DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claim 15 is rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter. The claims does not fall within at least one of the four categories of patent eligible subject matter because claim 15 is directed to a “computer-readable storage medium”. The specification discloses “there is disclosed a computer program stored in a computer-readable storage medium” (Paragraph [0022]). In the state of the art, transitory signals are commonplace as a medium for storing computer instructions and thus, in the absence of any evidence to the contrary, and given its broadest reasonable interpretation in light of the specification, the scope of the claimed "computer-readable storage medium", covers both non-transitory media and a transitory media (a signal per se). A transitory signal does not fall within the definition of a process, machine, manufacture, or composition of matter. A claim drawn to such a computer-readable storage medium that covers both transitory and non-transitory embodiments may be amended to narrow the claim to cover only statutory embodiments by adding the limitation "non-transitory" to the claim.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-15 are rejected under 35 U.S.C. 103 as being unpatentable over Francavilla et al. (US Pub. 2021/0311151), hereinafter Francavilla, in view of Qi et al. (US Pub. 2021/0290191), hereinafter Qi.
Regarding claim 1, Francavilla discloses a data processing method, the data processing method being performed by a computing device including at least one processor (Fig. 1; Paragraph [0008]: computer system includes: an interface circuit that communicates with an electronic device (such as a measurement device that performs measurements), a processor that executes program instructions, and memory that stores the program instructions), the data processing method comprising: grouping a plurality of coil images, generated based on signals received from a plurality of coils, into a plurality of groups (Paragraph [0126]: first class of approaches (which is referred to as ‘SENSE’, ‘ASSET’, ‘RAPID’ or ‘SPEEDER’) is image domain based after reconstruction of MR signals from individual RF pickup coils or antennas in receive channels (which are sometimes referred to as ‘coils’). In this approach, the number of dropped or skipped MR scan lines may equal the number of receive channels. However, a separate pre-scan is used to determine the coil sensitivities (or coil sensitivity maps) of the receive channels. This is because the measured MR signal using a given receive channel during an MR scan corresponds to a volume integral of the product of a coil sensitivity for the given receiver channel and the time-dependent magnetization of the sample. Moreover, because the polarized magnetic field received by a coil or antenna in the given receive channel depends on its position and orientation, in general each of the coils or antennas in the receive channels has a different coil sensitivity. By performing a pre-scan, the coils sensitivities can be predetermined. Then, in the image domain, sample properties (such as the spatially varying proton density) can be illustrated or presented); and for each of the plurality of groups, generating a group image by combining coil images included in each group (Fig. 11; Paragraphs [0185]-[0186]: FIG. 11 presents a drawing illustrating an example of MRI image reconstruction in accordance with an embodiment of the present disclosure. Notably, MRI may sample the Fourier transform (k-space) of the coil-weighted image. In multi-coil systems, multiple images may be acquired (e.g., at the same time. These coil-weighted images may be combined to form a single image…shown in FIG. 11, an inverse Fourier transform (ℑ−1) may convert raw scanner data into coil-weighted images, and a Fourier transform (ℑ) may convert the coil-weighted images into raw scanner data. Moreover, during the MRI image reconstruction, the coil-weighted images may be separated into MRI images and associated coil sensitivities. Note that for each voxel, the solution to the nonlinear optimization may provide the MR information (such as the one or more MR parameters) and the coefficients used in the coil sensitivities).
Francavilla does not explicitly disclose wherein a plurality of group images are noise-independent of each other.
However, Qi teaches MRI image reconstruction from coils (Paragraph [0005]: acquiring a large number of HdCT images is challenging due to the risk of radiation. In 2018, Lehtinen et al introduced a noise-to-noise (Noise2Noise) model to train DNNs without using any clean images; Paragraph [0093]), further comprising: wherein a plurality of group images are noise-independent of each other (Paragraph [0093]: MRI scanner 700 acquires k-space data (or scan data) of an object to be imaged, by the RF receiving coil, and the k-space data is used by the MRI data processor to reconstruct MR images. The k-space data can be used as the noisy data 115′ or 120′, or the scan data 145. The reconstructed MR images can be used as the noisy data 115 or 120). Qi teaches that this will allow for training an AI model without need for noiseless images (Paragraph [0005]). Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Francavilla with the features of above as taught by Qi so as to allow for training an AI model without need for noiseless images as presented by Qi.
