic resonance using nuclear magnetic resonance [NMR]
    • G01R33/48NMR imaging systems
    • G01R33/54Signal processing systems, e.g. using pulse sequences ; Generation or control of pulse sequences; Operator console
    • G01R33/56Image enhancement or correction, e.g. subtraction or averaging techniques, e.g. improvement of signal-to-noise ratio and resolution
    • G01R33/561Image enhancement or correction, e.g. subtraction or averaging techniques, e.g. improvement of signal-to-noise ratio and resolution by reduction of the scanning time, i.e. fast acquiring systems, e.g. using echo-planar pulse sequences
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R33/00Arrangements or instruments for measuring magnetic variables
    • G01R33/20Arrangements or instruments for measuring magnetic variables involving magnetic resonance
    • G01R33/44Arrangements or instruments for measuring magnetic variables involving magnetic resonance using nuclear magnetic resonance [NMR]
    • G01R33/48NMR imaging systems
    • G01R33/54Signal processing systems, e.g. using pulse sequences ; Generation or control of pulse sequences; Operator console
    • G01R33/56Image enhancement or correction, e.g. subtraction or averaging techniques, e.g. improvement of signal-to-noise ratio and resolution
    • G01R33/565Correction of image distortions, e.g. due to magnetic field inhomogeneities
    • G01R33/56509Correction of image distortions, e.g. due to magnetic field inhomogeneities due to motion, displacement or flow, e.g. gradient moment nulling

