During operation, a system may apply an external magnetic field and an RF pulse sequence to a sample. Then, the system may measure at least a component of a magnetization associated with the sample, such as MR signals of one or more types of nuclei in the sample. Moreover, the system may calculate a
During operation, a system may apply an external magnetic field and an RF pulse sequence to a sample. Then, the system may measure at least a component of a magnetization associated with the sample, such as MR signals of one or more types of nuclei in the sample. Moreover, the system may calculate at least a predicted component of the magnetization for voxels associated with the sample based on the measured component of the magnetization, a forward model, the external magnetic field and the RF pulse sequence. Next, the system may solve an inverse problem by iteratively modifying the parameters associated with the voxels in the forward model until a difference between the predicted component of the magnetization and the measured component of the magnetization is less than a predefined value. Note that the calculations may be performed concurrently with the measurements and may not involve performing a Fourier transform.
대표청구항▼
1. A method for determining model parameters associated with a sample, comprising operations of: applying, to the sample, an external magnetic field using a magnet and a radio-frequency pulse sequence using a transmission coil, wherein the external magnetic field is associated with polarization of t
1. A method for determining model parameters associated with a sample, comprising operations of: applying, to the sample, an external magnetic field using a magnet and a radio-frequency pulse sequence using a transmission coil, wherein the external magnetic field is associated with polarization of the sample;measuring, using a radio-frequency coil, a non-inductive sensor or both, at least a component of a magnetization response associated with the sample;calculating, with a computer that initiates a forward model with multiple voxels that represent the sample, at least a predicted value of the component of the magnetization response that was measured in the measuring operation, wherein the forward model simulates magnetic-resonance response physics occurring within the sample to a given external magnetic field and a given radio-frequency pulse sequence that are selected from a range of measurement conditions that includes the applied external magnetic field and the applied radio-frequency pulse sequence, and at least a different external magnetic field and at least a different radio-frequency pulse sequence,wherein the forward model is a function of the applied external magnetic field, the applied radio-frequency pulse sequence, the model parameters of the multiple voxels, and one of: Bloch equations, or Liouvillian computations,wherein the model parameters comprise a density of a type of nuclei, a longitudinal relaxation time along a direction parallel to the applied external magnetic field and a transverse relaxation time along a direction perpendicular to the applied external magnetic field, andwherein the model parameters are determined without performing a Fourier transform on the measured component of the magnetization response;computing, with the computer, a difference between at least the measured component of the magnetization response and at least the calculated predicted value of the component of the magnetization;solving, with the computer, an inverse problem as a function of the computed difference in order to determine revisions to the model parameters on a voxel-by-voxel basis;iteratively repeating instances of a group of operations comprising the operations of applying, measuring, calculating, computing and solving until the computed difference is less than a predefined value, wherein the magnetization of the sample either has not relaxed to a predefined state or is not reset to the predefined state between the instances of the group of operations, so that the sample has a dynamic magnetization state between at least some adjacent instances of the group of operations; andproviding the model parameters as an output to a user, another electronic device, a display or memory when the computed difference is less than the predefined value. 2. The method of claim 1, wherein the forward model further comprises an error term associated with the dynamic magnetization of the sample. 3. The method of claim 1, wherein the revisions to the model parameters are based on a Jacobian matrix and use Newton's method. 4. The method of claim 1, wherein the revisions to the model parameters are constrained based on the model parameters that have been determined during at least a previous instance of the group of operations. 5. The method of claim 1, wherein the sample comprises different types of tissue; and wherein initial values of the model parameters are within predefined parameter ranges for the different types of tissue that are present in the sample. 6. The method of claim 1, wherein the sample comprises different types of tissue; and wherein the method further comprises segmenting the tissue types that are present in the sample based on discontinuous changes in at least some of the model parameters along a direction between adjacent voxels in the representation of the sample. 7. The method of claim 1, wherein the model parameters are determined sequentially in the instances of the group of operations based on time scales associated with the model parameters; and wherein a model parameter that has a shortest time scale is determined first. 8. The method of claim 1, wherein the calculation of the component of the magnetization response and the revisions to the model parameters are performed concurrently with the measurement of the component of the magnetization response during a given instance of the group of operations. 