Regarding claim 2, Francavilla, in view of Qi teaches the data processing method of claim 1, Francavilla discloses wherein: the plurality of groups comprise a first group and a second group (Fig. 11; Paragraphs [0185]-[0186]: FIG. 11 presents a drawing illustrating an example of MRI image reconstruction in accordance with an embodiment of the present disclosure. Notably, MRI may sample the Fourier transform (k-space) of the coil-weighted image. In multi-coil systems, multiple images may be acquired (e.g., at the same time. These coil-weighted images may be combined to form a single image…shown in FIG. 11, an inverse Fourier transform (ℑ−1) may convert raw scanner data into coil-weighted images, and a Fourier transform (ℑ) may convert the coil-weighted images into raw scanner data. Moreover, during the MRI image reconstruction, the coil-weighted images may be separated into MRI images and associated coil sensitivities. Note that for each voxel, the solution to the nonlinear optimization may provide the MR information (such as the one or more MR parameters) and the coefficients used in the coil sensitivities); and grouping the plurality of coil images into the plurality of groups comprises grouping the plurality of coil images so that sensitivity characteristics of first group images of the first group correspond to sensitivity characteristics of second group images of the second group (Fig. 9; Paragraph [0126]: first class of approaches (which is referred to as ‘SENSE’, ‘ASSET’, ‘RAPID’ or ‘SPEEDER’) is image domain based after reconstruction of MR signals from individual RF pickup coils or antennas in receive channels (which are sometimes referred to as ‘coils’). In this approach, the number of dropped or skipped MR scan lines may equal the number of receive channels. However, a separate pre-scan is used to determine the coil sensitivities (or coil sensitivity maps) of the receive channels. This is because the measured MR signal using a given receive channel during an MR scan corresponds to a volume integral of the product of a coil sensitivity for the given receiver channel and the time-dependent magnetization of the sample. Moreover, because the polarized magnetic field received by a coil or antenna in the given receive channel depends on its position and orientation, in general each of the coils or antennas in the receive channels has a different coil sensitivity; Paragraph [0164]: the computer system may access (e.g., in memory) a predetermined set of coil magnetic field basis vectors (operation 912) associated with a surface (such as a closed surface, e.g., a cylindrical surface, a deformed sphere and, more generally, an arbitrary 2D manifold) surrounding the sample, where weighted superpositions of the predetermined set of coil magnetic field basis vectors may represent the coil sensitivities of coils in the MR apparatus. For example, a given coil sensitivity may be represented by a linear superposition of products of the coefficients and predetermined coil magnetic field basis vectors in the predetermined set of coil magnetic field basis vectors; Paragraph [0186]: shown in FIG. 11, an inverse Fourier transform (ℑ−1) may convert raw scanner data into coil-weighted images, and a Fourier transform (ℑ) may convert the coil-weighted images into raw scanner data. Moreover, during the MRI image reconstruction, the coil-weighted images may be separated into MRI images and associated coil sensitivities. Note that for each voxel, the solution to the nonlinear optimization may provide the MR information (such as the one or more MR parameters) and the coefficients used in the coil sensitivities).
Regarding claim 3, Francavilla, in view of Qi teaches the data processing method of claim 2, Francavilla discloses wherein grouping the plurality of coil images into the plurality of groups comprises grouping the plurality of coil images so that a standard deviation of a sensitivity map corresponding to the first group images and a standard deviation of a sensitivity map corresponding to the second group images are each minimized (Paragraphs [0126]-[0127]: there are two principal classes of existing MRI parallel imaging techniques. A first class of approaches (which is referred to as ‘SENSE’, ‘ASSET’, ‘RAPID’ or ‘SPEEDER’) is image domain based after reconstruction of MR signals from individual RF pickup coils or antennas in receive channels (which are sometimes referred to as ‘coils’). In this approach, the number of dropped or skipped MR scan lines may equal the number of receive channels. However, a separate pre-scan is used to determine the coil sensitivities (or coil sensitivity maps) of the receive channels. This is because the measured MR signal using a given receive channel during an MR scan corresponds to a volume integral of the product of a coil sensitivity for the given receiver channel and the time-dependent magnetization of the sample. Moreover, because the polarized magnetic field received by a coil or antenna in the given receive channel depends on its position and orientation, in general each of the coils or antennas in the receive channels has a different coil sensitivity. By performing a pre-scan, the coils sensitivities can be predetermined. Then, in the image domain, sample properties (such as the spatially varying proton density) can be illustrated or presented…in existing MRI scanners, the first class of approaches may involve the operations of: generating coil sensitivity maps, acquire partial k-space MR data, reconstruct partial field-of-view images from each coil, and unfold/combine partial field-of-view images using matrix inversion. Note, therefore, that the first class of approaches is recast as a linear problem, and which may, in part, be solved using a Fourier transform and an inverse Fourier transform; Paragraph [0161]: the estimated coil sensitivities, which were determined previously in the Maxwell parallel imaging technique, may be used to cast the original nonlinear problem into a linear one. This linear problem may still be ill-posed, because of the under-sampling of the k-space. Then, a final estimate of the WPD image may be obtained as the solution of a constrained convex optimization problem. Notably, the improved estimate of the WPD image may correspond to a minimizer of the total variation or the structure total variation subject to multiple constraints, whose number may equal to the number of MR coil measurements. Each of the constraints may enforce that the norm of the difference of the coil measurement and the corresponding observation or estimation model, which involves the solution, is less than or equal to a quantity proportional to the standard deviation of the noise effecting the specific coil measurements).