Definitions

  • the herein claimed method relates to the field of diagnostic assessment of fine textures in biological systems for pathology assessment and disease diagnosis, and in material and structural evaluation in industry and in engineering research. More specifically, the invention employs a method for repeat measurement of k-values associated with the spatial organization of biologic tissue texture, with the MRI machine gradients turned off and for k-values in an narrowly associated neighborhood with a low gradient. This allows assessment of tissue texture on a time-scale on the order of a msec, whereby the problem of patient motion becomes negligible. The method enables in vivo assessment, towards diagnosis and monitoring, of disease and therapy-induced textural changes in tissue.
  • Representative targets of the technique are: 1) for assessment of changes to trabecular architecture caused by bone disease, allowing assessment of bone health and fracture risk, 2) evaluation of fibrotic development in soft tissue diseases such as, for example, liver, lung, and heart disease, and 3) changes to fine structures in neurologic diseases, such as the various forms of dementia, or in cases of brain injury and downstream neuro-pathology as in, for example, Traumatic Brain Injury (TBI) and Chronic Traumatic Encephalopathy (CTE), or for characterization and monitoring of abnormal neurologic conditions such as autism and schizophrenia.
  • Other pathology applications include assessment of vascular changes such as in in the vessel network surrounding tumors or associated with development of CVD (Cerebrovascular Disease), and of changes in mammary ducting in response to tumor growth.
  • the invention also has applications in assessment of fine structures for a range of industrial purposes such as measurement of material properties in manufacturing or in geology to characterize various types of rock, as well as other uses for which measurement of fine structures/textures is needed.
  • CT Tomography
  • PET Positron Emission Tomography
  • Bone health is compromised by ageing, by bone cancer, as a side effect of cancer treatments, diabetes, rheumatoid arthritis, and as a result of inadequate nutrition, among other causes. Bone disease affects over ten million people annually in the US alone, adversely affecting their quality of life and reducing life expectancy.
  • the current diagnostic standard is Bone Mineral Density (BMD), as measured by the Dual Energy X-ray Absorptiometry (DEXA) projection technique.
  • This modality yields an areal bone density integrating the attenuation from both cortical and trabecular bone, similar to the imaging mechanism of standard x-ray, but provides only limited information on trabecular architecture within the bone, which is the marker linked most closely to bone strength.
  • KANIS, J. AND GLUER, C "An update on the diagnosis and assessment of osteoporosis with densitometry”; Osteoporosis International, Vol. 11, issue 3, 2000.
  • LEGRAND E. et al.; "Trabecular bone microarchitecture, bone mineral density, and vertebral fractures in male osteoporosis”; JBMR, Vol. 15, issue 1, 2000.
  • BMD correlates only loosely with fracture risk.
  • TBS Trabecular Bone Score
  • Fibrotic diseases occur in response to a wide range of biological insults and injury in internal organs, the development of collagen fibers being the body's healing response. The more advanced a fibrotic disease, the higher the density of fibers in the diseased organ. Fibrotic pathology occurs in a large number of diseases, from lung and liver fibrosis, to cardiac and cystic fibrosis, pancreatic fibrosis, muscular dystrophy, bladder and heart diseases, and myelofibrosis, in which fibrotic structures replace bone marrow. Fibrotic development is attendant in several cancers, such as breast cancer. A different pathology development is seen in prostate cancer, where the disease destroys healthy organized fibrous tissue.
  • tissue biopsy a highly invasive and often painful procedure with a non-negligible morbidity— and mortality— risk (patients need to stay at the hospital for post-biopsy observation for hours to overnight), and one that is prone to sampling errors and large reading variation.
  • Magnetic Resonance-based Elastography which has been under development for some time for use in assessment of liver disease, is not capable of early-stage assessment— the read errors are too large prior to significant fibrotic invasion (advanced disease). Further, this technique requires expensive additional hardware, the presence of a skilled technician, and takes as much as 20 minutes total set up and scanning time, making it a very costly procedure.
  • the ability to image fibrotic texture directly by MR imaging is compromised both by patient motion over the time necessary to acquire data and by lack of contrast between the fibers and the surrounding tissue. Even acquisition during a single breath hold is severely compromised by cardiac pulsatile motion and noncompliance to breath hold, which results in significant motion at many organs, such as liver and lungs.
  • SNR is low enough that motion correction by combining reregistered MR-intensity profiles obtained from successive echoes is extremely problematic.
  • assessment of the amount of cardiac fibrosis in early stage disease using MRI is seriously hampered by cardiac pulsation over the time of the measurement.
  • motion unlike Gaussian noise, a non-linear effect, it can't be averaged out— there must be sufficient signal level to allow reregistration before averaging for electronic noise-reduction.
  • a more sensitive (higher SNR), non-invasive technique capable of assessing textural changes throughout the range of fibrotic development, from onset to advanced pathology, is needed to enable diagnosis and monitoring of therapy response.
  • AD Alzheimer' s Disease
  • ALS Amyotrophic Lateral Sclerosis
  • Parkinson's disease conditions precipitated by Traumatic Brain Injury (TBI) such as Chronic Traumatic Encephalopathy (CTE)
  • TBI Traumatic Brain Injury
  • CTE Chronic Traumatic Encephalopathy
  • MS Multiple Sclerosis
  • CVD Cerebrovascular Disease
  • other neurologic diseases are often only diagnosable in advanced stages by behavioral and memory changes, precluding the ability for early stage intervention.
  • conditions such as epilepsy and autism have been associated with abnormal variations in fine neuronal structures, which, if clinically diagnosable, would allow targeted selection for testing therapy response.
  • PARA 14 Use of CSF biomarkers for dementia diagnosis is painful and highly invasive and cannot differentiate signal levels by anatomic position in the brain, as is possible with imaging biomarkers. As various forms of dementia are found to have different
  • CVD Another disease associated with various forms of dementia is CVD
  • PARA 16 Another possible neurologic application for the claimed method is to, in vivo, determine the boundaries of the various control regions of the cerebral cortex or the different Brodmann's areas of which these are comprised. Such ability would greatly aid data interpretation in brain function studies, such as those performed using, for example, FMRI (Functional Magnetic Resonance Imaging).
  • FMRI Magnetic Magnetic Resonance Imaging
  • PARA 17 The three classes of diseases listed above, bone disease, fibrotic diseases, and neurologic diseases are not an all-inclusive list.
  • Other disease states in which pathology- induced changes of fine structures occur for instance angiogenic growth of vasculature surrounding a tumor, or fibrotic development and changes in vasculature and mammary gland ducting in response to breast tumor development, also are pathologies wherein the ability to resolve fine tissue textures would enable early detection of disease, and monitoring of response to therapy.
  • a method for selective sampling to assess tissue texture using magnetic resonance (MR) is accomplished by applying a contrast mechanism enhancing the contrast between the component tissue types in a multiphase biologic sample being measured.
  • a volume of interest (VOI) is then selectively excited employing a plurality of time varying radio frequency signals and applied gradients.
  • An encoding gradient pulse is applied to induce phase wrap to create a spatial encode for a specific k-value and orientation, the specific k-value determined based on texture within the VOL
  • a time varying series of acquisition gradients is initiated to produce a time varying trajectory through 3D k-space of k-value encodes, a resulting k-value set being a subset of that required to produce an image of the VOL
  • Multiple sequential samples of the NMR RF signal encoded with the k-value set are simultaneously recorded.
  • the recorded NMR signal samples are then post-processed to produce a data set of signal vs k-values for k-values in the k-value set determined by the trajectory, to characterize the tissue in the VOL
  • FIG. 1 is a simulation showing the number of data samples required for averaging to achieve an output SNR > 20dB as a function of input SNR;
  • FIG. 2 is a simulation showing the number of data samples needed for averaging to achieve a SNR > 20db as a function of location in k-space;
  • FIG. 3 is an example timing diagram of a pulse sequence for the claimed method showing the timing of a single TR
  • FIG. 4 is a close-up of the example timing diagram of FIG. 3;
  • FIG. 5 is an example of a timing diagram for the claimed method, designed to acquire multiple measures of a select set of k-values, with a different number of samples acquired at each k-value to counteract the decrease in energy density at increasing k-value;
  • FIG. 6 is a simulation showing that the ability provided by the claimed method to acquire many repeats of a targeted k-value within a single TR enables robust signal averaging to boost SNR;
  • FIG. 7 is a simulation showing the results of attempting to acquire 90 samples for averaging using the conventional frequency-encoded echo approach, wherein acquisition of only a small number of repeats of a particular k-value are possible in each TR due to the long record time for each echo;
  • FIG. 8 is an example timing diagram for the claimed method designed to provide data acquisition over multiple refocused echoes within a single TR; and, [PARA 29] FIGs. 9 and 10 are a depiction of two possible shapes for the acquisition volume of interest (VOI);
  • FIG. 11 is an example timing diagram of a pulse sequence for the claimed hybrid method showing the timing of a single TR
  • FIG. 12 is a detailed view of the hybrid elements of the method at an expanded scale
  • FIG. 1 is a further detailed view of the very-low SNR acquisition mode portion of FIG. 12;
  • FIG. 14 is a further detailed view of the low SNR acquisition portion of FIG. 12;
  • FIG. 15 is a further detailed view of the high SNR acquisition portion of FIG. 12;
  • FIG. 16 is an example timing diagram of a pulse sequence for the claimed hybrid method showing data acquisition in a single echo
  • FIG. 17 is a further detailed view of the very-low SNR, low SNR and high SNR acquisition portions of FIG. 16;
  • FIG. 18 is a an example timing diagram of a pulse sequence for a low SNR acquisition.
  • FIG. 19 is a further detailed view of the low SNR acquisition mode of FIG. 18;
  • FIG. 20 is an example timing diagram of a pulse sequence for high SNR acquisition
  • FIG. 21 is a further detailed view of the high SNR acquisition mode of FIG. 20;
  • FIGs. 22A and 22B are pictorial representations of healthy an osteoporitic bone structure
  • FIG. 23 is a pictorial representation of fibrotic tissue in a liver
  • FIG. 24 is an example timing diagram of a pulse sequence implementing a first diffusion contrast
  • FIG. 25 timing diagram of a pulse sequence implementing a seconed diffusion contrast
  • FIG. 26 is a pictorial representation of VOIs dispers3ed in fibrotic tissue
  • FIG. 27 is a pictorial representation of cortical minicolumns in the brain
  • FIG. 28 is a pictorial representation of VOI placement in the brain
  • FIGs. 29A -29C are representations of three histology images showing progressive pathology with AD advancement.
  • FIG. 30 is exemplary representation of placement of a VOI in the brain cortex and application of gradients for k- value .
  • 180° inversion pulse RF pulse that inverts the spins in a tissue region to allow refocusing of the MR signal.
  • Adiabatic pulse excitation Adiabatic pulses are a class of amplitude- and frequency- modulated RF-pulses that are relatively insensitive to 6 inhomogeneity and frequency offset effects.
  • AWGN Additive White Gaussian Noise Additive white Gaussian noise is a basic noise model used in Information theory to mimic the effect of many random processes that occur in nature.
  • Biopsy A biopsy is a sample of tissue extracted from the body in order to examine it more closely.
  • Crusher gradients Gradients applied on either side of a 180° RF refocussing slice selection pulse to reduce spurious signals generated by imperfections in the pulse.
  • DEXA Dual Energy X-ray Absorptiometry is a means of measuring bone mineral density using two different energy x-ray beams.
  • Gradient set the set of coils around the bore of an MR scanner used primarily to spatially encode signal or to set a particular phase wrap in a selected direction
  • Isochromat A microscopic group of spins that resonate at the same frequency
  • k-value coefficient The coefficient in a Fourier series or transform reflecting the relative weight of each specific k-value in the series.
  • k-space The 2D or 3D Fourier transform of the MR image.
  • k-value One of the points in k-space reflecting the spacing of structural elements in a texture field.
  • MRE Magnetic Resonance Elastography an imaging technique that measures the stiffness of soft tissues using acoustic shear waves and imaging their propagation using MRI.
  • Noise floor is the measure of the signal created from the sum of all the noise sources and unwanted signals within a measurement system
  • PET Positron Emission Tomography is a functional imaging technique that produces a three-dimensional image of functional processes in the body using a positron-emitting radiotracer.
  • Phase coherence When referring to multiple measurements within a common VOI of a or multiple k-values indicates that the sample has the same position relative to the measurement frame of reference
  • Phase encode A phase encode is used to impart a specific phase angle to a transverse magnetization vector.
  • the specific phase angle depends on the location of the transverse magnetization vector within the phase encoding gradient, the magnitude of the gradient, and the duration of the gradient application.
  • Phase wrap The helical precession of the phase of the transverse magnetization along a phase encoded sample
  • Semi-crystalline texture a texture exhibiting regular spacing along one or more directions
  • slice Used interchangeably to indicate a non-zero thickness planar section of the Slice-selective refocusing Refocussing of spins through combination of a slice selective gradient and an RF pulse such that the bandwidth of the RF pulse selects a thickness along the direction of the gradient, and the RF pulse tips the net magnetization vector away from its equilibrium position Only those spins processing at the same frequency as the RF pulse will be affected.
  • T2 Defined as a time constant for the decay of transverse magnetization arising from natural interactions at the atomic or molecular levels.
  • T2* In any real NMR experiment, the transverse magnetization decays much faster than would be predicted by natural atomic and molecular mechanisms; this rate is denoted T2* ("T2-star”). T2* can be considered an "observed” or “effective” T2, whereas the first T2 can be considered the “natural” or “true” T2 of the tissue being imaged. T2* is always less than or equal to T2.
  • TBS Trabecular Bone Score is a technique that looks for texture patterns in the DEXA signal for correlation with bone microarchitecture for assessing bone health
  • TbTh trabecular thickness for bone measurement TbTh trabecular thickness for bone measurement.
  • Echo Time is the time between the 90° RF pulse and MR signal sampling, corresponding to maximum of echo.
  • the 180° RF pulse is applied at time TE/2.
  • Repetition Time is the time between 2 excitations pulses (time between two 90° RF pulses).
  • Textural frequency the number of texture wavelength repeats per unit length in a texture Texture wavelength the characteristic spacing between structural elements in a texture TR Spin Echo sequences have two parameters: Echo Time (TE) is the time between the 90° RF pulse and MR signal sampling, corresponding to maximum of echo. The 180° RF pulse is applied at time TE/2. Repetition Time is the time between 2 excitations pulses (time between two 90° RF pulses).
  • TE Echo Time
  • Repetition Time is the time between 2 excitations pulses (time between two 90° RF pulses).
  • Vector combination gradient A magnetic gradient resulting from any vector combination of the gradient coil set
  • windowing function In signal processing, a window function (also known as an apodization function or tapering function) is a mathematical function that is zero- valued outside of some chosen interval
  • x-ray diffraction is a tool used for identifying the atomic and molecular structure of a crystal
  • the embodiments disclosed herein provide an MR-based technique that enables in vivo, non-invasive, high-resolution measurement and assessment of fine biologic textures, enabling monitoring of texture formation and/or change in response to disease onset and progression in a range of pathologies.
  • This same method can be applied to fine- texture characterization in other biologic and physical systems. It enables MR-based resolution of fine textures to a size scale previously unattainable in in vivo imaging.
  • the method while described herein with respect to biological systems for examination of tissue, is equally applicable for assessment of fine structures in a range of industrial purposes such as measurement of material properties in manufacturing or in geology to characterize various types of rock, as well as other uses for which measurement of fine
  • the acquisition TR can be repeated as many times as necessary, changing the encode as needed to span the desired extent of real and of k-space required.
  • the set of one or more k- values output from each TR are now high SNR due to the ability to average repeats without motion effects, and since the measure of interest is textural spacing, and not development of an image, the lack of phase coherence between TRs is of no concern.
  • the method claimed herein consists of acquiring MR signal from within an inner volume to encompass a specific tissue region of interest, such as a lesion, an organ, a location in an organ, a specific region of bone, or a number of regions in a diseased organ for sampling.
  • This inner volume may be excited by one of a number of methods, including but not limited to: intersecting slice- selective refocusing, selective excitation using phased-array transmit in combination with appropriate gradients, adiabatic pulse excitation to scramble signal from the tissue outside the region of interest, outer volume suppression sequences, and other methods of selectively exciting spins in an internal volume including physically isolating the tissue of interest, to name a few,
  • VOI volume of interest
  • the gradient is turned off, and multiple samples of signal centered at a specific k-value, the spread of which is defined by receiver BW and sampling length, are acquired. This measurement is repeated only in specified directions within the VOI rather than trying to map all of k-space, as is required to generate an image.
  • One or more samples of a particular k-value are acquired within an acquisition block during a single TR and the k-value subsequently incremented or decremented, allowing further multiple samples of other k- values as desired during the same TR.
  • This method allows multiple sampling of each k-value of interest over a time period of milliseconds, providing immunity to subject motion.
  • the process can then be repeated in further TRs, the requirement on motion between k-value acquisitions being only that the VOI remain within the tissue region of interest.
  • Build up of a magnitude spectrum of spatial frequencies may be accomplished without the need to acquire it in a spatially coherent manner. Because the quantities of interest are the relative intensities of the various k- values (textural spacings) present in the sample volume, as long as the acquisition volume remains within a representative sample of tissue, any motion between the blocks does not compromise the measurement.
  • the data acquired can, with reference to positioning images, be mapped spatially.
  • Either the VOI can be moved in successive TRs or interleaved acquisition done within a single TR by exciting additional volumes during the time that the signal is recovering in advance of the next TR. The requirement is that successive VOIs be excited in new tissue, that does not overlap the previous slice selects. Spatial variation of pathology can be determined by this method. This can also be used to monitor temporal progression of a pathology through an organ if the measure is repeated longitudinally.
  • PARA 59 The method claimed herein can be used in conjunction with time-dependent contrast schemes that target blood flow.
  • Some of these contrast techniques are Blood Oxygenation Level Dependent (BOLD) imaging, Arterial Spin Labeling (ASL) imaging, and Dynamic Susceptibility Contrast (DSC) imaging.
