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 .
Status of Claims
Claims 1, 4-10, 12-14, 16, 17, and 19 have been amended.
Claims 3 and 11 have been cancelled. Claims 21 and 22 are newly added.
Claims 1, 2, 4-10, and 12-22 are pending.
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.
Claims 1, 2, 4-10, and 12-22 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception without significantly more. The claims recite:
[a] ascertaining an anatomical region of interest associated with a magnetic resonance imaging (MRI) exam of a patient;
[b] determining, based on further MRI imaging data, control data for identifying the anatomical region of interest in the MRI imaging data by (i) determining one or more registration parameters for registering the further MRI imaging data with the MRI imaging data, (ii) determining a contour of the anatomical region of interest based on the further MRI imaging data, and (iii) generating a masking image based on the contour; and
[c] masking, based on the control data and using the MR scanner, one or more regions surrounding the anatomical region of interest for each of one or more MRI images obtained from the MRI imaging data registering the contour with each of the one or more MRI images based on the one or more registration parameters and generating one or more masked MRI images by performing element-wise multiplication between the masking image and each of the one or more MRI images.
Claim limitations [a], [b], and [c] recites an abstract idea.
Claim limitation [a], as drafted and under its broadest reasonable interpretation, recites a mental process that can be performed in the human mind. (MPEP 2106.04(a)(2), III). It involves “observations, evaluations, judgments, and opinions” with or without a physical aid. (Id). As acknowledged in Applicant’s disclosure, the ascertaining may be based on “a purpose of the MRI exam, …and a capability of a physician who will give a diagnosis of the MRI exam.” (emphasis added) ([0028]).
Claim limitation [b] as drafted and under its broadest reasonable interpretation, also recites a mental process that can be performed in the human mind. (MPEP 2106.04(a)(2), III). Determining control data for identifying the anatomical ROI, determining one or more registration parameters, and determining a contour of the ROI includes “observations, evaluations, judgments, and opinions” in determining factors that should be considered when identifying the ROI. For example, the control data may include a contour of the anatomical ROI, a landmark associated with the anatomical ROI, a threshold associated with a distance to a respective one of one or more target objects within anatomical ROI, or a masking MRI image (or mask). ([0029]).
Claim limitation [c] as drafted and under its broadest reasonable interpretation, recites a mathematical concept (e.g., mathematical relationship) because masking one or more regions surrounding the anatomical region of interest involves “organizing information and manipulating information through mathematical correlations.” (MPEP 2106.04(a)(2), III, A, iv). For example, the claim recites performing element-wise multiplication between the masking image and each of the one or more MRI images. Moreover, the mask is applied to remove or filter data where the mask is located. However, image data that is not covered by the mask is not filtered.
This judicial exception is not integrated into a practical application. An additional element includes the steps of “acquiring, using a magnetic resonance (MR) scanner, MRI imaging data for a measurement volume including the anatomical region of interest” and “acquiring, using the MR scanner, further MRI imaging data associated with the anatomical region of interest” However, these steps are extra-solution activity (i.e., the pre-solution activity of gathering data for the mathematical formula/calculations) and does not integrate the judicial exception into a practical application or meaningfully limit the judicial exception. (MPEP 2106.05(g)).
With respect to claim 19, other additional elements are a processor and a memory storing program code, that when executed by the processor, performs the abstract ideas. However, the processor, memory, and program code are recited at a high-level of generality in the claims. As such, the claim limitations amount to no more than mere instructions to apply the judicial exception, which does not impose a meaningful limit on the judicial exception. (MPEP 2106.05(f)).
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional element, “obtaining…MRI imaging data,” is a well-understood, routine, conventional activity that only adds insignificant extra-solution activity to the judicial exception. (See MPEP 2106.05, I, A). Moreover, the processor, memory, and program code are purely functional and generic components that are widely prevalent and in common use throughout the medical/healthcare industries.