Regarding claim 4, Francavilla, in view of Qi teaches the data processing method of claim 2, Francavilla discloses wherein grouping the plurality of coil images into the plurality of groups comprises grouping the plurality of coil images so that a difference between a sensitivity map corresponding to the first group images and a sensitivity map corresponding to the second group images is minimized (Paragraph [0137]: because of the low frequency (the precession frequency for a proton in an external magnetic field of 1.5 T is 63.87 MHz) and the near-field condition, the currents on the surface may be similar to each other. Consequently, there may be a set of coil magnetic field basis vectors that encompasses or includes the majority of the energy or power in the different coil magnetic fields. For example, a singular value decomposition or an eigenvalue-decomposition technique may be used on the different numerically simulated coil magnetic fields to determine the set of coil magnetic field basis vectors. Then, a given coil magnetic field (and, thus, a given coil sensitivity) may be a linear superposition of the set of coil magnetic field basis vectors; Paragraph [0185]: the mapping from equivalent sources to coil sensitivities may admit or allow fast factorizations, which may reduce the computational cost to store the mapping in memory and to evaluate the mapping from an order of O(M.Math.N) to an order of O(M+N), where M and N are the number of voxels and of mesh cells, respectively. For example, M may be between 1-100 M and N may be between 100-4,000, and the number of coils may be between 8-128 or 8-4,096. Note that the mapping from physical to virtual coils may be one-to-one (i.e., the number of coefficients may equal the number of coils). In some embodiments, applicable methods or technique to reduce the computational cost of the mapping may include: fast multipole methods (FMM), a multi-level fast multipole approach (MLFMA), hierarchical (H) and nested hierarchical (H2) matrix decompositions, hierarchical semi-separable (HSS) matrix representations, hierarchical off-diagonal low rank (HODLR) representations, interpolative decompositions (ID), an adaptive cross approximation (ACA), a multi-level matrix decomposition algorithm (MLMDA), and/or fast Fourier transform (FFT)-based approaches; Paragraph [0192]: the disclosed computation techniques may provide a better representation of the coil sensitivities at different locations (or voxels) in a sample, better regularization, faster computations (e.g., 10x faster), reduced memory usage, and/or faster factorization with reduced or low rank. These capabilities may accelerate MR scans because, unlike existing techniques, the coil sensitivities may not need to be determined prior to every scan, and the coil sensitivities may not be assumed to be smooth. Instead, as described previously, the coil sensitivities may be determined at the same time as the image reconstruction, and may obey Maxwell's equations. The calculated coil sensitivities may be more accurate than the measured MR signals (because of a finite signal-to-noise ratio), which may make the forward model more accurate so that the inverse problem can be solved more accurately).