  • BOLD Blood Oxygenation Level Dependent
  • ASL Arterial Spin Labeling
  • DSC Dynamic Susceptibility Contrast
  • PARA 60 The method claimed herein can also be used in conjunction with other MR- based measurement techniques, including DWI and DTI, to provide front end information toward parameter selection for the diffusion techniques as well as correlation with their measurements of tissue health.
  • the echo record window can be designed such that recording begins with the highest k-values of interest, as signal level is highest at the echo peak. This enables recording of fine structures currently unachievable with in vivo MR imaging.
  • [PARA 66] Coil combination is also simplified by having higher SNR for each k- value, hence providing a significant improvement in overall SNR. This is especially beneficial as the trend in MRI is towards coil arrays composed of many small element coils. As the acquisition volumes targeted in the method claimed herein are small, correction for phase across the sample volume is not needed. Only one phase and gain value for each coil is needed for combining the multiple element channels. These can be combined using the Maximal Ratio Combining (MRC) method, which weights the coil with the highest SNR most heavily, or other multi-signal combination methods.
  • MRC Maximal Ratio Combining
  • Phase and gain for the elements of a given coil array can be determined once from a phantom and applied to patient data.
  • PARA 67 Signal acquisition and data sampling in a standard MRI scan is done by acquiring complex- valued samples of multiple echoes, while applying a gradient sequence concurrently, as well as in sequence with the echoes. Imaging relies on frequency encode for one of the dimensions because this allows a line in k-space to be acquired with each phase encode rather than a single point. For 3-dimensional imaging, two dimensions in k- space normally rely on phase encode to generate the targeted filling of k-space, with the third dimension frequency-encoded.
  • Phase encode acquisition in imaging usually entails acquisition of on the order of 256 k-values in each of the phase-encode directions, hence is is a relatively slow process.
  • Clinical MRI scans take on the order of 10-15 minutes to generate an image.
  • the aim in image construction is to acquire sufficient k-space coverage to fill out all the coefficients in the 2 or 3 -dimensional Fourier series, which is why in standard MR resolution is limited by subject motion.
  • the method claimed herein is in direct contrast to standard MR data acquisition, with its focus on image generation. Image formation is plagued by blurring resulting from subject motion over the long time necessary to acquire the large data matrix required. Since the target of the method claimed herein is texture rather than image, the only requirement on subject motion is that the sampled volume remain within a region of similar tissue properties over the course of acquiring data. This is a much less stringent and easy to achieve target than the requirement of structural phase coherence, as the scale of the allowable motion is then large enough, and of a temporal order, to be easily correctable by real-time motion assessment and correction techniques. The speed of acquisition for the method claimed herein is such that, in most cases, real-time motion correction may not be necessary at all. While other methods have focused on post-processing of images to try to extract textural measures, the method claimed herein eliminates the need for image generation, focusing instead on directly measuring texture, hence enabling a more sensitive and robust measure.
  • k-space sampling is considered synonymous with sampling of an echo in the presence of a gradient set.
  • the approach to k-space filling is to acquire only the set of k-values needed for texture evaluation in the targeted pathology, with data acquired after the gradient is switched off. This method enables such rapid acquisition of single -k-value repeats for averaging for noise reduction that subject motion does not degrade the data.
  • MR echo sampling provides specific samples vs. time of a time-dependent echo.
  • the echo is comprised by the gradients applied concurrently (for the frequency- encode axis) and prior to (for a phase encode axis), but also contains the isochromats associated with the different chemical species of the sample, as well as the envelope (T2 & T2*) associated with spin-spin interactions.
  • a common technique for noise reduction in signal acquisition is through repeat sampling of a signal and subsequent combination of the multiple measurements.
  • linear noise sources such as Gaussian noise
  • this technique improves SNR through cancellation of the random noise on the signal, the cancellation effect increasing with the number of samples, N.
  • ⁇ ' is the complex- valued signal
  • ⁇ ' represents the average of the two acquisitions.
  • [PARA 90] Within a given acquisition in standard MR practice, there are M samples which are acquired of the echo. Instead of acquiring a sample at each k-value, N ⁇ M of those samples could be used for estimation of the (complex-valued) underlying signal value at a specific k-value. Multiple samples within an acquisition can be combined with much less concern of movement than across acquisitions because they are much closer in time. [PARA 91] If the entire echo is used to measure one k-value, the receive bandwidth can be adjusted so as to pass the most abundant resonant peaks in the underlying NMR spectrum, and attenuate frequencies above them.
  • the claimed method enables acquisition of values in regions of k-space which have very low signal levels, such as would be found for higher k- values (shorter textural-wavelengths)— the fine texture range that has hitherto remained elusive.
  • the full NMR spectrum may be extracted (without any phase encoding gradients: just volume selection) to obtain a baseline of the underlying signal strength (and associated frequencies), which in turn will be spatially modulated, providing insight into textural wavelengths through knowledge of the chemical species expected in the textural elements under study.
  • the isochromats of interest can be extracted by acquiring N samples of the echo, then taking the Fourier transform. Since the echo is being played out with no gradient, the strength of the resulting signal at the Isochromat of interest will correspond to the (complex-valued) k-value coefficient of interest.
  • represents the noise variance
  • ' ' represents the squared magnitude of the isochromat(s) of interest
  • ° ⁇ £ ⁇ 1 represents the allowable error of the estimate.
  • NF e s is the effective noise figure of the receiver
  • 3 ⁇ 4 is Boltzmann' s constant
  • T is the temperature in Kelvin of the biological sample
  • B is the receiver bandwidth.
  • N can be used as a guide to the number of samples that need to be acquired within a given acquisition in order to create a reasonable estimate.
  • FIG. 3 shows an example timing diagram for a pulse sequence for data acquisition using the method claimed herein.
  • RF pulses included in trace 302 are employed to excite selected volumes of the tissue under investigation, as in typical MR imaging.
  • a first RF pulse, 304 is transmitted coincidentally with a gradient pulse 308 on the a first magnetic field gradient, represented in trace 306 .
  • This excites a single slice, or slab, of tissue the positioning of which is dependent on the orientation and magnitude of the first gradient, and the frequencies contained in the RF pulse.
  • the negative gradient pulse, pulse 310 rephases the excitation within the defined thickness of the slice or slab.
  • this second RF pulse 312 tips the net magnetic vector to antiparallel to Bo, it results in inversion of spins and subsequent refocusing, thus leading to a signal echo at a time after the 180 degree RF pulse equivalent to the time between the 90° and 180° RF pulses.
  • An initial higher value gradient pulse, 318, at the start of gradient pulse 316 is a crusher, or "spoiler" gradient, designed to induce a large phase wrap across the tissue volume.
  • the second RF pulse in combination with the applied second gradient, provides slice selective refocusing of the signal in a region defined by the intersection of the first slice and the second slice set by this second gradient.
  • [PARA 103] An encoding gradient pulse 326, on trace 314, sets an initial phase wrap, hence k-value encode, along the direction of gradient pulse 326.
  • the k-value encode can be oriented in any direction, by vector combination of the machine gradients but for ease of visualization is represented as on the second gradient.
  • the negative prephasing gradient pulse 326 winds up phase such that, at the signal echo following the second 180° RF pulse, signal acquisition starts at high k- value, which may then be subsequently decremented (or incremented or varied in orientation) for further acquisitions, as will be described below.
  • energy density in the signal is generally proportional to k "1 , this method ensures k-values with lower SNR are acquired first, before T2 effects have caused much overall signal reduction.
  • a receive gate 333 is opened to receive the RF signal, which is shown in FIG. 3 as pulse 334 on trace 336.
  • the RF signal in trace 336 is a representation showing only the signal present in the receive gate window without showing the actual details of the RF signal outside the window.
  • Sampling occurs as represented by trace 338 beginning with the initial k-value, 340a, seen on trace 324. Note that, at the scale of the drawing, the sampling rate is high enough that the individual triggers of the analog to digital converter (A/D) have merged together in trace 338. (The expanded time scale in FIG. 4 described below shows the individual A/D triggers.)
  • a k-value selection gradient pulse 342a is then applied and the resultant k-value 340b is sampled. (Though shown in the figure as a negative pulse on the second gradient, decrementing the k-value, in practice this pulse and subsequent k-value gradient pulses can be designed through any vector combination of gradients to select any k-value or orientation.)
  • the k-value selection gradient pulse 342b selects a third k-value 340c which is sampled by the A/D. Each gradient pulse changes the phase wrap, selecting a new k-value.
  • Application of a k-value selection gradient pulse (342c - 342f) followed by multiple sampling of the resultant k- value coefficient is repeated as many times as desired.
  • the gradient orientations for slice and k- value select may be coincident with the machine gradients, which are aligned to lie coincident or orthogonal to the Bo field.
  • the acquisition directions and k- value encodes may be selected using gradients that are a vector combination of all three machine-gradient axes.
  • the prewinding encoding gradient pulse can be set such that the fist k-value to be measured is the desired low SNR k-value.
  • the prewinding gradient pulse can be set to zero so that the first k-value measured is kO.
  • a measurement of kO may be desired for the purpose of determining the systems receiver sensitivity to the particular VOI, determining the relative prevalence of isochromats (e.g., water vs. lipids) irrespective of texture in the VOI, or for the purpose of establishing a reference value for normalization of the other k- values measured in a VOI or for comparison with k- values from other VOI.
  • a strategy for gathering a specified set of k- values for a VOI may include measuring the low SNR k- values (typically the higher k-values) in a first set of multiple TR and then measuring kO and other higher SNR k-values in other TRs while remaining in the same VOI.
  • the signal reaches a maximum at the time of the spin echo. It is also shown diagrammatically that the signal is varying throughout the acquisition of the multiple RF measurements of a k-value and more so between successive blocks of measurements of k-values. Alignment in time of the measurement of the low SNR k-values with the highest echo signal enhances the SNR of the k-value measurement, alternatively alignment of higher SNR k-values with lower echo signal allows gathering additional useful k-value acquisitions during the echo.
  • FIG. 4 shows a close-up of the pulse sequence of FIG. 3 during the initial portion of the RF sampling window 338 between 7.25 and 8.00 msec. Multiple samples of the same k-value, taken in rapid succession with all gradients off, provide the input for signal averaging to reduce AWGN when SNR is low.
  • a first block 344a of the sampling window 338 multiple samples 346a are taken of the first k-value 340a.
  • transition samples 348a are taken.
  • multiple samples 346b are taken at the second It- value 340b.
  • k-value selection gradient pulse 342b then occurs with associated transition samples 348b, and subsequent acquisition of samples 346c of the third k-value 340c after the gradients are switched off.
  • the underlying signal is minimally impacted by motion due to the very short time window used to acquire data at each given k- value. Since the data is acquired with gradients off, there is no issue with chemical shift and the effective T2* is longer, boosting the signal value.
  • PARA 112 A consistent number of samples at each k-value can be acquired, or an alternative sequence may be employed where, as k- values decrease, hence increasing in signal amplitude, fewer samples are acquired.
  • a pulse sequence designed for this type of acquisition is illustrated in FIG. 5. Multiple samples of each k-value targeted in the acquisition are acquired in rapid succession, with all gradients off. These repeats provide the input for signal averaging in low SNR signals.
  • the underlying signal is minimally impacted by motion due to the very short time window in which data is acquired for a given k-value.
  • Samples within the portions of the sample window 344a - 344g outlined on FIG. 5 correspond to the number of samples acquired for a given k-value 340a - 340g each induced by an unwinding pulse 342a - 342f of the k-value selection gradient.
  • N k the number of samples associated with a given k-value, can be selected based upon expected SNR, tissue contrast, contrast to noise, pathology, texture size, and/or texture bandwidth. For the example in FIG. 5 it can be seen that a decreasing number of samples are taken for progressively smaller k-values (larger textural features).
  • [PARA 114] Refocusing the echo, and/or a new TR can be used to build up a library of It- space samples. Acquisition of multiple k- values within one TR can be facilitated by application of multiple refocusing gradients and/or RF pulses, to increase the time over which the additional k- values can be sampled within a TR. These later echoes would presumably be used to acquire the coefficients of the lower k- values in the selected set, as their energy density in the continuum of values is generally higher so the effect of T2 decay on overall signal will not affect them as severely as it would the higher k- values. In this way a larger portion of the required k-space filling can be accomplished over fewer TRs, allowing more rapid data acquisition, minimizing the need for repositioning the VOL
  • FIG. 8 shows an extension of the basic sequence of the method claimed herein, using spin-echo refocusing to extend the record time for the TR.
  • Application of a refocusing RF pulse 802 with an associated gradient pulse 804 results in slice- selective refocusing.
  • a second sampling window 806 is opened by the receive gate 808.
  • Multiple k-value selection gradient pulses 810 are applied to increment the selected k-value and, after switching off each successive gradient pulse, multiple samples of the selected k-value are acquired in the sampling window.
  • a second slice-selective refocusing RF pulse 812 with associated gradient pulse 814 again inverts the spins and, after application of each in the multiplicity of k-value selection gradient pulses 820, data is acquired in a third sampling window 816, opened by the receive gate 818.
  • a third sampling window 816 opened by the receive gate 818.
  • an increasing number of k-values may be sampled with each refocusing. Refocusing can be repeated until the decrease in signal level from T2 and other effects makes further signal acquisition ineffective.
  • Another method to extend the record time by exciting multiple signal echoes is to use one, or a series of, gradient recalled echoes (GRE). GRE are different from the SE in that they cannot refocus the effects of stationary inhomogeneities, so T2* effects limit the number of repeats.
  • GRE gradient recalled echoes
  • the k-values associated with particular pathology will be part of the determination of the number of samples needed for signal averaging, N k -
  • the wavelength of pertinent textures is in the range of 400 microns, i.e. a k-value of 2.5 cycles/mm. This is similar to the textural spacings seen in fibrotic development in many other diseases, such as cardiac fibrosis.
  • the spacing of elements in trabecular bone varies a lot, but the minimum spacing of interest is the width of trabecular elements, which are approximately 80 microns, setting a maximum k- value of 12.5cycles/mm.
  • many of the textures of interest are very fine, on the scale of 50 microns, equivalent to a k-value of 20 cycles/mm.
  • the signal level obtainable will depend on anatomy to some extent. For instance, though the resolution needed is highest in brain, the proximity of the cortex to the surface of the head ensures that use of a surface coil will provide significant signal boost for cortical structures. Lower resolution is required in liver, as the structures of interest are on the order of several hundred, rather than tens of microns. But, the organ is deeper (further from the coil) reducing the measured signal. Using the in-table coil for spine data acquisition yields modest signal level and good stabilization. Also, bone is a high contrast target, so the SNR requirement is not as stringent. For all these reasons, the exact number of repeats needed for averaging depends on more than the k-value range targeted.
  • FIG. 6 shows a simulation demonstrating that the ability provided by the claimed method to acquire many repeats of a targeted k-value within a single TR enables robust signal averaging to boost SNR.
  • a subject displacement rate which has in practice been measured clinically over the course of several scans
  • the situation is actually worse due to the fact that to acquire 90 repeats using conventional spatially-encoded echoes would require several TRs, making the acquisition time significantly longer, and the signal degradation due to motion much more severe.
  • the potential SNR gain due to multiple sample combination has been nullified by the effects of motion.
  • VOI shape, dimensions, orientation, and positioning within an organ/anatomy affects the data measured and its interpretation.
  • the VOI shape can be chosen to maximize usefulness of the acquired data.
  • Data can be acquired in different directions, and at different textural wavelengths (k- values) within a VOI enabling assessment of textural anisotropy. Texture can be sampled in multiple VOIs, either interleaved within a single TR, or in successive TRs, towards assessment of pathology variation across an organ.
  • Standard interleaving processes for the VOI may be used within a TR to provide additional data by applying additional encoding pulses on vector combination gradients and associated k-value selection gradient pulses for k- values in the interleaved VOI.
  • additional excitation RF pulses with associated slice selection gradients may be repeated within the same TR by exciting a volume of interest with a gradient set in each repeat having at least a first gradient with an alternative orientation from the first gradient pulse 308 applied initially in the TR, to define an additional VOI for excitation in new tissue, that does not overlap any previous VOI in the TR (fourth, fifth and sixth gradients in a first repeat and succeeding incremental gradients in subsequent repeats).
  • This response can be mapped, or the several measures taken and averaged, whatever is appropriate for the targeted pathology. This is similar to the multi- positioning of tissue biopsy. However, in the case of tissue biopsy, the number of repeats is limited due to the highly invasive nature of the technique. The minimum number of structural oscillations to be sampled at a specific k-value dictates a minimum VOI dimension in the direction of sampling— the length required varying inversely with targeted k-value.
  • the VOI dimension in the sampled direction can be held constant for all k-values in the targeted range, with the result that the number of structural oscillations sampled will vary with k-value. This is a simple solution, requiring the sampling dimension be set by the lowest k-value (longest wavelength structure). Using this approach, the sampling dimension of the VOI is larger than required for the highest k-value in the range, thus providing less localization within the tissue than would be otherwise possible.
  • the VOI may be held constant and the vector combination gradient for the encoding and k-selection pulses may be altered from TR to TR for assessing feature size.
  • k-value encodes can be applied in multiple directions by changing the applied vector combination gradients for encoding and k-selection pulsing.
  • the exact form of the VOI and sampling direction can be used to yield much textural information.