The dependent claims do not render the claimed invention patent eligible. For example:
With respect to claim 2, ascertaining the anatomical region of interest based on a purpose of the MRI exam, a type of a radiofrequency (RF) coil used during the MRI exam, and/or a capability of a diagnosing physician for the MRI exam is still a “mental process” because the step involves “observations, evaluations, judgments, and opinions.”
With respect to claims 5, 8, 9, 12, and 13, the additional limitations regarding how control data is determined for identifying the anatomical ROI still involve “observations, evaluations, judgments, and opinions” when selecting known method/factors in MRI imaging. For example, each of the following is a mental process that encompasses a human performing a step mentally with or without a physical aid: determining registration parameters and a contour of the ROI (claim 5); determining landmarks (claim 8); determining a threshold associated with a distance to a respective one of the one or more target objects (claim 9); determining control data based on imaging data associated with the anatomical region of interest that is acquired using one or more other imaging modalities (claim 12); and determining control data based on a difference between a signal-to-noise ratio (SNR) of the anatomical region of interest and a SNR of the one or more regions surrounding the anatomical region of interest (claim 13).
With respect to claim 10, applying a trained machine-learning algorithm to the MRI imaging data to identify the anatomical region of interest in the MRI imaging data is an example of “[m]erely reciting the words ‘apply it’ (or an equivalent) with the judicial exception” and does not integrate the judicial exception into a practical application. (MPEP 2106.05(f)).
With respect to claims 4 and 7, the additional elements do not integrate the judicial exception into a practical application or are insufficient to amount to significantly more than the judicial exception. For example, none of the following meaningfully limit the claim by going beyond generally linking the use of the judicial exception to a particular technological environment: the further MRI imaging data including localizer data and/or one or more localizer images (claim 4); and acquiring the further MRI imaging data using a different RF coil and/or a different MRI sequence (claim 7). Moreover, each of the above is a well-understood, routine, conventional activity in MRI imaging and image processing. (see, e.g., 102/103 rejections of claims 3, 4, and 7 below).
With respect to certain additional limitations of claims 5 and 8, these additional limitations regarding what the masking of one or more regions is based on still recite a mathematical concept (e.g., mathematical relationship) because masking one or more regions surrounding the anatomical region of interest involves “organizing information and manipulating information through mathematical correlations.” (MPEP 2106.04(a)(2),III,A,iv).
With respect to claims 6 and 14-18, the claim limitations either introduce new abstract ideas and/or recite additional elements that do not integrate the judicial exception into a practical application or are insufficient to amount to significantly more than the judicial exception.
With respect to claim 6, selecting a reference image and adjusting a position of the contour in the reference image is a mental process. Registering the contour with the one or more MRI images is also a mental process and/or mathematical concept. Moreover, these are well-understood, routine, conventional activities in MRI imaging and image processing. (see, e.g., 103 rejection of claim 6 below).
With respect to claim 14, registering each of the optical images with MRI imaging data is a mental process and/or mathematical concept and capturing one or more optical images does not meaningfully limit the claim such that the judicial exception is integrated into a practical application or include elements that are sufficient to amount to significantly more than the judicial exception. Moreover, these are well-understood, routine, conventional activities in MRI imaging and image processing. (see, e.g., 103 rejection of claim 14 below).
With respect to claim 15, obtaining a graphical representation associated with the anatomical region of interest based on the one or more masked MRI images does not meaningfully limit the claim such that the judicial exception is integrated into a practical application or include elements that are sufficient to amount to significantly more than the judicial exception. Moreover, this is a well-understood, routine, conventional activity in MRI imaging and image processing. (see, e.g., 103 rejection of claim 15 below).
With respect to claim 16, storing the control data together with the graphical representation does not meaningfully limit the claim such that the judicial exception is integrated into a practical application or include elements that are sufficient to amount to significantly more than the judicial exception. Moreover, this is a well-understood, routine, conventional activity in MRI imaging and image processing. (see, e.g., 103 rejection of claim 16 below).