Regarding claim 5, Francavilla, in view of Qi teaches the data processing method of claim 1, Francavilla discloses wherein generating the group image by combining the coil images included in each group comprises combining the coil images included in each group so that sensitivity characteristics of individual group images correspond to each other (Fig. 9; Paragraph [0126]: first class of approaches (which is referred to as ‘SENSE’, ‘ASSET’, ‘RAPID’ or ‘SPEEDER’) is image domain based after reconstruction of MR signals from individual RF pickup coils or antennas in receive channels (which are sometimes referred to as ‘coils’). In this approach, the number of dropped or skipped MR scan lines may equal the number of receive channels. However, a separate pre-scan is used to determine the coil sensitivities (or coil sensitivity maps) of the receive channels. This is because the measured MR signal using a given receive channel during an MR scan corresponds to a volume integral of the product of a coil sensitivity for the given receiver channel and the time-dependent magnetization of the sample. Moreover, because the polarized magnetic field received by a coil or antenna in the given receive channel depends on its position and orientation, in general each of the coils or antennas in the receive channels has a different coil sensitivity; Paragraph [0164]: the computer system may access (e.g., in memory) a predetermined set of coil magnetic field basis vectors (operation 912) associated with a surface (such as a closed surface, e.g., a cylindrical surface, a deformed sphere and, more generally, an arbitrary 2D manifold) surrounding the sample, where weighted superpositions of the predetermined set of coil magnetic field basis vectors may represent the coil sensitivities of coils in the MR apparatus. For example, a given coil sensitivity may be represented by a linear superposition of products of the coefficients and predetermined coil magnetic field basis vectors in the predetermined set of coil magnetic field basis vectors; Paragraph [0186]: shown in FIG. 11, an inverse Fourier transform (ℑ−1) may convert raw scanner data into coil-weighted images, and a Fourier transform (ℑ) may convert the coil-weighted images into raw scanner data. Moreover, during the MRI image reconstruction, the coil-weighted images may be separated into MRI images and associated coil sensitivities. Note that for each voxel, the solution to the nonlinear optimization may provide the MR information (such as the one or more MR parameters) and the coefficients used in the coil sensitivities).
Regarding claim 6, Francavilla, in view of Qi teaches the data processing method of claim 5, Francavilla discloses wherein generating the group image by combining the coil images included in each group comprises generating the group image by adaptively combining the coil images included in each group by using a sensitivity adjustment coefficient determined based on sensitivity of coils corresponding to each group (Paragraph [0032]: a computer system that determines coefficients in a representation of coil sensitivities and MR information associated with a sample is described. During operation, the computer system may acquire MR signals associated with a sample from a measurement device or memory. Then, the computer system may access a predetermined set of coil magnetic field basis vectors associated with a surface (such as a closed surface) surrounding the sample, where the coil sensitivities of coils in the measurement device are represented by weighted superpositions of the predetermined set of coil magnetic field basis vectors using the coefficients, and where the predetermined coil magnetic field basis vectors are solutions to Maxwell's equations. Next, the computer system may solve, on a voxel-by-voxel basis for voxels associated with the sample, a nonlinear optimization problem for the MR information associated with the sample and the coefficients using: a forward model that uses the MR information as inputs and simulates response physics of the sample, the MR signals and the predetermined set of coil magnetic field basis vectors, where the MR information includes quantitative values of one or more MR parameters in the voxels associated with the sample; Paragraph [0186]: shown in FIG. 11, an inverse Fourier transform (ℑ−1) may convert raw scanner data into coil-weighted images, and a Fourier transform (ℑ) may convert the coil-weighted images into raw scanner data. Moreover, during the MRI image reconstruction, the coil-weighted images may be separated into MRI images and associated coil sensitivities. Note that for each voxel, the solution to the nonlinear optimization may provide the MR information (such as the one or more MR parameters) and the coefficients used in the coil sensitivities).