9. The method of claim 1, wherein at least one of the applied external magnetic field and the applied radio-frequency pulse sequence is modified between at least two adjacent instances of the group of operations. 10. The method of claim 9, wherein the method further comprises determining the current dynamic magnetization state of the sample using the forward model; wherein the forward model is a function of the applied external magnetic field and the applied radio-frequency pulse sequence; andwherein the determined current dynamic magnetization state is used as an initial condition in the calculating and the determining operations that occur in a subsequent instance of the group of operations. 11. The method of claim 1, wherein at least one of a magnitude and a direction of the applied external magnetic field is changed as a function of time during acquisition of the measurement. 12. The method of claim 1, wherein at least one of the applied external magnetic field and the applied radio-frequency pulse sequence is modified between at least two adjacent instances of the group of operations; and wherein at least the one of the modified applied external magnetic field and the modified applied radio-frequency pulse sequence are selected to reduce the computed difference. 13. A non-transitory computer-readable storage medium for use in conjunction with a computer, the computer-readable storage medium configured to store program instructions that, when executed by the computer, cause the computer to perform operations comprising: applying, to a sample, an external magnetic field using a magnet and a radio-frequency pulse sequence using a transmission coil, wherein the external magnetic field is associated with polarization of the sample;measuring, using a radio-frequency coil, a non-inductive sensor or both, at least a component of a magnetization response associated with the sample;calculating, with the computer that initiates a forward model with multiple voxels that represent the sample, at least a predicted value of the component of the magnetization response that was measured in the measuring operation, wherein the forward model simulates magnetic-resonance response physics occurring within the sample to a given external magnetic field and a given radio-frequency pulse sequence that are selected from a range of measurement conditions that includes the applied external magnetic field and the applied radio-frequency pulse sequence, and at least a different external magnetic field and at least a different radio-frequency pulse sequence,wherein the forward model is a function of the applied external magnetic field, the applied radio-frequency pulse sequence, model parameters of the multiple voxels, and one of: Bloch equations, or Liouvillian computations,wherein the model parameters comprise a density of a type of nuclei, a longitudinal relaxation time along a direction parallel to the applied external magnetic field and a transverse relaxation time along a direction perpendicular to the applied external magnetic field, andwherein the model parameters are determined without performing a Fourier transform on the measured component of the magnetization response;computing, with the computer, a difference between at least the measured component of the magnetization response and at least the calculated predicted value of the component of the magnetization;solving, with the computer, an inverse problem as a function of the computed difference in order to determine revisions to the model parameters on a voxel-by-voxel basis;iteratively repeating instances of a group of operations comprising the operations of applying, measuring, calculating, computing and solving until the computed difference is less than a predefined value, wherein the magnetization of the sample either has not relaxed to a predefined state or is not reset to the predefined state between the instances of the group of operations, so that the sample has a dynamic magnetization state between at least some adjacent instances of the group of operations; andproviding the model parameters as an output to a user, another electronic device, a display or memory when the computed difference is less than the predefined value. 14. The non-transitory computer-readable storage medium of claim 13, wherein the forward model further comprises an error term associated with the dynamic magnetization of the sample. 15. The non-transitory computer-readable storage medium of claim 13, wherein at least one of the applied external magnetic field and the applied radio-frequency pulse sequence is modified between at least two adjacent instances of the group of operations; and wherein at least the one of the modified applied external magnetic field and the modified applied radio-frequency pulse sequence are selected to reduce the computed difference. 16. The non-transitory computer-readable storage medium of claim 13, wherein the calculation of the component of the magnetization response and the revisions to the model parameters are performed concurrently with the measurement of the component of the magnetization response during a given instance of the group of operations. 17. The non-transitory computer-readable storage medium of claim 13, wherein at least one of the applied external magnetic field and the applied radio-frequency pulse sequence is modified between at least two adjacent instances of the group of operations. 18. The non-transitory computer-readable storage medium of claim 13, wherein at least one of a magnitude and a direction of the applied external magnetic field is changed as a function of time during acquisition of the measurement. 