  • the organization of cortical neuron fiber bundles is semi-crystalline, as the bundles in healthy tissue form in columns. Because of this, the measure of textural spacing perpendicular to the bundles is very sensitive to orientation. When the orientation is exactly normal to the columns, a very sharp signal maximum is expected, the signal falling off rapidly as the orientation varies on in either rotational direction away from this maximum.
  • One way to measure the spacing and organizational integrity would be to "rock" the acquisition axis around this maximum looking for a resonance in signal intensity. This approach of looking for "textural resonances" by looking for signal maxima can be applied in any tissue region. As pathology degrades the organizational integrity, the sharpness of this peak will degrade and the signal maximum will be reduced.
  • the randomness of the spacing in certain textures can be assessed by varying the length of tissue sampled in a specific, or in multiple directions, with subsequent change in acquisition length. The selected value for that length can be varied over multiple TRs to test the sensitivity of the measured coefficient to this parameter.
  • the VOI can be selectively excited by a number of methods, for instance intersecting slice- selective refocusing, selective excitation using phased-array transmit in combination with appropriate gradients, adiabatic pulse excitation to scramble signal from the tissue outside the region of interest, as examples. Parameter selection for the various methods can be done with SNR optimization in mind.
  • the VOI generated by a slice selective excitation and two additional mutually- orthogonal slice selective refocusing pulses as by VOI 904 in FIG. 9.
  • the shape of the VOI can be designed so that the edges are smooth and more approximate a windowing function, as shown in FIG. 10.
  • These windowing functions provide the volume selection without adverse impact on the spatial frequencies.
  • each spectral line is smeared by the convolution of a Fourier transform of the window function. It is desirable to minimize this smearing of the underlying spectrum, as it decreases the energy spectral density, and adversely impacts SNR.
  • the VOI can be moved from place to place within an organ or anatomy under study to measure the variation of texture/pathology. This response can be mapped, or the several measures taken and averaged, as appropriate for the targeted pathology. This is similar to the multi-positioning of tissue biopsy. However, in the case of tissue biopsy, the number of repeats is limited due to the highly invasive nature of the technique.
  • k-space is probed to reveal texture in such a way as to eliminate the loss of signal resolution that results from subject motion blurring.
  • the focus here is on acquisition of a select few k-values per TR, with sufficient repeats of each to yield high SNR.
  • Each of the individual acquisitions is centered on a single k- value.
  • the spatial encode is, to first order, a single spatial frequency sinusoidal encode, there are a number of factors which have the effect of broadening the spatial frequency selectivity of the k- value measurement.
  • One significant factor affecting the broadness, or bandwidth, of the k- value measurement is the length of the sampled tissue region.
  • an aspect of the claimed method is the ability to set the bandwidth of the k- value measurements by appropriate selection of the sampling length determined by the VOI dimensions or determined by the acquisition dimensions. Using this method, the bandwidth of the measurement can be set according to the desired k-space resolution appropriate to the tissue being evaluated.
  • the k-value can be varied by keeping the magnitude of k-constant but sweeping the vector over the surface of the sphere, or the same angular orientation may be maintained, and the magnitude of k varied, or both can be varied simultaneously
  • a number of approaches to "dither" k-value to reduce speckle or to tailor width in k-space may be employed.
  • a first approach employs constant k-magnitude plus sweeping through a range of angles by keeping gradients on during acquisition and combining the measures using correlative information to eliminate speckle.
  • the same direction in k-space may be maintained but the magnitude varied by leaving gradient on during acquisition and combining the measures using expected correlation.
  • both magnitude and direction may be varied simultaneously or over an acquisition series, essentially performing the other two alternatives simultaneously to both reduce noise and provide a better assessment of the representative k magnitude in a structure in a "small" region around a specific k-value, i.e., to reduce speckle.
  • a dynamic k-space acquisition is therefore employed.
  • the acquisition mode is dynamically chosen based upon the Signal to Noise Ratio (SNR) of the signal at various k-space locations.
  • SNR Signal to Noise Ratio
  • the gradient, applied during signal acquisition, post-acquisition receive bandwidth, and estimation algorithms used are dynamically adjusted based upon the expected SNR values in k-space to optimize acquisition time and post-processed SNR.
  • a single sample at a given k value may be a sufficient estimator. This requires a relatively wide receive bandwidth to accommodate the relatively rapid signal variations in the receive chain as k is changed rapidly (due to the large gradient).
  • Correlation may be introduced in k-space due to selected windowing in profile space.
  • windowing To enable combination of sequential samples from the ADC so as to improve SNR, correlation among successive samples will be increased by proper choice of windowing in profile space, a shorter window driving a greater correlation distance across sequential values in k-space and a longer window resulting in lower sample to sample correlation.
  • Inducing correlation of neighboring points in k-space by windowing in profile- space is a mathematical tool that can, in many cases, help to measure the underlying texture in a low SNR environment. Basically, windowing blurs the data so that the k- value power spectrum is smeared out through k-space, so that sequential measures can be
  • sample-to-sample spacing determined by the analog to digital converter speed and gradient height
  • correlation can be used in post-acquisition processing to form better estimates.
  • receive bandwidth in these regions can be decreased, which further decreases the noise floor.
  • FIG. 11 An example Hybrid pulse sequence is shown in FIG. 11. This particular sequence is an example of a rapid acquisition with refocused echoes (RARE) type sequence wherein three different levels of gradient are used for acquisition as illustrated across three separate echoes.
  • the shown pulse sequence for selecting the desired VOI and initial phase wrap for k is as described in FIG. 3 and is numbered consistently in FIG. 11. While this exemplary pulse sequence for establishing the VOI is employed in the various examples disclosed herein, the determination of VOI made be made by any of numerous approaches including, as an example, time varying RF pulses with commensurately time varying gradients applied.
  • the encoding gradient pulse may be applied for selection of the initial k-value.
  • the data recording starts with acquiring values in regions where lkl»0, given that the signals are smallest there and should be acquired first.
  • k-value windup (as determined by the gradient height and pulse duration) can be acquired in one echo rather than in multiple echoes as will be described subsequently. Multiple combinations of the differing k-value windup also can be acquired within one echo. Additionally, while refocusing is disclosed in the drawings as employing an RF pulse, gradient refocusing may also be employed.
  • a small gradient may be defined as a gradient inducing samples in k-space which will be sufficiently closely spaced so that the samples are highly correlated. These samples can then be post-processed by an estimator which takes advantage of the high inter-sample correlation to improve the resulting SNR. Quantitatively an exemplary "small" gradient might be up to 20% of the magnitude of the encoding gradient pulse. As seen in the figure samples of multiple values in a relatively small neighborhood, Ak, in k-space are obtained.
  • the spacing of Ak can be chosen such that, due to the windowing of the VOI, there is high correlation between neighboring samples.
  • the correlation is exploited in the estimation algorithm to generate an optimal estimate of the signal levels across a neighborhood in k- space. This is appropriate for regions of k-space which have low SNR, but whose values, because of the correlation induced by the windowing function, vary slowly across k-space in a small neighborhood.
  • a third echo from RF pulse 1108 with gradient pulse 1110 is acquired with a relatively larger time dependent phase encode gradient 1112.
  • the higher gradient employed herein creates sequential measurements in k-space which have a lower degree of correlation across the neighborhood of k-space under study. In this case, there is lower inter-sample correlation available for SNR improvement.
  • a higher gradient may be employed for sampling k-space locations whose values have a high SNR to begin with (such as would be seen in lower textural frequency (low k- value) regions). As seen in the figure samples relatively widely spaced across k-space, well outside the inter-sample correlation imposed by the window function, are generated over the entire pulse.
  • Non-zero gradient acquisition allows sweeping across a curvilinear path in k- space.
  • At gradient magnitude, G, w(X), and the total number of samples acquired, N
  • the neighborhood of "high" correlation can be adjusted to be M ⁇ N. This in turn would allow estimation of a multiplicity of distinct values within k-space by using a subset of M samples for each output estimation.
  • Textural data from within a tissue region defined by the VOI can be acquired with the non-zero gradient to enable determination of the local distribution of power density of k- values within a neighborhood in k-space.
  • the extent of k-space sampled with a gradient pulse played out during acquisition is determined by the gradient height and the gradient pulse width (pulse duration).
  • the spacing between signal samples in k-space is determined by the gradient height and the sampling rate (limited by the maximum speed of the analog to digital converter).
  • the correlation between sequential samples in k-space is determined then by the spacing between samples, by subject motion, by the window used to bracket the acquisition in physical space, and by the underlying texture.
  • [PARA 155] A useful method for selecting the acquisition parameters is with reference to the degree of correlation needed within a set of values to be combined.
  • the underlying textural signals must not be shifted in phase by a significant percentage of ⁇ tex ture relative to each other.
  • the phase shift across the set of samples to be combined should be no greater than 80% of 2 ⁇ .
  • Resolution in MR imaging is limited by subject motion during image acquisition. This limitation can be very severe with non-compliant patients.
  • the resolution achievable in MR imaging which is exemplary art comparable to the present invention, depends on several factors, such as tissue contrast, organ, coil type, proximity to coil. Robust imaging of structures below about 5mm in extent is problematic, and anything below about 1mm is outside the realm of routine clinical imaging. This is a clear shortcoming as many tissue textures in the range of about 5mm down to ⁇ develop and change in response to pathology development, hence measurement of these textures can provide much diagnostic information— these tissue changes are most often the first harbinger of disease. It is this textural wavelength range, from about 5mm down to ⁇ that is targeted with the presently disclosed method.
  • the range of wavelengths in real space which can be resolved i.e. the wavelengths of the textures pertinent to the particular pathology, are in the range of several mm down to microns. This is the range made inaccessible (blurred) in imaging due to patient motion.
  • k is defined as 1/wavelength
  • a range of k-values from about 0.2 mm "1 to 100 mm "1 is employed in exemplary embodiments to define the textures of interest.
  • the method of the embodiments herein for sample acquisition and post processing may all be conducted in k- space.
  • the only localization in real space is the positioning of the VOL Just enough of the neighborhood around a point in real-space is sampled to measure texture— i.e. to determine the power distribution within a neighborhood in k-space around the selected point in real space.
  • [PARA 158] The exact range needed varies with the targeted pathology.
  • [PARA 159] Osteoporotic development in bone microarchitecture.
  • the variation in average trabecular spacing (TbSp) from healthy to osteoporotic bone brackets a wavelength range of about 0.3mm to 3mm; the equivalent range of k-values is 0.34mm- 1 to 3.4mm- 1.
  • TbSp average trabecular spacing
  • k-values 0.34mm- 1 to 3.4mm- 1.
  • Vessel-to-vessel range is 0.4mm to 1.5mm translating to k- values of 0.67mm “1 to 2.5mm "1 while lobule-to-lobule spacing of approximately 1mm to 4mm, translating to k-values from O ⁇ Smm ⁇ tolmm “1 .
  • the acquisition parameters can be chosen such that (1) the gradient height/duration generates a range of k-encodes spanning the neighborhood of k- space over which it is desired to inspect the power density present in the targeted tissue texture, (2) the samples to be combined must occur close enough in time that there is no significant blurring due to subject motion across the acquisition time of a block of samples to be combined, and (3) the tolerable amount of motion depends on the neighborhood of k- space under investigation (i.e., the wavelength).
  • [PARA 161] Acquisition of textural data from within a targeted VOI with the non-zero gradient enables determination of the local variation of power density of k-values within a neighborhood of the initial k- value in k-space.
  • the extent of k-space sampled at each gradient pulse is determined by the gradient height and the pulse width. Spacing between the samples in k-space is determined by the gradient height and the sampling rate.
  • PARA 162 These parameters are selected (1) to allow acquisition of sufficient data for combining toward significant SNR improvement, before subject motion can blur the data significantly relative to the texture to be measured, (2) to ensure sufficient correlation across the blocks of k-values from the acquisition to be combined to maintain a SNR > 0.5dB, and (3) to set the extent of k-space over which the power density of k-values present in the texture is desired.
  • Blocks of sequential signal samples to be recombined for SNR improvement can be non-overlapping, or overlapping by a selected number of points, or a sliding block used so as to combine, for example, measures 1-4, 2-5, 3-6 and so on as will be described subsequently. Additionally, the number of samples in each block may be varied from block to block across the extent in k-space of the acquisition, this variation in number of samples to be combined being determined by the requirement for sufficient correlation to maintain SNR sufficient to provide a robust measurement.
  • the approximate noise level can be determined independently by several methods well known to the industry including measuring noise in the absence of signal input.
  • [PARA 164] Acquiring data with different magnitude gradients within one echo, TR, or scan may be accomplished with the successive gradient heights being selected to enable best SNR of the combined signal at the various targeted regions of k-space.
  • correlation among successive samples can be increased by proper choice of windowing in profile space, a shorter window driving a greater correlation distance across sequential values in k-space and a longer window resulting in lower sample to sample correlation.
  • the window width selected is defined by both the desire for correlation across many samples in k-space, which dictates a shorter window, and the need to sample a sufficient extent of texture in real space to provide robust measure, especially when measuring highly amorphous textures.
  • Non-zero gradient acquisition may be employed to intentionally vary the direction and magnitude of the k- vector over a range during data acquisition, for the purpose of smoothing signal speckle— which will manifest as a time varying signal during the data acquisition— that results from interference of the varying phases and amplitudes of the individual spin signals.
  • the selected variation in k- value direction and magnitude during data acquisition is chosen to provide sufficient combined measures to get a an estimation of the representative power within a neighborhood of k-space, with a SNR of OdB, where the neighborhood is within 20% of the 3D orientation and magnitude of the centroid of the neighborhood.
  • correction for change in k across a set neighborhood created by application of a non-zero gradient may be accomplished by employing proscribed k encodes for a specific set of k measurement. Additionally, correlation within a set of k measurements acquired within a time period and from a selected VOI can be induced by selecting the time period such that the biological motion is sufficiently small that the phase shift in the data induced by patient motion is less than 50% of the wavelength corresponding to the targeted textural k- value range. Alternatively, a windowing function may be selected such that there is sufficient correlation between individual measurements and the set estimate that a desired SNR can be achieved.
  • the k-value is constant at an initial value 1614a, a second value 1614b induced by k value selection gradient pulse 1600a and a third value 1614c induced by k value selection gradient pulse 1600b. Note that the k values are decremented as opposed to incremented in the example of FIG. 12. This is again the previously described pulse sequence where multiple repeats of signal at the same K value are rapidly sampled 1622a, 1622b and 1622c, all of which are then combined into one estimate.
  • a pulse sequence for selecting the desired VOI and initial phase wrap to set the k-value region is illustrated and is as described in FIG. 3 and is numbered consistently in FIG. 20.
  • a higher non-zero gradient 2002 acting as a time dependent phase encode is applied and data samples 2004 are taken from an initial k-value 2006 for more rapidly time varying k-values, seen in trace segment 2008.
  • the initial phase wrap may be selected to provide an initial k-value with a magnitude corresponding to a higher SNR region.
  • the encoding gradient 326 may be employed to wind up to the lowest or highest k-value in a targeted texture and the non-zero magnitude gradient pulse is imposed in the necessary direction (increasing or decreasing k) to reach the other limit in k-space to define the texture.
  • the acquired samples may be outside the inter-sample correlation imposed by the window function. However, signal levels for the k-values are relatively large and have high SNR. Additionally as previously described, rapid acquisition of samples in subsets 2010a, 2010b, 2010c and 2010d as exemplary, may be accomplished in a manner that the samples within the subset may remain sufficiently correlated and may provide desired data in structures having predetermined or anticipated texture. One can combine however many sequential values are correlated enough to yield an improvement in SNR via the combination (averaging being one simple form of combining). Then, the set of combined data points is used to characterize the power distribution across the entire acquisition, to get a better measure of the underlying texture within the VOI.
  • — ⁇ - acts like a low-pass filter to the G (2nk) spectrum, which tends to smooth out the signal: the larger the value of a, the narrower the low-pass filter.
  • X corresponds to a 3D vector in image space
  • g(X) corresponds to the value of the image at a given 3D spatial location
  • G(K) corresponds to the value in k-space of the image g.
  • PARA 189 For initial simplicity, the time-dependency of this signal is ignored which in turn depends upon Tl, T2, T2*, as well as signal contribution due to differing isochromats (different chemical species within the Volume) etc. In the sequel the effect of these is taken into account
  • the rate at which SNR decreases is typically expressed as SNR o lkl " " where a is in the range of 1-3.
  • sampling rate combined with the magnitude of the gradient will set the sample spacing (Ak) density in k-space for a given VOL
  • r represents the real valued 3-Dimensional spatial coordinates with units of meters (m).
  • I(r) represents the image which is a non-negative Real function of spatial coordinates r.
  • k represents the real valued 3-Dimensional k-space coordinates with units in cycles/meter (m- 1 )
  • S(k) represents the Fourier Transform of ' and is generally a complex-valued function of k
  • g(t) is a real- valued 3 -Dimensional function of time representing the gradient strength with units in T/m. This function, is a design input as part of the pulse sequence whose purpose is to manipulate the proton spins in some desired way.
  • equation (6) Making the dependence on time more explicit, equation (6) can be expressed as
  • ' represents the complex-valued baseband signal one might obtain during an MRI echo experiment which is played in conjunction with a gradient sequence encoded ing(t).