Regarding claim 7, Francavilla, in view of Qi teaches the data processing method of claim 6, Francavilla discloses wherein: the plurality of groups comprise a first group and a second group (Fig. 11; Paragraphs [0185]-[0186]: FIG. 11 presents a drawing illustrating an example of MRI image reconstruction in accordance with an embodiment of the present disclosure. Notably, MRI may sample the Fourier transform (k-space) of the coil-weighted image. In multi-coil systems, multiple images may be acquired (e.g., at the same time. These coil-weighted images may be combined to form a single image…shown in FIG. 11, an inverse Fourier transform (ℑ−1) may convert raw scanner data into coil-weighted images, and a Fourier transform (ℑ) may convert the coil-weighted images into raw scanner data. Moreover, during the MRI image reconstruction, the coil-weighted images may be separated into MRI images and associated coil sensitivities. Note that for each voxel, the solution to the nonlinear optimization may provide the MR information (such as the one or more MR parameters) and the coefficients used in the coil sensitivities); and generating the group image by combining the coil images included in each group comprises generating a first group image of the first group and a second group image of the second group based on a first sensitivity adjustment coefficient determined to be applied to pixels of first locations for coil images of the first group and a second sensitivity adjustment coefficient determined to be applied to pixels of first locations for coil images of the second group (Paragraph [0032]: a computer system that determines coefficients in a representation of coil sensitivities and MR information associated with a sample is described. During operation, the computer system may acquire MR signals associated with a sample from a measurement device or memory. Then, the computer system may access a predetermined set of coil magnetic field basis vectors associated with a surface (such as a closed surface) surrounding the sample, where the coil sensitivities of coils in the measurement device are represented by weighted superpositions of the predetermined set of coil magnetic field basis vectors using the coefficients, and where the predetermined coil magnetic field basis vectors are solutions to Maxwell's equations. Next, the computer system may solve, on a voxel-by-voxel basis for voxels associated with the sample, a nonlinear optimization problem for the MR information associated with the sample and the coefficients using: a forward model that uses the MR information as inputs and simulates response physics of the sample, the MR signals and the predetermined set of coil magnetic field basis vectors, where the MR information includes quantitative values of one or more MR parameters in the voxels associated with the sample; Paragraph [0094]: the first convolutional layer filters a 224×224×3 input image with 96 kernels of size 11×11×3 with a stride of four pixels (this is the distance between the receptive field centers of neighboring neurons in a kernel map). Note that the second convolutional layer may take as input the (response-normalized and pooled) output of the first convolutional layer and may filter it with 256 kernels of size 5×5×48. Furthermore, the third, fourth, and fifth convolutional layers may be coupled to one another without any intervening pooling or normalization layers; Paragraph [0186]: shown in FIG. 11, an inverse Fourier transform (ℑ−1) may convert raw scanner data into coil-weighted images, and a Fourier transform (ℑ) may convert the coil-weighted images into raw scanner data. Moreover, during the MRI image reconstruction, the coil-weighted images may be separated into MRI images and associated coil sensitivities. Note that for each voxel, the solution to the nonlinear optimization may provide the MR information (such as the one or more MR parameters) and the coefficients used in the coil sensitivities).
Regarding claim 8, Francavilla, in view of Qi teaches the data processing method of claim 7, Francavilla discloses wherein, with respect to the pixels of the first locations, when sensitivity of coils corresponding to the first group is higher than sensitivity of coils corresponding to the second group, the first sensitivity adjustment coefficient is smaller than the second sensitivity adjustment coefficient (Paragraph [0126]: in existing MRI scanners, the first class of approaches may involve the operations of: generating coil sensitivity maps, acquire partial k-space MR data, reconstruct partial field-of-view images from each coil, and unfold/combine partial field-of-view images using matrix inversion. Note, therefore, that the first class of approaches is recast as a linear problem, and which may, in part, be solved using a Fourier transform and an inverse Fourier transform; Paragraph [0176]: the problem of regularizing coil sensitivities may be formulated as an inverse source reconstruction problem, whereby equivalent surface fields are to be computed on a closed surface. Such equivalent surface fields may be constrained to match magnetic fields, e.g., coil sensitivities, which are: solutions to Maxwell's equations; and compatible with the data (such as MR signals) acquired from an MR apparatus).
Regarding claim 9, Francavilla, in view of Qi teaches the data processing method of claim 1, Francavilla discloses wherein the plurality of coil images are generated based on signals obtained by imaging an object once (Fig. 1; Paragraphs [0046]-[0047]: measurement device 114 may include one or more RF pickup coils or another magnetic sensor (such as a magnetometer, a superconducting quantum interference device, opto-electronics, etc.) that measure time-varying or time-domain electrical signals corresponding to the dynamic behavior of nuclear spins in the one or more types of nuclei or at least an average component of the magnetization corresponding to the aggregate dynamic behavior of the nuclear spins (which is sometimes referred to as a ‘magnetic response’) of at least the portion of sample 112. For example, measurement device 114 may measure the transverse magnetization of at least a portion of sample 112 as it processes in the xy plane…the measurements provided by measurement device 114 may be other than or different from an image. For example, the measurements may be other than MRI results. For example, the measurements may include or may correspond to (such as one or more components of) a free-induction-decay of the nuclear spins in sample 112. Consequently, in some embodiments the measurements may not involve performing a Fourier transform on the measured electrical signals (and, thus, may not be performed in k-space and may not involve pattern matching in k-space, such as MR fingerprinting). However, in general, the measurements may be specified in the time domain and/or the frequency domain. Therefore, in some embodiments, a variety of signal processing).