With respect to claim 17, adjusting imaging parameters associated with the MRI exam based on the graphical representation does not meaningfully limit the claim such that the judicial exception is integrated into a practical application or include elements that are sufficient to amount to significantly more than the judicial exception. Moreover, this is a well-understood, routine, conventional activity in MRI imaging and image processing. (see, e.g., 103 rejection of claim 17 below).
With respect to claim 18, the imaging parameters being parameters associated with the measurement volume does not meaningfully limit the claim such that the judicial exception is integrated into a practical application or include elements that are sufficient to amount to significantly more than the judicial exception. Moreover, this is a well-understood, routine, conventional activity in MRI imaging and image processing. (see, e.g., 103 rejection of claim 18 below).
With respect to claim 21, the claim limitation recites an extra-solution activity (i.e., the pre-solution activity of gathering data for the mathematical formula/calculations) and does not integrate the judicial exception into a practical application or meaningfully limit the judicial exception. (MPEP 2106.05(g)).
With respect to claim 22, the claim limitation recites a mental process that can be performed in the human mind. (MPEP 2106.04(a)(2), III). Determining control data for identifying the anatomical ROI involves “observations, evaluations, judgments, and opinions” in determining factors that should be considered when identifying the ROI. For example, the control data may include a contour of the anatomical ROI, a landmark associated with the anatomical ROI, a threshold associated with a distance to a respective one of one or more target objects within anatomical ROI, or a masking MRI image (or mask). ([0029]).
Accordingly, the claims are not patent eligible.
RESPONSE TO APPLICANT’S ARGUMENTS
Applicant relies upon the added claim limitations and argues that these limitations “clarify the practical implementation of Applicant’s method as part of an MRI imaging system.” (p.10, Response). More specifically, Applicant argues that the claims require two different MRI datasets. “This integration of multi-dataset registration with specific image processing operations represents a concrete technological implementation that goes well beyond abstract mental processes.” (p.10, Response).
Examiner disagrees that the claim limitations integrate the judicial exception into a practical application. “A claim that integrates a judicial exception into a practical application will apply, rely on, or use the judicial exception in a manner that imposes a meaningful limit on the judicial exception, such that the claim is more than a drafting effort designed to monopolize the judicial exception.” (MPEP 2106.04(d)). As explained above, while the added limitations acquire and use multiple MRI datasets, these limitations do not impose meaningfully limits on the claims.
Applicant also alleges that the claimed invention is similar to Example 3 in the USPTO SME Guidelines, which was held to be patent eligible. (p.11, Response). While there are similarities, the claimed invention, unlike Example 3, recites multiple steps that are both mental processes and/or mathematical concepts. Moreover, the USPTO based the decision in Example 3 on the practical benefits provided by the method. For example, a computer would use less memory and the computations would be produced faster while providing an improved digital image. Applicant does not appear to allege these benefits but relies upon an alleged “concrete result” of masked MRI images that show the anatomical regions of interest. However, masked MRI images are known and it does not appear that Applicant’s method generates the masked images using, for example, less memory or performing the computations more quickly. For similar reasons, the claims do not provide “significantly more” and the additional elements are well-understood, routine, and convention activities as explained below with respect to the Section 103 rejection.
Finally, Applicant argues that the claims do not recite a mental process because it contains limitations that cannot be practically performed in the human mind. (p.12, Response). However, courts do not distinguish “between claims that recite mental processes performed by humans and claims that recite mental processes performed on a computer.” (MPEP 2106.04(a)(2), III). “Claims can recite a mental process even if they are claimed as being performed on a computer.” (Id).
In order to overcome the Section 101 rejection, Applicant may wish to consider reciting the anatomical regions of interest that are desired to be viewed (e.g., teeth, jaw, and nearby soft tissue) and those that are being removed (e.g., brain).