  • w is a complex-valued zero-mean
  • the sequence ⁇ " is defined by a sequence of increments which is determined by the integral between samples of the gradient function.
  • a structure may be imposed on the values of by applying a multiplicative window function in the image domain. This is accomplished by leveraging two Fourier Transform identities:
  • Multiplication in one domain corresponds to convolution in the reciprocal domain.
  • the parameter X (and the window function) are chosen so that the resulting weighted sum across the neighborhood of is "wide enough" so that ⁇ where C is a complex- valued constant, but not so wide so as to lose significant spectral resolution
  • [PARA 226] ⁇ ' is treated as an impulse response of a linear filter which is applied to the complex- valued signal in k-space to produce a complex- valued output signal N(k)
  • the autocorrelation function of the output signal NN ⁇ 1 ' 2 ' can be expressed as a function the autocorrelation of the input signal ⁇ ' ⁇ 2 ⁇ , and the impulse
  • R NN (« " 3 ⁇ 4) j j R RR (« ⁇ ⁇ A ⁇ 2 - ⁇ ) ⁇ ( ) ⁇ ( ⁇ ) ⁇ ⁇
  • ⁇ 22 is the autocorrelation function of the impulse response and is given by
  • R 7Z ⁇ K): ° z(k)Z(k + K)dk
  • Equations (50) and (53) provide the defining relationship between the
  • window function impulse response the gradient strength G , the sample interval ⁇ 1 , the number of samples N , and the correlation lower bound ⁇ ram .
  • An exemplary embodiment maintains a constant ratio of textural wavelength to length of VOI acquisition axis. As the targeted k- value varies, the length of the VOI acquisition axis is varied such that the ratio of the corresponding textural wavelength to the acquisition length remains constant. The aim here is to keep the number of textural "cells" sampled constant. In this way the differential broadening observed at specific points in k- space, Ak, is expected to arise from sources other than sampled length in real space, such as the finite width of the RF pulse or the edges of the gradient pulse.
  • MR-based diagnostic Techniques may be combined. Certain MR-based techniques designed to look at very fine tissue structure provide data that may be difficult to interpret in certain pathologies, as they provide only an indirect measure of the underlying structures. Diffusion weighted imaging and Magnetic Resonance Elastography (MRE) are two such techniques. The method of this provisional filing is a direct measure and hence would provide, in many cases, a better measure of fine texture, and in some cases provides complementary data to increase diagnostic capability. Combining acquisition techniques can provide more robust measure of texture, and hence of pathology.
  • MRE Magnetic Resonance Elastography
  • [PARA 241] The method disclosed herein can provide direct measure of fibrotic development in the liver and, as such, would provide additional data on disease progression or response to therapy in the case of the various triggers of fibrotic liver disease. It provides a local measure within the targeted anatomy for calibration of other, indirect measures, such as MRE, DTI, DWI, etc.
  • [PARA 242] The embodiments disclosed may be used in combination with, or replacement for, diffusion weighted imaging in tumors. The ability to detect the edge of tumors with high accuracy would facilitate accurate surgical removal. Can acquire data using the method disclosed herein in VOIs along a direction through a tumor region looking for the edge of the region of angiogenic vasculature.
  • PARA 244 The embodiments disclosed may be used as a bone degradation measure in oncology. It is well known that radiation and/or chemotherapy often compromise bone health. A measure of changes to bone resulting from cancer therapy would help in tailoring therapy and determine if there is need for intervention to protect bone health.
  • [PARA 245] Currently, as a follow on to surgery and treatment for breast cancer, patients are routinely put in the MR scanner to image the breast tissue. The sternum is within the field of view for such exams, enabling easy application of a short add-on sequence of this method to measure changes to trabecular bone and thus obtain a measure of bone health.
  • the embodiments disclosed may be used to measure and quantify hyperplasiac development of mammary duct growth in response to tumor formation and development or to measure and quantify angiogenic growth of vasculature surrounding tumors to stage development, type, and response to therapy.
  • k-value distribution vs. fracture in femur may be evaluated, or changing k- values in cortical neuron bundles can be measured and correlated with performance on Alzheimer's mini-mental state exams or other assessment of AD, or the local k- values in liver compared with other inferences of liver disease over a huge population;
  • Such machine learning can indicate, for example, if a disease is defined by appearance of a specific k-value appearing in the diseased tissue.
  • tissue textures in the brain in both white and grey matter, change in response to disease onset and progression.
  • the ability to measure the early-stage changes in disease, those affecting fine tissue textures, will enable early stage diagnosis, thus enabling earlier treatment, subject targeting for trial inclusion, and sensitive monitoring of therapy response.
  • PARA 250 One of the most valuable features of the disclosed method is that it can be used in conjunction with most contrast methods applied in MRI. As the method results in a texture measurement, as opposed to an image, it needs only to have contrast between the tissue textural elements. This contrast can be generated in many ways, selected to optimize tissue contrast in specific pathologies. The tissue texture measurement yields high spatial resolution due to its relative immunity to subject motion. As the acquisition time for the methods previously disclosed herein is short, the data can be acquired interspersed with image acquisition in various sequences.
  • T2 contrast can be used to highlight fluid towards determining if a bone lesion is lytic or sclerotic, as there may be little fat remaining around the calcified bone to provide signal.
  • T2 weighted imaging has a host of applications, including abdominal lesion imaging, imaging of iron deposition in the brain, and cardiac imaging, hence use in conjunction with the methods previously disclosed herein enables highlighting of tissue texture in these organs/pathologies.
  • MRI contrast generation has become increasingly sophisticated over time.
  • exogenous contrast agents such as gadolinium, and the standard Tl, T2, T2*, proton density contrast, fat suppression, Inversion Recovery sequences as examples.
  • Many new contrast techniques often dependent on functional contrast, have been developed to highlight different tissues involved in pathology.
  • MR angiography a method of visualizing vasculature and blood flow, makes use of MR signal saturation, or induced phase contrast in flowing blood, to assess vascular density and permeability.
  • BOLD Breast Oxygenation Level Dependent contrast uses metabolic changes in blood to image active brain regions.
  • Diffusion weighting both DWI (Diffusion Weighted Imaging) and DTI (Diffusion Tensor Imaging) is used to assess pathology in an increasing range of diseases, providing a signal reflective of the microscopic state of the targeted tissue.
  • ASL Arterial Spin Labelling
  • Perfusion imaging is used to assess blood microcirculation in capillaries, another measure of functional response. In both these contrast schemes, the time-course of blood flow, is followed to assess the state of the vasculature near a tumor, as this is a key feature in the diagnosis of gliomas. Blood vessels are present in higher numbers within tumors than in normal brain tissue, and they tend to have a larger volume.
  • DWI diffusion weighting in its simplest form, uses the random Brownian motion of water molecules to generate contrast in an MR image.
  • pathology pathology
  • Obstacles such as macromolecules, fibers, and membranes also affect water diffusion in tissue. Water molecule diffusion patterns can therefore reveal microscopic details about tissue state. By measuring the differential rate of water diffusion across a region of tissue, a map of diffusion rates, reflecting local pathology, can be produced. Diffusion weighting is particularly useful in tumor characterization, vasculature typing, and diagnosing/monitoring cerebral ischemia, among other pathologies.
  • the methods disclosed herein entail acquisition within either a single VOI or within multiple, interleaved VOIs, within one TR. Data is acquired without use of a spatial encoding gradient to form an image. This significantly shortens the acquisition time and, combined with the narrowly targeted acquisition in k-space, enables acquisition of the requisite data fast enough to provide immunity to subject motion. Though multiple measures of single volume acquisition can be mapped across the anatomy under study, each measure is acquired rapidly within a single volume. High SNR is assured by this single volume technique as, before motion effects set in, there is time for repeat measure of each targeted k- value. The number of repeats and number of/range of k- values for which data is acquired is limited by the requirement to keep the acquisition fast enough to provide the requisite motion immunity.
  • crusher gradients are used on either side of the 180° gradients to eliminate focusing of noise signal generated during the 180° pulse. Replacing these crusher gradients with diffusion weighting gradients, allows acquisition of both the diffusion weighted signal as well as the subsequent restricted k- value signals. As such, the diffusion weighting would be a measure in the VOL
  • DTI diffusion Tensor Imaging
  • diffusion occurs preferentially along one direction, diffusion along the nerve/axon tracts being much preferred to that of the across-track orientation.
  • the degree of directionality, or anisotropy, in tissue is an indicator of pathology, as many neurologic conditions degrade the order of neurologic structures, such as the minicolumns ordering of cortical neurons, or lead to order degradation through demyelination of the axons that form the white matter tracts in the brain.
  • anisotropic diffusion the value of the diffusion constant varies with direction.
  • anisotropy is a measure of pathology advancement
  • measurement of the diffusion constant in multiple directions can be used to yield the "fractional anisotropy" arising from the tissue structures and hence provide a measure of pathology advancement.
  • at least six non-collinear gradients are used to measure fractional anisotropy, leading to a symmetric 9 x 9 matrix, the "diffusion tensor", the eigenvalues of which yield the major diffusion axes in the 3 orthogonal directions.
  • the diffusion tensor mapped across the brain can be used to delineate the path of white matter tracts. This is called tractography.
  • tractography A possible application of the methods disclosed herein is to measure texture in the white matter tracts affected in multiple sclerosis (MS) using standard Tl or T2 contrast or using the method disclosed herein in conjunction with diffusion weighting to determine anisotropy of the measure for correlational input for machine learning with standard DTI acquisitions.
  • a contrast is applied using any one of the previously described mechanisms enhancing the contrast between the component tissue types in a multiphase biologic sample being measured.
  • the contrast mechanism and its application may occur at various locations within the NMR inducing pulse sequence.
  • pulse sequencing such as that described with respect to FIGs. 3 and 8 a volume of interest (VOI) is selectively excited employing a plurality of time varying radio frequency signals and applied gradients.
  • An encoding gradient pulse as also previously described with respect to FIGs.
  • a time varying series of acquisition gradients is initiated to produce a time varying trajectory through 3D k-space of k- value encodes as previously described with respect to, for example, FIGs. 8, 11, 16 or 18, with the k-value set being a subset of that required to produce an image of the VOI.
  • Multiple sequential samples of the NMR RF signal encoded with the specific k-values are simultaneously recorded. The recorded NMR signal samples are then post-processed to produce a data set of signal vs k-values for the k- values in the set determined by the trajectory.
  • a first tissue for application of the methods herein is bone. Though the effect on quality of life is huge, no accurate and non-invasive method for sensitive assessment of bone fracture likelihood exists.
  • the current gold standard measure, DEXA which relies on x-ray absorption, measures areal density of bone.
  • the main determinant of bone strength which predicts fracture, is trabecular microarchitecture, a measure of which is not currently available in vivo. The method disclosed herein enables this measure.
  • TbTh the trabecular element thickness
  • TbSp the repeat spacing of trabecular cells
  • TbN the trabecular number
  • TbSp increases faster along the primary load bearing direction than it does normal to this direction, the variation between the two measures being a marker for bone degradation.
  • variability in the measure of the trabecular spacing, TbSp increases, due to the thinning of the trabecular struts to the point where they break, causing discontinuities in the measure of TbSp.
  • TbSp can be used as a correlative measure for fracture likelihood
  • Data can be acquired using the method disclosed herein by positioning the VOI within the targeted bone region, acquiring data within a single TR or across multiple TRs to enable acquisition across the pertinent range of k-values spanned by TbTh and TbSp. Data acquired quickly within one TR can be averaged to improve SNR, the only requirement being that the data is acquired from similar bone tissue.
  • PARA 267 The requisite contrast is between bone and marrow. Tl weighting provides high signal in marrow, bone offering negligible MR signal. Alternatively, an IR sequence, which results in heightened Tl contrast, can be used.
  • TbSp and TbTh are closer in magnitude than they are in diseased bone. This can be seen in FIGs. 22A and 22B, by comparing the image of healthy bone, FIG. 22A, with the image of highly osteoporotic bone, FIG. 22B. The exact form of this relationship varies somewhat, the difference in these two morphometric parameters being higher in spine, for instance, than in the hip across pathology. The increasing difference in measure of these two morphometric parameters provides a marker of disease development.
  • the methods disclosed herein can be applied within and around the location of a bone lesion identified with, for example, conventional Tl, T2 or proton density contrast, or flow or diffusion contrast MR imaging, to assess the state of the trabecular bone in the region.
  • the lesion type is the lesion indicative of an erosive tumor, or is the lesion in a region of inflammation/ degraded bone surrounding a fracture.
  • Some lesions are dangerous and erosive tumors, some lesions are benign.
  • Acquiring data by the method disclosed herein in multiple VOIs in the area of the lesion would enable a determination of whether the trabecular structure is degraded in the vicinity, and how progressed the degradation is, both spatially and temporally.
  • Further biomarkers can be derived by inputting the MR images of lesions to machine learning algorithms and correlating them with the trabecular data of TbTh, TbSp, TbN, anisotropy and measure variability.
  • T2 contrast can be used in conjunction with the method disclosed herein to highlight fluid in an oncological bone lesion to type it as either lytic or sclerotic. In such pathology there may be little fat/marrow remaining around the calcified bone to provide signal. In an oncologic bone lesion there is usually a mix of fluid, and of marrow in various states of inflammation. To yield a signal from outside the hard bone, proton density can be used. Alternatively, diffusion weighting would return signal based on diffusion of water molecules in the fluid imbued marrow phase.
  • Biomarkers can be derived directly from the signal vs. k-value data by inputting it into machine learning algorithms and correlating it with, for instance, fracture occurrence data from the same patient, DEXA measurements, or bone biopsy report data.
  • liver tissue or other tissues subject to fibrotic invasion provide a second example of the use the methods disclosed herein.
  • the salient feature of the disease is the development of fibrotic depositions within the liver. Disease onset and progression are marked by increasing accumulation of proteinaceous deposits, mainly collagen fibers, on the hepatic structures.
  • fibrotic development can, in the short term, promote healing, if the disease is left untreated, the healing response itself becomes a problem, the excess of proteinaceous substance impeding the normal functioning of the organ. In the case of liver disease, this process, left unchecked, advances to cirrhosis— with attendant carcinoma and/or liver failure.
  • liver fibrosis the pathology, and the application of the method disclosed herein, is similar to that of a range of diseases characterized by fibrotic invasion.
  • a partial list of these diseases cardiac fibrosis, cystic fibrosis, idiopathic pulmonary fibrosis, pancreatitis, kidney disease.
  • pathologies such as prostate disease, lose proteinaceous deposits in response to disease progression. Though the mechanism is reversed, the tissue texture assessment needed for diagnosis and monitoring is the same.
  • fibrotic development starts on the portal triads, then progresses, eventually forming bridges linking the portal triads to the central veins (Stages Fl through F3). These bridges enlarge and coalesce, forming islands of regenerative tissue surrounded by fibrotic deposition.
  • vessel to vessel structural spacing in tissue texture becomes gradually replaced by lobule to lobule spacing (Stages F3 to F4).
  • a clear marker of disease progression is the shift in distribution of textural wavelengths from shorter to longer wavelength (decreasing k- value), this shift being from about 0.5mm to about 2mm, and often longer.
  • Contrast between the collagen decorating the various hepatic structures and the underlying tissue can be achieved using either endogenous or exogenous contrast: signal from fibrosis is dark in standard Tl imaging, and can be bright in T2 imaging, due to the large water content in the fibrotic structure. Use of Gd contrast agent shortens Tl such that on Tl weighting the fibrosis shows up bright against the background tissue. Higher contrast makes for a more robust measure.
  • standard MR imaging is not capable of sufficient resolution to discern the pattern of fibrotic development on the various hepatic structures, which characterizes early stage disease. Patient motion blurs the image, even when using breath-hold imaging or respiratory triggering.
  • liver disease can only be assessed at the more advanced stages, when the liver may be irreversibly damaged. Though advanced disease is diagnosable, what is needed for therapy justification and response monitoring is early stage diagnosis.
  • MRE Magnetic Resonance Elastography
  • DWI diffusion-weighted imaging
  • MR perfusion imaging can yield some information on liver disease, though none are capable of robust diagnosis in the earlier stages of the disease.
  • a major difficulty with them is that they rely on surrogate markers for fibrotic development.
  • MRE relies on stiffness measurement
  • perfusion imaging measures blood perfusion parameters
  • DWI Diffusion Weighted Imaging
  • ADC Apparent Diffusion Coefficient
  • PARA 284 One of the features of the method disclosed herein that makes it novel, is that it can be used in conjunction with most contrast mechanisms.
  • One application of this method is its use in conjunction with diffusion weighting, using diffusion-weighted contrast (see FIG. 23 below), but acquiring signal only in the k- value ranges of the fibrotic deposits in early stage disease rather than the entire image acquired in standard DWI.
  • the data is acquired with a much finer spatial resolution than is possible with diffusion-weighted MR imaging.
  • the texture being measured is on the scale of the fiber-decorated structures, between actual fiber clumps, rather than an averaged measure affected by partial volume imaging.
  • Fibrotic deposition lowers the diffusion coefficient for water, the lower ADC (apparent diffusion coefficient) in areas of fibrosis making for a brighter signal than the surrounding tissue.
  • the structural signal obtained will measure highly localized water diffusion.
  • the lobular unit transforms from one with no delineated boundaries to collagen decoration of the hexagonal boundary and then to filling in the entire lobule, water diffusion at the boundaries will be impeded, increasing diffusion weighted signal intensity.
  • a textural signature can be obtained by using diffusion contrast.
  • Two approaches to positioning of the diffusion weighting gradients are as shown in FIGs. 24 and 25.
  • the shown pulse sequence for selecting the desired VOI and initial phase wrap for k is as described in FIG.
  • FIG. 24 shows positioning of the diffusion weighting gradients 2402, 2404 on either side of the second 180° slice selection pulse, while in FIG. 25, the diffusion weighting gradients 2502, 2504 are positioned before the first, and after the second, 180° slice select pulses to provide more diffusion time for the same TE than would be available when placing the pair either side of the last 180° slice-selection pulse.