19. A system configured to determine model parameters, comprising: a magnet and a transmission coil configured to generate magnetic fields;a measurement device configured to perform measurements, wherein the measurement device comprises: a radio-frequency coil, a non-inductive sensor or both;a processor, coupled to the magnet, the measurement device and memory, configured to execute program instructions; andthe memory, coupled to the processor, configured to store the program instructions that, when executed by the processor, cause the system to perform operations comprising: applying, to a sample, an external magnetic field using the magnet and a radio-frequency pulse sequence using the transmission coil, wherein the external magnetic field is associated with polarization of the sample;measuring, using a measurement device, at least a component of a magnetization response associated with the sample;calculating, with the processor that initiates a forward model with multiple voxels that represent the sample, at least a predicted value of the component of the magnetization response that was measured in the measuring operation, wherein the forward model simulates magnetic-resonance response physics occurring within the sample to a given external magnetic field and a given radio-frequency pulse sequence that are selected from a range of measurement conditions that includes the applied external magnetic field and the applied radio-frequency pulse sequence, and at least a different external magnetic field and at least a different radio-frequency pulse sequence,wherein the forward model is a function of the applied external magnetic field, the applied radio-frequency pulse sequence, model parameters of the multiple voxels, and one of: Bloch equations, or Liouvillian computations,wherein the model parameters comprise a density of a type of nuclei, a longitudinal relaxation time along a direction parallel to the applied external magnetic field and a transverse relaxation time along a direction perpendicular to the applied external magnetic field, andwherein the model parameters are determined without performing a Fourier transform on the measured component of the magnetization response;computing, with the processor, a difference between at least the measured component of the magnetization response and at least the calculated predicted value of the component of the magnetization;solving, with the processor, an inverse problem as a function of the computed difference in order to determine revisions to the model parameters on a voxel-by-voxel basis;iteratively repeating instances of a group of operations comprising the operations of applying, measuring, calculating, computing and solving until the computed difference is less than a predefined value, wherein the magnetization of the sample either has not relaxed to a predefined state or is not reset to the predefined state between the instances of the group of operations, so that the sample has a dynamic magnetization state between at least some adjacent instances of the group of operations; andproviding the model parameters as an output to a user, another electronic device, a display or the memory when the computed difference is less than the predefined value. 20. The system of claim 19, wherein the operations comprise determining the current dynamic magnetization state of the sample using the forward model; wherein the forward model is a function of the applied external magnetic field and the applied radio-frequency pulse sequence; andwherein the determined current dynamic magnetization state is used as an initial condition in the calculating and the determining operations that occur in a subsequent instance of the group of operations. 21. The system of claim 19, wherein at least one of the applied external magnetic field and the applied radio-frequency pulse sequence is modified between at least two adjacent instances of the group of operations; and wherein at least the one of the modified applied external magnetic field and the modified applied radio-frequency pulse sequence are selected to reduce the computed difference. 22. The system of claim 19, wherein the model parameters are determined sequentially in the instances of the group of operations based on time scales associated with the model parameters; and wherein a model parameter that has a shortest time scale is determined first. 23. The system of claim 19, wherein the forward model further comprises an error term associated with the dynamic magnetization of the sample.
연구과제 타임라인
LOADING...
LOADING...
LOADING...
LOADING...
LOADING...
이 특허에 인용된 특허 (21)
Boernert, Peter; Koken, Peter; Nehrke, Kay; Katscher, Ulrich; Graesslin, Ingmar, B1-mapping and B1L-shimming for MRI.
Reddy, Ravinder; Hariharan, Hari; Haris, Mohammad; Cai, Kejia; Singh, Anup; Kogan, Feliks, CEST MRI methods for imaging of metabolites and the use of same as biomarkers.
Gjesdal, Kjell-Inge; Grøvik, Endre, Dynamic MR imaging of patients with breast cancer—establishment and comparison of different analytical methods for tissue perfusion and capillary permeability.
Pla Frederic Ghislain ; Hedeen Robert Arvin ; Dobberstein Robert James ; Ebben Thomas Gerard ; Mansell Scott Thomas ; Obasih Kemakolam Michael ; Radziun Michael James ; Sue Peter Ping-Liang ; Edelste, Low noise MRI scanner.
Assmann, Stefan; Ekinci, Okan, Magnetic resonance data acquisition system and method with recursively adapted object-specific measurement parameter adjustment during patient movement through the MRI system.
Chen Songhua ; Georgi Daniel T., Methods for acquisition and processing of nuclear magnetic resonance signals for determining fluid properties in petroleum reservoirs having more than one fluid phase.
Bergstrom Paul M. ; Daly Thomas P. ; Moses Edward I. ; Patterson ; Jr. Ralph W. ; Schach von Wittenau Alexis E. ; Garrett Dewey N. ; House Ronald K. ; Hartmann-Siantar Christine L. ; Cox Lawrence J. , System and method for radiation dose calculation within sub-volumes of a monte carlo based particle transport grid.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.