Claim Rejections - 35 USC § 103
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, 2, 4, 5, 13, 19, and 22 are rejected under 35 U.S.C. 103 as being unpatentable over U.S. Patent Appl. Publ. No. 2021/0319539 A1 (hereinafter referred to as “AGARWAL”) in view of U.S. Patent Appl. Publ. No. 2014/0314289 A1 (hereinafter referred to as “SPOTTISWOODE”) and U.S. Patent Appl. Publ. No. 2021/0216878 A1 (hereinafter referred to as “NORMAN”).
With respect to claim 1, AGARWAL discloses a computer-implemented method (see, e.g., [0023]: “the methods described herein may be encoded as executable instructions embodied in a tangible, non-transitory, computer readable medium...”) that includes:
ascertaining an anatomical region of interest associated with a magnetic resonance imaging (MRI) exam of a patient. AGARWAL teaches that the MRI images include an area of interest or “useful portion” that includes internal anatomical structures of a patient (i.e., images were acquired with a purpose). ([0028]). AGARWAL teaches that localizer images can be acquired to determine the scan position (i.e., ascertain the anatomical region of interest) for the subsequent scans. ([0005] and [0054]). “The area of interest or the useful portion or the foreground image is [a] portion of the image showing internal structures of the brain and the abdomen in FIGS. 1(a) and 1(b) respectively.” ([0028]; see also, e.g., [0054]).
acquiring, using a magnetic resonance (MR) scanner, MRI imaging data for a measurement volume including the anatomical region of interest. AGARWAL discloses obtaining high resolution MRI images of a measurement volume after localizer scans. For example, “[t]ask (Task 1) of acquiring images through a localizer scan precedes any other tasks (tasks 2 to task n) for acquired other type of high resolution images.…” ([0030]). “The image acquisition system (110) in the present example may be a magnetic resonance imaging (MRI) device.” ([0029]). “The image acquisition system (110) of a medical imaging device acquires the data in k-space that is reconstructed into an image by image reconstruction module (120) which is also referred as image reconstruction pipeline. Image reconstruction module typically involves reconstruction of image from k-space data and processing it further to get human readable, diagnostic quality DICOM images.” ([0030]).
acquiring, using the MR scanner, further MRI imaging data associated with the anatomical region of interest. Applicant’s disclosure describes that the “further MRI imaging data” may include “localizer data and/or one or more localizer images associated with the anatomical region of interest.” ([0069] of Applicant’s disclosure). AGARWAL teaches acquiring localizer images prior to the high-resolution images. “Task (Task 1) of acquiring images through a localizer scan precedes any other tasks (tasks 2 to task n) for acquired other type of high resolution images.…” ([0030]). “Localizer images are pilot images or scouts that are obtained from same imaging system. The localizer image is an image of a portion of the body that includes the area from which the image series is taken but localizer images are generated for a larger area than the area of actual scan from where the image series is generated. Localizer images are also used to identify the location of the area of interest.” ([0054]; see also [0055]-[0056] and Figures 4 and 5).
determining, based on the further MRI imaging data, control data for identifying the anatomical region of interest in the MRI imaging data by…generating a masking image…. AGARWAL teaches determining control data by generating a binary mask to identify the anatomical region of interest. “Localizer images are reconstructed from k-space and during this reconstruction, a foreground mask of the deep learning module (425) is used to mask the foreground pixels and suppress the background pixels. Before applying the mask, morphological operations are performed on the mask to recover any potential loss of anatomical regions and images are returned to ongoing localizer reconstruction process (420).” ([0055]). AGARWAL discloses generating a binary mask “that is used by image reconstruction module (120) to carry out suppressing of background pixels by setting the value of background pixels to zero and not processing them from that point onwards in each subsequent reconstruction operation thereby suppressing the background pixels.” ([0031]; see also [0028] and [0055]). Applicant’s disclosure describes control data as including, among other things, “binary masks.” (see, e.g., p.6, line 20).