  • Fibrotic texture development can also be assessed using the methods disclosed herein in conjunction with contrast, such as Tl or T2 weighting, with or without exogenous agents such as Gd.
  • contrast such as Tl or T2 weighting
  • exogenous agents such as Gd.
  • the methods disclosed herein enables fast acquisition of signal at the requisite k-values, enabling robust assessment of pathologic tissue texture at a specific location in the liver— providing a measure of the textural frequencies present at that location, immune from the subject motion-induced blurring that limits current MR imaging methods.
  • the problem of respiratory motion is circumvented by the speed of acquisition of the requisite data.
  • VOIs 2602a-2602d can be positioned at various locations in the liver as represented in FIG. 26, using either interleaved acquisition within one TR, or measurements in separate multiple TRs.
  • texture coherence is maintained within each defined VOI throughout the data acquisition at specific k-values or k- value ranges, so as to enable SNR maximization through signal averaging.
  • repeat sampling can be made in subsequent TRs at the same location, to obtain an average measure of the degree of fibrotic invasion at that location, for assessment of stage of disease progression.
  • Fibrotic development is the hallmark of lung disease (e.g. cystic fibrosis, idiopathic pulmonary fibrosis), myocardial fibrosis, muscle fibrosis, pancreatic fibrosis and kidney disease.
  • lung disease e.g. cystic fibrosis, idiopathic pulmonary fibrosis
  • myocardial fibrosis e.g. myocardial fibrosis
  • muscle fibrosis e.g. cystic fibrosis, idiopathic pulmonary fibrosis
  • pancreatic fibrosis e.g. cystic fibrosis, idiopathic pulmonary fibrosis
  • kidney disease e.g. cystic fibrosis, idiopathic pulmonary fibrosis
  • myocardial fibrosis e.g. cystic fibrosis, idiopathic pulmonary fibrosis
  • myocardial fibrosis e.g. cystic fibrosis,
  • PARA 287 Many neurologic diseases and conditions have a vascular component that may serve as a marker for disease onset and progression, allowing diagnosis and therapy tracking which provides a third exemplary implementation of the methods disclosed herein.
  • the ability to sensitively assess changes in micro-vessels would enable monitoring of pathology progression in a number of diseases which are often not diagnosed until pathology is well advanced.
  • Angiogenesis formation of new blood vessels from pre-existing micro vessels, is necessary for tumor growth and metastasis. Rather than the ordered formation of vessels that exist in healthy tissue, pathogenic angiogenesis tends to form chaotic, tortuous vessels, replete with blocked, dead end structures— see FIG. 23. Vessel diameter and wall thickness are highly variable in angiogenic micro- vessels, with marked vessel permeability in places.
  • AD Alzheimer's disease
  • HD Huntington's disease
  • PD Parkinson's disease
  • Frontotemporal dementia are also found to have
  • the salient cause of the dementia appears to be pathogenic vasculature in the brain, such as CVD.
  • CVD pathogenic vasculature
  • PARA 291 Chronic inflammation is another important factor that can lead to abnormal neurovascular structure, exhibiting permeability and hemorrhage. Some microvasculature pathogenesis is linked to permeability of the blood brain barrier. Multiple Sclerosis, a brain disorder with pathology associated with inflammation and axonal demyelination, exhibits microvessel disruption. Stroke and the resultant ischemia result in development of angiogenesis modifying the capillary network, as the body attempts to heal the damage.
  • angiogenesis features increased vascularity, involving both structural and functional alterations within the neurovascular system, this increased density, and the high variability in vessel spacing, is a promising biomarker that can be used for the characterization of ischemic conditions in the brain following stroke.
  • tumor development, ischemic stroke, and brain pathology in dementia a means of assessing the micro- vasculature in brain tissue, is needed— both for determination of pathology advancement, and for assessment of therapy response.
  • Perfusion is the irrigation of tissues via blood delivery through the microvessels. Because the state of the vessels changes the dynamics of blood flow, such measure can be used to assess vascular health.
  • endogenous contrast is most commonly provided by use of Gd-based contrast agent.
  • Endogenous contrast is obtained through a technique known as ASL (Arterial Spin Labelling) in which the blood flowing into a region of the brain is magnetically labelled. In both cases, sequential images are made via fast imaging techniques, as the contrast moves into, and exits, the imaging plane.
  • ASL Arterial Spin Labelling
  • differential contrast can be obtained via subtraction of the image obtained with no contrast agent/blood tagging in the imaging plane from that obtained when contrast is at a maximum in the imaging plane.
  • PARA 293 When using a contrast agent for this measure, a bolus of the agent is injected intravenously and successive images are acquired as the contrast agent passes through the microcirculation.
  • At least one image is taken following passage of the bolus, or of the labelled spins, through the microvasculature, when contrast between the blood and surrounding tissue is minimal. This image is then subtracted from the early images to allow calibration of the absolute signal level from the
  • Angiogenic vasculature is denser, and more varied in vessel diameter and spacing, than are healthy vessels.
  • the high spatial variation in vessel thickness and spacing is one of the hallmarks of angiogenic vasculature and hence, along with increased vessel density, serves as a marker for angiogenesis-related pathology.
  • image resolution in perfusion imaging is not high enough to determine detailed vascular morphology.
  • Flow contrast highlights locally averaged signal variation due to the pathogenic flow parameters, offering indirect assessment of vessel morphometry.
  • the methods disclosed herein can be used to directly measure vessel density, and vessel spacing variability, to provide direct, robust assessment of angiogenic vessel development. Using the methods disclosed herein to acquire signal vs.
  • k- value data disclosing tissue texture enables robust resolution of the morphometric features of the vessels.
  • This acquisition can be done in one TR, fast enough that the sequence can be injected into the multi-image-acquisition perfusion series. To provide best resolution, this morphometry acquisition would be done near peak contrast, either acquiring data for one TR or for multiple TRs acquired either sequentially or interspersed at various time points with the perfusion image acquisitions.
  • hybrid acquisition is possible wherein a gradient is on for some part of the data acquisition within one TR and off for part of the acquisition.
  • the aim here is to, while acquiring a range of k-values, ensure sufficient repeats of, say, a set of highly correlated k-values, to allow SNR maximization by averaging, while ensuring fast enough acquisition to provide immunity to subject motion.
  • the methods disclosed herein can be used in conjunction with any blood contrast method to measure vessel morphology directly, in regions exhibiting pathogenic flow parameters.
  • vessel spacing and the variability in this measure are known markers of angiogenesis, the vasculature spacing becoming more random with degree of pathology.
  • structural contrast such as T2 or Tl weighting, which yield, respectively, bright or dark blood
  • black blood and bright blood flow contrast can be achieved by various standard methods, including arterial spin labelling. This structural measure of the vessels can be carried out in as many tissue regions, using as many acquisition directions, as desired.
  • Angiogenic vasculature would be expected to exhibit a high degree of anisotropy, so varying the orientation of the acquisition axis between acquisitions provides another marker of pathology. Correlation of the flow data from the perfusion imaging with the structural vessel data from application of the methods disclosed herein, can be made through machine learning.
  • the VOI can be positioned in the vasculature near the cortical region(s) implicated in the dementia. Data can be acquired in one or more VOIs in one TR. Additionally, in scanners capable of parallel imaging, multiple VOIs can be defined with simultaneous record of data to sample extended areas of the brain vasculature. For example, in dementia in which multiple cortical regions seem to be damaged, VOIs can be placed in the vasculature feeding these different regions and data recorded simultaneously.
  • the correlational data could be the output from any of those tests individually. With sufficient number of cases, this would enable finer gradations to be defined in disease staging, for instance, steps in between each stage— F0 and Fl, between F0 and F2, and between F0 and F3, would be possible to define using this method.
  • outcomes progression to more advanced pathology, or therapy-induced healing— can provide correlational data for machine learning algorithms for correlation with textural assessment from application of the methods disclosed herein. .
  • [PARA 305] The assessment stage obtained by the methods disclosed herein can be mapped on top of a standard MRI morphology image of the diseased liver. (For easier viewing, an icon can replace the staging number.) This will facilitate visualization of the disease variability through the organ. Additionally, these staging values can be correlated by machine learning with the imaging output obtained on the same patient by MRE, standard DWI, or perfusion, for instance to track possible correlations.
  • a final example of pathology assessment employing the methods herein is brain tissue. Brain pathology is often problematic to diagnose and treat because of the sensitivity of the organ to intervention. Further, changes in cognition and behavior can occur over a long time span such that the underlying pathology can go unchecked for years.
  • AD Alzheimer's disease
  • the predictive value of the novel biomarker provided by the textural data acquired by the method disclosed herein, towards assessment of the degree of AD pathology, can be defined by correlation with a range of diagnostic information from the same patient.
  • the main correlation marker will be drawn from patient outcomes— i.e. definitive diagnosis of AD or other dementia— as this has the highest diagnostic information content, though definitive diagnosis is well-downstream from the pathology we are assessing.
  • Additional correlation will be drawn from patient MRI imaging data on hippocampal shrinkage, a proven, and continuous, marker of advancing AD (as well as other forms of dementia). This correlation will be made longitudinally with disease progression, if possible.
  • a third correlative marker is FDG-PET, as decline in glucose metabolism is expected to occur early relatively in disease progression.
  • the MMSE Mini Mental State Exam
  • Genetic predilection for AD provides an additional marker for correlation with the textural measure acquired by the method disclosed herein. While the previous markers provide downstream correlative value (on the outcome side), genetic markers exist in advance of any pathology development. Correlation of this varied set of biomarkers with the data acquired by the method disclosed herein in the hippocampus and entorhinal cortex, across a broad range of patients, will enable a clear definition of diagnostic content of the use of the method disclosed herein for early stage prediction of AD pathology.
  • PARA 310 Current machine learning algorithms are capable of pathology level classification of non-specific features, as will be obtained from MR data acquisition by the method disclosed herein. As such, the disclosed sources of correlational data above will be input into machine learning algorithms to highlight the correlation with textural features and disease.
  • the hippocampus may be the earliest affected cortical structure with AD progression, its depth within the brain results in lower SNR due to distance from the MR sensing coil. Texture within the neocortex provides a target for assessment of dementia and other brain pathology that, due to its proximity to the skull, offers higher SNR.
  • very ordered neuronal architecture is found in the neocortex. The neurons form in bundles of approximately 50microns in width and 80 micron spacing, with about 80 to 100 myelinated neurons grouped together in each bundle. This is the minicolumn organization visible in histology of neocortical tissue. In specific regions of the brain that are seen by histology studies to be affected early in AD
  • AD Alzheimer's Disease
  • FIG. 27 The structure of these minicolumns in healthy brain can be seen in FIG. 27, a histology image stained to reveal myelin 2702— the coating sheathing the neurons.
  • FIGs. 29A - 29C are a series of three histology images stained to reveal the pyramidal neuron cells in the bundles.
  • FIG. 29 A is of neuronal order in healthy brain and
  • FIGs. 29B and 29C show progressive pathology with AD advancement— the columnar spacing shrinks and the ordered structure becomes increasingly random.
  • FIG. 28 is a representation showing possible positioning of the VOI 2802a, 2802b, 2802c and 2802d in the neocortex 2804.
  • PARA 315 As the axonal component of the neurons forming the minicolumn bundles are sheathed in myelin, a fatty substance, Tl contrast can be used to highlight the axon bundles against the background tissue and hence is a good choice for contrast when assessing these structures.
  • a VOI is positioned in the center of the cortex height, aligning the acquisition axis parallel to the top and bottom surfaces at the VOI midpoint, as closely as possible. 6) Signal vs. k-value data is then acquired with the gradient on or off to measure the minicolumn spacing; measurement across a broad range of k-values encompassing on the average spacing of the minicolumns as indicated in the literature (approximately 80 ⁇ ) will ensure coverage of the width distribution. Signal intensity maximum should occur when the acquisition gradient is oriented normal to the columns. 6) The acquisition gradient is then rocked in small angular increments to look for the signal resonance— the sharpness of the signal resonance vs. angular deviation reflects the order of the minicolumns.
  • a sharp resonance indicates ordered structure.
  • a broad resonance as a function of angular deviation indicates columnar degradation has introduced randomness into the minicolumn order.
  • a sharp peak (high q-value) in the signal vs. k-value curve indicates ordered structure, the broadness of the curve is indicative of the degree of loss of order. Locating the resonance in the signal vs. acquisition angle and in the signal vs. k- value distribution can be accomplished as an interactive process.
  • [PARA 320] Data can be acquired at other positions in the cortex or nearby the original VOI, either within one TR or multiple TRs.
  • Optimal VOI dimensions for characterization of the cortical minicolumns are determined by 1) the need to fit the VOI entirely within the cortex, which is 2-3 mm in height, 2) the requirement to sample sufficient textural repeats along the encode axis for accurate assessment of the textural wavelength, and 3) by signal requirements. Additionally, the
  • a change in spacing of neuronal columns indicates pathology advancement /aging; this can be determined by longitudinal monitoring of the peak of the signal- magnitude vs. K- value distribution.
  • a variation on this disease marker is the degree of anisotropy of the columnar order. As the columnar order degrades with progressing pathology, the degree of anisotropy of the columnar texture also lessens and the overall cortical tissue texture becomes more isotropic.
  • the degree of anisotropy can be measured by use of Tl, or other, contrast using the method disclosed herein, with the VOI 3002 positioned as above, midway between the two cortical surfaces 3004 as seen in FIG. 30 and comparing the signal vs It- value distribution with the acquisition axis normal to the cortical surfaces (parallel to the minicolumns), with the signal vs. k-value distribution obtained with the acquisition axis aligned tangential to the cortical surfaces 3004 therefore (normal to the minicolumns) as shown in FIG. 30.
  • Diffusion weighted imaging provides an indirect measure of structure at the cellular level, by applying gradients that first dephase and then rephase signal in a targeted location.
  • the difficulty with the technique is that, by design, it is extremely sensitive to motion. It is also low SNR, due to the late echo time resulting from the need for the long diffusion gradients.
  • Use of the method disclosed herein for data acquisition when using diffusion contrast can remedy the motion problem as, though the echo time is still long, data acquisition is fast enough that the signal loss and blurring due to motion is minimized.
  • the method disclosed herein can be used with diffusion weighting contrast to assess the spacing and order/randomness of the minicolumns.
  • This measure can be made with the diffusion gradients applied parallel to the surfaces of the cortex (normal to the minicolumns), and then normal to the cortical surfaces (parallel to the minicolumn direction). These two measures enable assessment of the anisotropy, which will be highest in healthy brain, pathology then inducing increasing isotropy as the columns degrade.
  • a refinement on this measure is through application of the diffusion gradient in multiple directions for data acquisition and development of the diffusion tensor similarly to diffusion tensor imaging (DTI), but with data acquisition being by the method disclosed herein.
  • Development of a diffusion tensor requires using at least 6 non-collinear directions of diffusion gradient orientation to yield sufficient data to generate the tensor, the eigenvalues of which determine the level of Fractional Anisotropy (FA) in the cortex, reflective of the order of the minicolumns.
  • the FA should change, moving toward more isotropic organization as columnar organization degrades— an FA value of 1 indicates highest anisotropy, and a value of 0 indicates maximum isotropy of the underlying diffusion, hence revealing the order of the columnar texture.
  • the targeted k-values are selected by a combination of knowledge of the approximate location in k-space of the minicolumns from the literature and from measure when they are still sufficiently ordered to define a clear textural wavelength signature, and pre-measure to determine the distribution of signal vs. k- value.
  • One method to achieve this is with a gradient on to provide sufficient spread in k-space during data acquisition to enable finer sampling on the subsequent acquisition(s), though this measure can also be made using gradient off acquisition.
  • the mean diffusivity (MD) of water is found to decrease with increasing dementia.
  • MD mean diffusivity
  • the signal vs. k- value data obtained using the method disclosed herein can be input into a machine learning algorithm, with correlational data from cognitive evaluation tests such as the MMSE exam, downstream neuropathology outcomes, and serum and imaging data.
  • the method disclosed herein can be used to assess degree of pathology in any of these conditions. Correlation for machine learning to determine the association between measured data and pathology can be obtained from cortical atrophy segmentation, MMSE, doctor' s evaluation of degree of pathology from observational data, etc.
  • the method disclosed herein can be used with stationary contrast mechanism to highlight tissue texture changes and flow contrast to highlight vasculature changes in the vicinity of a lesion showing up on an MR image, that may indicate stroke or tumor related pathology.
  • the predictive value of the novel biomarker provided by the textural data acquired by the method disclosed herein, towards assessment of the degree of AD pathology, can be defined by correlation with a range of diagnostic information from the same patient.
  • the main correlation marker will be drawn from patient outcomes— i.e. definitive diagnosis of AD or other dementia— as this has the highest diagnostic information content, though definitive diagnosis is well-downstream from the pathology we are assessing.
  • Additional correlation will be drawn from patient MRI imaging data on hippocampal shrinkage, a proven, and continuous, marker of advancing AD (as well as other forms of dementia). This correlation will be made longitudinally with disease progression, if possible.
  • a third correlative marker is FDG-PET, as decline in glucose metabolism is expected to occur early relatively in disease progression.
  • the MMSE Mini Mental State Exam
  • Genetic predilection for AD provides an additional marker for correlation with the textural measure acquired by the method disclosed herein. While the previous markers provide downstream correlative value (on the outcome side), genetic markers exist in advance of any pathology development. Correlation of this varied set of biomarkers with the data acquired by the method disclosed herein in the hippocampus and entorhinal cortex, across a broad range of patients, will enable a clear definition of diagnostic content of the use of the method disclosed herein for early stage prediction of AD pathology.