Regarding claim 10, Francavilla, in view of Qi teaches the data processing method of claim 9, Francavilla discloses wherein the plurality of coil images are generated based on signals obtained by performing accelerated imaging on the object (Paragraph [0062]: conjunction with the generated dataset, one or more predictive models can be used to select regularization that accelerates the initial data acquisition and/or denoising. Moreover, the one or more predictive models can also be used to accelerate simulations or reconstruction using the forward model. For example, a predictive model can provide initial model parameters for use in the forward model, which may reduce the number of iterations required for the measurements and the simulations to converge on a solution that has an accuracy exceeding the predefined value. Thus, if the initial model parameters result in predicted response that are very different from the measurements, this can be feedback into the subsequent measurements and simulations to improve the model parameters and, thus, the predicted response; Paragraph [0122]: the gradient coils in an MR scanner phase encode (temporally) MR signals, which allows the output MR signals to be distinguished from each other. Moreover, when there are multiple receive channels, there is redundancy in the collected phase-encoded MR signals. In principle, by exploiting the different phase profiles, the redundancy allows some of the phase-encoded MR signals (such as some of the MR scan lines) to be skipped and subsequently reconstructed from the other phase-encoded MR signals, thereby accelerating the MR scan time).
Regarding claim 11, Francavilla, in view of Qi teaches the data processing method of claim 1, Francavilla discloses further comprising generating input and label images to be input to an artificial neural network model that outputs a high-quality medical image based on a low-quality medical image by combining the plurality of group images (Paragraph [0102]: for a type of tissue (such as a particular organ), the model parameters determined using different layers in a neural network may be iteratively refined as the size of the voxels is progressively decreased (and, thus, the number of voxels is increased) in the different layers. This analysis may be driven by the error between the measurements and simulated or predicted responses using the forward model. Progressing through successive layers in a neural network, the focus may be on the residual regions with errors that are larger than a convergence or an accuracy criterion. For example, the model parameters for the forward model in a layer in a neural network may be based on measurements at one magnetic-field strength and then the error may be determined based on the predicted response of the forward model at another magnetic-field strength. Furthermore, note that initially the predictive model or the forward model may assume that there is no contribution or interaction between different voxels. However, as the error and the voxel size are reduced, such contributions and/or interactions may be included in subsequent layers in a neural network. In some embodiments, when there are multiple candidate model-parameter solutions (having similar errors) to the inverse problem for a layer in a neural network, at least some of these candidates may be kept for use in a subsequent layer; Paragraph [0153]: the computation techniques address the problem of MRI reconstruction using multiple MR coils and under-sampled k-space measurements. By solving this problem, the computation techniques may significantly reduce the MR acquisition or scan time, but without compromising the quality of the restored or reconstructed image. This problem is known as ‘parallel imaging’ or MRI parallel imaging).
Regarding claim 12, the limitations of this claim substantially correspond to the limitations of claim 1 (except for the memory configured to store a plurality of coil images generated based on signals received from a plurality of coils, which is disclosed by Francavilla, Paragraph [0144]: a computer may acquire MR signals (operation 710) from or associated with a sample. This may involve having an MR apparatus applying an external magnetic field, a gradient magnetic field, and/or one or more RF pulse sequences, and measuring MR signals using receivers or receive channels. Alternatively or additionally, the computer may access MR signals stored in memory, which were previously acquired by an MR apparatus or measurement device); thus they are rejected on similar grounds.