masking, based on the control data and using the MR scanner, one or more regions surrounding the anatomical region of interest for each of one or more MRI images obtained from the MRI imaging data… AGARWAL discloses using the binary mask “to carry out suppressing of background pixels by setting the value of background pixels to zero and not processing them from that point onwards in each subsequent reconstruction operation thereby suppressing the background pixels.” ([0031]). “Localizer images are reconstructed from k-space and during this reconstruction, a foreground mask of the deep learning module (425) is used to mask the foreground pixels and suppress the background pixels… During all subsequent scan tasks, same mask is loaded and applied to (setting background pixels value to zero) to get background suppressed images during reconstruction and the identified background pixels will be skipped for remaining stages of reconstruction for Digital Imaging and Communications in Medicine (DICOM) image generation.” ([0055]; see generally [0053]-[0056]).
generating one or more masked MRI images by performing element-wise multiplication between the masking image and each of the one or more MRI images, AGARWAL teaches that a “masked image is obtained by performing pixel wise intensity multiplication of mask pixels and pixels in the original image.” ([0052]; see also [0053]).
However, AGARWAL does not teach determining the anatomical contour from the further MRI data and then generating the masking image based on the contour. However, AGARWAL does teach generating a mask based on localizer images in which the mask is used for subsequent scans. ([0030] and [0053]).
In the same field of endeavor, SPOTTISWOODE teaches determining patient-specific anatomical regions from MRI localizer images. SPOTTISWOODE teaches a computer-implemented method in which “[a]n image mask is created based on predetermined identification information about the tissue of interest. Next, an updated image mask is created based on a largest connected region in the image mask. The updated image mask is applied to the T1-mapping image to yield a masked image.” (Abstract). SPOTTISWOODE teaches that “a-priori information about the location” of patient-specific anatomy may be determined by automatically fitting a model of a sub-region to MRI localizer images. ([0034]; see also [0031]). This model is then used to determine contours of the anatomical region. “Once the T1 image is spatially aligned with the stack of localizer images, a re-slicing process may be applied to the model to estimate a left ventricular epicardial border, a left ventricular endocardial border, and the whereabouts of myocardium using the projection of the fitted mesh polygons. FIG. 6 shows three images 605, 610, 615 which depict the estimating of epicardial and endocardial contours for the T1-map, generated according to some embodiments of the present invention.” ([0037]).
SPOTTISWOODE teaches generating an image mask (e.g., “determining the image mask based on the first sub-region of the anatomical area and the one or more additional sub-regions of the anatomical area.”) ([0008]; see also [0041]: “Next, a-priori information about structure and anatomy in the area of interest are used to create a mask, as shown in image 915.”)
It would have been obvious to one having ordinary skill in the art at the time of filing to configure the AGARWAL method to determine the anatomical contour from the further MRI data, as taught in SPOTTISWOODE, and generate an image bask based on the contour. One of ordinary skill in the art would modify AGARWAL’s localizer-based identification of an anatomical region of interest to delineate a desired sub-region of the anatomical region of interest using SPOTTISWOODE’s fitted anatomical model. As such, the regions retained would be defined by anatomical boundaries within the image and not simply defined by non-anatomical background. There would have been a reasonable expectation of success as SPOTTISWOODE teaches a process that can determine the anatomical contour based on scout images.
However, AGARWAL does not explicitly teach determining one or more registration parameters for registering the further MRI imaging data with the MRI imaging data or, when masking the one or more regions, registering the contour with each of the one or more MRI images based on the one or more registration parameters. However, AGARAWAL teaches a desire to mask foreground pixels and suppress background using a binary mask generated by localizers. (see, e.g., [0031] and [0055]). Moreover, one having ordinary skill in the art would understand that, in order for the masking to be effective in the subsequent scans, the mask should be registered with the subsequent scans.