Landscapes

  • Health & Medical Sciences (AREA)
  • Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Molecular Biology (AREA)
  • High Energy & Nuclear Physics (AREA)
  • Biophysics (AREA)
  • Biomedical Technology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Medical Informatics (AREA)
  • Radiology & Medical Imaging (AREA)
  • Surgery (AREA)
  • Animal Behavior & Ethology (AREA)
  • Pathology (AREA)
  • Public Health (AREA)
  • Veterinary Medicine (AREA)
  • Condensed Matter Physics & Semiconductors (AREA)
  • General Physics & Mathematics (AREA)
  • Signal Processing (AREA)
  • Optics & Photonics (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Magnetic Resonance Imaging Apparatus (AREA)

Abstract

Les modes de réalisation de l'invention portent sur un procédé pour l'acquisition de données RM à des résolutions pouvant descendre jusqu'à des dizaines de microns pour l'application en diagnostic in vivo et la surveillance d'une pathologie pour laquelle des changements de fines textures de tissu peuvent être utilisés comme marqueurs de l'apparition et de la progression d'une maladie. Les maladies osseuses, les tumeurs, les maladies neurologiques et les maladies impliquant la croissance et/ou la destruction des fibres sont toutes des pathologies cibles. En outre, la technique peut être utilisée dans n'importe quel système biologique ou physique pour une caractérisation très haute résolution d'une morphologie à petite échelle. Le procédé concerne l'acquisition rapide de valeurs sélectionnées dans l'espace k, avec plusieurs acquisitions successives de valeurs k individuelles prises sur une échelle de temps de l'ordre de la microseconde, dans un volume de tissu défini et la combinaison subséquente des multiples mesures de manière à augmenter au maximum le rapport signal/bruit. Le volume d'acquisition réduit et l'acquisition uniquement des valeurs sélectionnées dans l'espace k le long de directions sélectionnées permettent une bien meilleure résolution in vivo que ce qu'il est possible d'obtenir avec les techniques d'IRM actuelles.
PCT/US2016/056147 2015-10-07 2016-10-07 Échantillonnage sélectif pour évaluer des fréquences spatiales structurales avec des mécanismes de contraste spécifiques WO2017062882A1 (fr)

Priority Applications (5)