Regarding claim 13, Francavilla, in view of Qi teaches the data processing device of claim 12, Francavilla discloses wherein the processor combines the coil images included in each group based on sensitivity characteristics of coils corresponding to each group and a sensitivity adjustment coefficient reflecting pixel locations of each coil image therein (Fig. 9; Paragraph [0126]: first class of approaches (which is referred to as ‘SENSE’, ‘ASSET’, ‘RAPID’ or ‘SPEEDER’) is image domain based after reconstruction of MR signals from individual RF pickup coils or antennas in receive channels (which are sometimes referred to as ‘coils’). In this approach, the number of dropped or skipped MR scan lines may equal the number of receive channels. However, a separate pre-scan is used to determine the coil sensitivities (or coil sensitivity maps) of the receive channels. This is because the measured MR signal using a given receive channel during an MR scan corresponds to a volume integral of the product of a coil sensitivity for the given receiver channel and the time-dependent magnetization of the sample. Moreover, because the polarized magnetic field received by a coil or antenna in the given receive channel depends on its position and orientation, in general each of the coils or antennas in the receive channels has a different coil sensitivity; Paragraph [0164]: the computer system may access (e.g., in memory) a predetermined set of coil magnetic field basis vectors (operation 912) associated with a surface (such as a closed surface, e.g., a cylindrical surface, a deformed sphere and, more generally, an arbitrary 2D manifold) surrounding the sample, where weighted superpositions of the predetermined set of coil magnetic field basis vectors may represent the coil sensitivities of coils in the MR apparatus. For example, a given coil sensitivity may be represented by a linear superposition of products of the coefficients and predetermined coil magnetic field basis vectors in the predetermined set of coil magnetic field basis vectors; Paragraph [0176]: the problem of regularizing coil sensitivities may be formulated as an inverse source reconstruction problem, whereby equivalent surface fields are to be computed on a closed surface. Such equivalent surface fields may be constrained to match magnetic fields, e.g., coil sensitivities, which are: solutions to Maxwell's equations; and compatible with the data (such as MR signals) acquired from an MR apparatus; Paragraph [0186]: shown in FIG. 11, an inverse Fourier transform (ℑ−1) may convert raw scanner data into coil-weighted images, and a Fourier transform (ℑ) may convert the coil-weighted images into raw scanner data. Moreover, during the MRI image reconstruction, the coil-weighted images may be separated into MRI images and associated coil sensitivities. Note that for each voxel, the solution to the nonlinear optimization may provide the MR information (such as the one or more MR parameters) and the coefficients used in the coil sensitivities).
Regarding claim 14, Francavilla, in view of Qi teaches the data processing device of claim 13, Francavilla discloses wherein the processor: groups the plurality of coil images into a first group and a second group (Fig. 11; Paragraphs [0185]-[0186]: FIG. 11 presents a drawing illustrating an example of MRI image reconstruction in accordance with an embodiment of the present disclosure. Notably, MRI may sample the Fourier transform (k-space) of the coil-weighted image. In multi-coil systems, multiple images may be acquired (e.g., at the same time. These coil-weighted images may be combined to form a single image…shown in FIG. 11, an inverse Fourier transform (ℑ−1) may convert raw scanner data into coil-weighted images, and a Fourier transform (ℑ) may convert the coil-weighted images into raw scanner data. Moreover, during the MRI image reconstruction, the coil-weighted images may be separated into MRI images and associated coil sensitivities. Note that for each voxel, the solution to the nonlinear optimization may provide the MR information (such as the one or more MR parameters) and the coefficients used in the coil sensitivities); and with respect to pixels of first locations for the individual coil images, when sensitivity of coils corresponding to the first group is higher than sensitivity of coils corresponding to the second group, combines the coil images by applying a sensitivity adjustment coefficient smaller than a sensitivity adjustment coefficient of coil images of the second group to coil images of the first group (Paragraph [0126]: in existing MRI scanners, the first class of approaches may involve the operations of: generating coil sensitivity maps, acquire partial k-space MR data, reconstruct partial field-of-view images from each coil, and unfold/combine partial field-of-view images using matrix inversion. Note, therefore, that the first class of approaches is recast as a linear problem, and which may, in part, be solved using a Fourier transform and an inverse Fourier transform; Paragraph [0176]: the problem of regularizing coil sensitivities may be formulated as an inverse source reconstruction problem, whereby equivalent surface fields are to be computed on a closed surface. Such equivalent surface fields may be constrained to match magnetic fields, e.g., coil sensitivities, which are: solutions to Maxwell's equations; and compatible with the data (such as MR signals) acquired from an MR apparatus).
Regarding claim 15, the limitations of this claim substantially correspond to the limitations of claim 1 (except for the a computer-readable storage medium, which is disclosed by Francavilla, Paragraph [0015]: a computer-readable storage medium for use with the computer system. This computer-readable storage medium includes program instructions that, when executed by the computer system, causes the computer system to perform at least some of the aforementioned operations); thus they are rejected on similar grounds.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MATTHEW D SALVUCCI whose telephone number is (571)270-5748. The examiner can normally be reached M-F: 7:30-4:00PT.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, XIAO WU can be reached at (571) 272-7761. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/MATTHEW SALVUCCI/Primary Examiner, Art Unit 2613