SPOTTISWOODE, on the other hand, teaches that localizer-derived anatomical locations can become spatially misaligned with subsequent MRI scans. “The localizer images and T1-mapping images in this image were obtained during separate breath holds at the same diastolic phase. The two arrows show misregistration due to slight differences in breath hold position and cardiac phase.” ([0035]). SPOTTISWOODE further teaches that “the stack of localizer images can be re-sliced/interpolated to create a new image at the same plane as the T1-mapping image, and two-dimensional non-linear image coregistration can be used to correct for mismatches in the breath hold position and diastolic phase of the two scans.” ([0036]).
In the same field of endeavor, NORMAN teaches deep learning based coregistration of medical images. With respect to Figure 3, NORMAN teaches registering an image that has a corresponding segmentation mask with another medical image (see, e.g., “a pair of 2D cardiac SAX MR images (source image 302 and target image 303) from a database of medical images (301), where one of the images has a corresponding segmentation mask (304) of ventricular contours, for instance, to include the left ventricular endocardium (LV endo), left ventricular epicardium (LV epi), and/or right ventricular endocardium (RV endo), for example.”) ([0033]). “In at least some of the implementations, the segmentation mask (304) may correspond to the target image (303).” ([0033]). “Using the trained DeformationNet (305), a DDF (306) with respect to the source image is inferred. This DDF (306) is applied to the segmentation mask (304) corresponding to the target image (303) creating a warped segmentation mask (307) that has been warped onto the source image.” ([0033]). Notably, DDF is a dense deformation field (i.e., one or more registration parameters). (see also claim 36: “wherein the warped target image is obtained by applying the dense deformation field to an original target image.”).
It would have been obvious to one having ordinary skill in the art at the time of filing to modify the AGARWAL process to include, prior to masking the MRI images, determining registration parameters for registering the further MRI imaging data with the MRI imaging data and then registering the contour with each of the one or more MRI images based on the one or more registration parameters. While SPOTTISWOODE teaches a method for determining an anatomical contour and generating a mask based on the contour, SPOTTISWOODE also acknowledges that localizer-derived image data can be misaligned with subsequent scans. To avoid misalignment, one having ordinary skill in the art would have been motivated to determine registration parameters (i.e., DDF) for registering the further MRI imaging data with the MRI imaging data and then register the contour with each of the one or more MRI images based on the one or more registration parameters in order to implement ARGAWAL’s method of applying masks to MRI images. There would have been a reasonable expectation of success as SPOTTISWOODE teaches that masks can be based on contours and NORMAN teaches that contours may be registered to MRI images.
With respect to claim 2, AGARWAL discloses wherein the anatomical region of interest is ascertained based on: a purpose of the MRI exam, a type of a radiofrequency (RF) coil used during the MRI exam, and/or a capability of a diagnosing physician for the MRI exam. “The area of interest or the useful portion or the foreground image is portion of the image showing internal structures of the brain and the abdomen in FIGS. 1(a) and 1(b) respectively.” ([0028]). In AGARWAL, the images include a “useful portion” (i.e., images that were acquired with a purpose) and a physician of at least some capability will diagnose, see supervised learning affirmation by human experts [0038] and support of physician’s experience/healthcare practitioners workflow [0043]. Moreover, AGARWAL seeks to remove “clinically irrelevant information” ([0007]), meaning some determination must be made as to what is relevant (i.e., purpose of the MRI exam).
With respect to claim 4, AGARWAL discloses wherein the further MRI imaging data comprise localizer data associated with the anatomical region of interest and/or one or more localizer images associated with the anatomical region of interest. AGARWAL discloses that the control data (e.g., binary mask) can be formed from localizer scans ([0030]) that are subsequently processed by a deep learning module ([0031]). AGARWAL discloses that a “localizer scan” ([0030]) associated with the anatomical region of interest (brain) is acquired and processed by a deep learning module ([0031]) to generate a binary mask. NOTE: Applicant’s disclosure acknowledges that a localizer scan is an example of “further MRI imaging data.” (see, e.g., page 24, lines 33-34).