Application Number Priority Date Filing Date Title
EP16854502.8A EP3359034A4 (fr) 2015-10-07 2016-10-07 Échantillonnage sélectif pour évaluer des fréquences spatiales structurales avec des mécanismes de contraste spécifiques
AU2016334250A AU2016334250B2 (en) 2015-10-07 2016-10-07 Selective sampling for assessing structural spatial frequencies with specific contrast mechanisms
CA3000765A CA3000765C (fr) 2015-10-07 2016-10-07 Echantillonnage selectif pour evaluer des frequences spatiales structurales avec des mecanismes de contraste specifiques
JP2018517540A JP6906507B2 (ja) 2015-10-07 2016-10-07 特定のコントラスト機構を用いて構造的空間周波数を評価するための選択的サンプリング
CN201680071304.4A CN108366753B (zh) 2015-10-07 2016-10-07 用于评估具有指定对比度机制的结构空间频率的选择性采样

Applications Claiming Priority (8)

Application Number Priority Date Filing Date Title
US201562238121P 2015-10-07 2015-10-07
US62/238,121 2015-10-07
US201662302577P 2016-03-02 2016-03-02
US62/302,577 2016-03-02
US15/167,828 2016-05-27
US15/167,828 US9664759B2 (en) 2014-09-01 2016-05-27 Method for assessing structural spatial frequencies using hybrid sampling with non-zero gradient for enhancement of selective sampling
US201662382695P 2016-09-01 2016-09-01
US62/382,695 2016-09-01

Publications (1)

Publication Number Publication Date
WO2017062882A1 true WO2017062882A1 (fr) 2017-04-13

Family

ID=58488603

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2016/056147 WO2017062882A1 (fr) 2015-10-07 2016-10-07 Échantillonnage sélectif pour évaluer des fréquences spatiales structurales avec des mécanismes de contraste spécifiques

Country Status (6)

Country Link
EP (1) EP3359034A4 (fr)
JP (1) JP6906507B2 (fr)
CN (1) CN108366753B (fr)
AU (1) AU2016334250B2 (fr)
CA (1) CA3000765C (fr)
WO (1) WO2017062882A1 (fr)

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2019010381A1 (fr) * 2017-07-06 2019-01-10 bioProtonics, LLC Procédé pour mesurer une texture de tissu en utilisant la spectroscopie rmn pour identifier des espèces chimiques d'éléments de texture de composant dans une région ciblée de tissu
WO2020136571A1 (fr) * 2018-12-26 2020-07-02 Analytics For Life Inc. Procédés et systèmes de configuration et d'utilisation de réseaux neuronaux dans la caractérisation de systèmes physiologiques
JP2020520786A (ja) * 2017-05-24 2020-07-16 バイオプロトニクス インコーポレイテッドBioprotonics,Inc. 特定のコントラスト機構を用いて構造的空間周波数を評価するための選択的サンプリング
CN112184845A (zh) * 2020-09-02 2021-01-05 东软医疗系统股份有限公司 弥散加权图像重建模型的生成方法及装置
US11341734B2 (en) 2018-12-17 2022-05-24 Shanghai United Imaging Intelligence Co., Ltd. Systems and methods for image segmentation
WO2023055471A1 (fr) * 2021-09-29 2023-04-06 The Johns Hopkins University Identification automatisée de grappes de coefficients de diffusion apparents à hyperintensité pour une thérapie thermique interstitielle laser guidée par irm
CN116112139A (zh) * 2018-09-21 2023-05-12 中兴通讯股份有限公司 波束失败恢复方法及装置

Families Citing this family (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3874288A4 (fr) * 2018-11-20 2022-12-28 Bioprotonics, Inc. Procédé de mesure de texture tissulaire à l'aide d'une spectroscopie par résonance magnétique nucléaire (rmn) avec une longueur de volume d'intérêt (voi) dans une direction d'analyse définie par une bande passante de récepteur
US20220022804A1 (en) * 2018-12-10 2022-01-27 Oxford University Innovation Limited Brain imaging
CN109754394B (zh) * 2018-12-28 2021-02-23 上海联影智能医疗科技有限公司 三维医学图像处理装置及方法
JP7220727B2 (ja) * 2019-01-31 2023-02-10 富士フイルム株式会社 学習済みモデル、学習方法、及びプログラム、並びに医用情報取得装置、方法、及びプログラム
CN110322408A (zh) * 2019-06-11 2019-10-11 浙江大学 基于云平台的多中心磁共振图像自动质量控制方法
CN110599447B (zh) * 2019-07-29 2023-10-20 广州市番禺区中心医院(广州市番禺区人民医院、广州市番禺区心血管疾病研究所) 一种肝癌病灶数据的处理方法、系统和存储介质
CN110501664B (zh) * 2019-08-23 2021-07-16 苏州朗润医疗系统有限公司 磁共振波谱相位校正方法及应用该方法的磁共振成像系统
TWI754858B (zh) 2019-11-21 2022-02-11 國立中央大學 光學斷層造影系統及造影方法
US11085979B1 (en) * 2020-02-24 2021-08-10 Uih America, Inc. System and method for magnetic resonance imaging with radial sampling of k-space
CN111803068A (zh) * 2020-08-21 2020-10-23 上海中医药大学附属曙光医院 一种基于虚拟磁共振弹性成像的无创评估肝纤维化的方法
CN112304811B (zh) * 2020-10-22 2023-02-03 无锡鸣石峻致医疗科技有限公司 一种对运动不敏感的低场核磁共振表观扩散系数测量方法
CN112958840B (zh) * 2021-02-10 2022-06-14 西南电子技术研究所(中国电子科技集团公司第十研究所) 精密零件加工削力信号自动分段方法
CN113156350B (zh) * 2021-03-25 2022-07-19 华南理工大学 无创光不透明微米级活体组织磁共振断层成像方法及系统
CN116509453B (zh) * 2023-06-29 2023-09-01 南京科进实业有限公司 一种基于超声波的胫骨骨密度检测系统及方法

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5565777A (en) * 1993-09-13 1996-10-15 Kabushiki Kaisha Toshiba Method/apparatus for NMR imaging using an imaging scheme sensitive to inhomogeneity and a scheme insensitive to inhomogeneity in a single imaging step
US20070167717A1 (en) * 2005-11-27 2007-07-19 James Timothy W Structure assessment using spatial-frequency analysis
US20120262171A1 (en) * 2011-04-15 2012-10-18 Hans Weber Method of magnetic resonance imaging for the selection and recording of curved slices
US20160274203A1 (en) 2014-09-01 2016-09-22 bioProtonics, LLC Method for assessing structural spatial frequencies using hybrid sampling with non-zero gradient for enhancement of selective sampling

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS63130056A (ja) * 1986-11-21 1988-06-02 旭化成株式会社 部分選択核磁気共鳴情報導出方法
DE4334038C1 (de) * 1993-10-06 1995-03-23 Bruker Medizintech Geräuscharmes NMR-Bildgebungsverfahren mit Einzelpunktaufnahme (SPI)
US20060155186A1 (en) * 2005-01-12 2006-07-13 James Timothy W Bone health assessment using spatial-frequency analysis
US7602180B2 (en) * 2006-11-02 2009-10-13 Siemens Aktiengesellschaft System and method for fast texture-based tensor field visualization for DT-MRI
EP2756323A1 (fr) * 2011-09-13 2014-07-23 Acuitas Medical Limited Procédé fondé sur la résonance magnétique pour évaluation de la maladie d'alzheimer et de pathologies associées
US9360542B2 (en) * 2012-04-20 2016-06-07 Wisconsin Alumni Research Foundation System and method for spectrally-resolved three-dimensional magnetic resonance imaging without frequency-encoding gradients
CN106716167B (zh) * 2014-09-01 2019-03-12 生物质子股份有限公司 用于评估结构空间频率的基于选择性采样磁共振的方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5565777A (en) * 1993-09-13 1996-10-15 Kabushiki Kaisha Toshiba Method/apparatus for NMR imaging using an imaging scheme sensitive to inhomogeneity and a scheme insensitive to inhomogeneity in a single imaging step
US20070167717A1 (en) * 2005-11-27 2007-07-19 James Timothy W Structure assessment using spatial-frequency analysis
US20120262171A1 (en) * 2011-04-15 2012-10-18 Hans Weber Method of magnetic resonance imaging for the selection and recording of curved slices
US20160274203A1 (en) 2014-09-01 2016-09-22 bioProtonics, LLC Method for assessing structural spatial frequencies using hybrid sampling with non-zero gradient for enhancement of selective sampling

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3359034A4

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2020520786A (ja) * 2017-05-24 2020-07-16 バイオプロトニクス インコーポレイテッドBioprotonics,Inc. 特定のコントラスト機構を用いて構造的空間周波数を評価するための選択的サンプリング
JP7273803B2 (ja) 2017-05-24 2023-05-15 バイオプロトニクス インコーポレイテッド 特定のコントラスト機構を用いて構造的空間周波数を評価するための選択的サンプリング
WO2019010381A1 (fr) * 2017-07-06 2019-01-10 bioProtonics, LLC Procédé pour mesurer une texture de tissu en utilisant la spectroscopie rmn pour identifier des espèces chimiques d'éléments de texture de composant dans une région ciblée de tissu
CN116112139B (zh) * 2018-09-21 2024-01-30 中兴通讯股份有限公司 波束失败恢复方法及装置
CN116112139A (zh) * 2018-09-21 2023-05-12 中兴通讯股份有限公司 波束失败恢复方法及装置
US11836925B2 (en) 2018-12-17 2023-12-05 Shanghai United Imaging Intelligence Co., Ltd. Systems and methods for image segmentation
US11341734B2 (en) 2018-12-17 2022-05-24 Shanghai United Imaging Intelligence Co., Ltd. Systems and methods for image segmentation
WO2020136571A1 (fr) * 2018-12-26 2020-07-02 Analytics For Life Inc. Procédés et systèmes de configuration et d'utilisation de réseaux neuronaux dans la caractérisation de systèmes physiologiques
US11589829B2 (en) 2018-12-26 2023-02-28 Analytics For Life Inc. Methods and systems to configure and use neural networks in characterizing physiological systems
US11989652B2 (en) 2018-12-26 2024-05-21 Analytics For Life Inc. Methods and systems to configure and use neural networks in characterizing physiological systems
CN112184845A (zh) * 2020-09-02 2021-01-05 东软医疗系统股份有限公司 弥散加权图像重建模型的生成方法及装置
CN112184845B (zh) * 2020-09-02 2023-05-05 东软医疗系统股份有限公司 弥散加权图像重建模型的生成方法及装置
WO2023055471A1 (fr) * 2021-09-29 2023-04-06 The Johns Hopkins University Identification automatisée de grappes de coefficients de diffusion apparents à hyperintensité pour une thérapie thermique interstitielle laser guidée par irm

Also Published As

Publication number Publication date
CA3000765C (fr) 2024-01-23
AU2016334250A1 (en) 2018-05-17
EP3359034A1 (fr) 2018-08-15
JP6906507B2 (ja) 2021-07-21
CN108366753A (zh) 2018-08-03
JP2018529474A (ja) 2018-10-11
EP3359034A4 (fr) 2019-08-07
CN108366753B (zh) 2019-06-07
AU2016334250B2 (en) 2020-04-02
CA3000765A1 (fr) 2017-04-13

Similar Documents

Publication Publication Date Title
US11175363B2 (en) Selective sampling for assessing structural spatial frequencies with specific contrast mechanisms
AU2016334250B2 (en) Selective sampling for assessing structural spatial frequencies with specific contrast mechanisms
US9664760B2 (en) Selective sampling for assessing structural spatial frequencies with specific contrast mechanisms
AU2021201998B2 (en) Selective sampling magnetic resonance-based method for assessing structural spatial frequencies
US9664759B2 (en) Method for assessing structural spatial frequencies using hybrid sampling with non-zero gradient for enhancement of selective sampling
JP7273803B2 (ja) 特定のコントラスト機構を用いて構造的空間周波数を評価するための選択的サンプリング
US11353533B2 (en) Methods and devices for contrast agent magnetic resonance imaging
Horie et al. Magnetic resonance imaging technique for visualization of irregular cerebrospinal fluid motion in the ventricular system and subarachnoid space
Andronesi et al. Whole brain mapping of water pools and molecular dynamics with rotating frame MR relaxation using gradient modulated low-power adiabatic pulses
Martinez Introduction to MRI physics
AU2018297168A1 (en) A method to measure tissue texture using NMR spectroscopy to identify chemical species of component textural elements in a targeted region of tissue
Mendez et al. Diffusion breast MRI: current standard and emerging techniques
Nunes et al. Diffusion mri outside the brain
Bihan-Poudec et al. Cardiovascular effects on high-resolution 3D multi-shot diffusion MRI of the rhesus macaque brain
Habib Development & optimization of diffusion tensor imaging at high field strengths in translational research
Marx Improved MRI Thermometry for MR-Guided Focused Ultrasound in the Brain
Cohen In-Vivo Three dimensional Proton Hadamard Spectroscopic Imaging in the Human Brain

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16854502

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 3000765

Country of ref document: CA

WWE Wipo information: entry into national phase

Ref document number: 2018517540

Country of ref document: JP

NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 2016854502

Country of ref document: EP

ENP Entry into the national phase

Ref document number: 2016334250

Country of ref document: AU

Date of ref document: 20161007

Kind code of ref document: A