With respect to claim 5, AGARWAL further teaches that the masking of the one or more regions surrounding the anatomical region of interest includes blurring, removing, or fading out the one or more regions outside the contour for each of the one or more MRI images. (see, e.g., [0052]: “A mask image of an input image is an image where intensity values at some pixels of the input image are set to zero, and intensity values at other pixels of the input image are set to non-zero.”). However, AGARWAL does not teach the determining limitations or the remaining masking limitations of claim 5.
With respect to claim 13, AGARWAL teaches wherein the determining of the control data is further based on a difference between a signal-to-noise ratio (SNR) of the anatomical region of interest and a SNR of the one or more regions surrounding the anatomical region of interest. The masks in AGARWAL are configured to remove surrounding regions with higher background noise. “Some of the advantages of the method include removal of background noise or background artefact as part of image reconstruction while no data of anatomical regions is lost.” ([0065]).
With respect to claim 21, AGARWAL teaches wherein the further MRI imaging data includes additional anatomical regions compared to the MRI imaging data such that the further MRI imaging data comprises parts of the anatomical region of interest and parts of the regions surrounding the anatomical region of interest. AGARAWAL teaches that localizer images or scout images “are low resolution, larger field of view images.” As such, the further MRI imaging data includes parts of the anatomical region of interest and parts of the regions surrounding the anatomical region of interest
With respect to claim 22, the combination of ARGAWAL, SPOTTISWOODE, and NORMAN teach that determining the contour of the anatomical region of interest comprises identifying the anatomical region of interest and the regions surrounding the anatomical region of interest in the further MRI imaging data. ARGAWAL teaches the general desire to obtain useful information regarding the anatomical region of interest from the image and discard irrelevant information. “The area of interest or the useful portion or the foreground image is portion of the image showing internal structures of the brain and the abdomen in FIGS. 1(a) and 1(b) respectively. The background of the image is clinically irrelevant, non-anatomical portion of the image.” ([0028]).
SPOTTISWOODE further teaches determining an image mask based on multiple sub-regions that are near each other or part of a larger structure (see, e.g., “determining the image mask based on the first sub-region of the anatomical area and the one or more additional sub-regions of the anatomical area.”). ([0008]). “The sub-regions will corresponds to the anatomical area of interest. For example, in one embodiment, the first sub-region of the anatomical area corresponds to a left ventricle region and the one or more additional sub-regions of the anatomical area corresponds to a left ventricular epicardial border and a left ventricular endocardial border.” ([0008]). NOTE: Each of the sub-regions can be considered an “anatomical region of interest” and the nearby, adjacent sub-regions can be considered “regions surrounding the anatomical region.”
It would have been obvious to one having ordinary skill in the art at the time of filing to modify the AGARWAL method to identify pixels that correspond to anatomical regions of interest and pixels that correspond to regions that surround the anatomical regions of interest. AGARWALL teaches the general desire to acquire useful information about the anatomical region of interest, which includes the regions that surround the anatomical region of interest (e.g., various chambers of the heart). NORMAN teaches a known technique for classifying pixels. Applying the known techniques taught in NORMAN to the regions surrounding the anatomical region of interest would predictably provide a more anatomically accurate boundary for the desired region.
Claims 6, 8, and 9 are rejected under 35 U.S.C. 103 as being unpatentable over U.S. Patent Appl. Publ. No. 2021/0319539 A1 (hereinafter referred to as “AGARWAL”) in view of U.S. Patent Appl. Publ. No. 2014/0314289 A1 (hereinafter referred to as “SPOTTISWOODE”) and U.S. Patent Appl. Publ. No. 2021/0216878 A1 (hereinafter referred to as “NORMAN”) as applied to claim 1 above, and further in view of International Patent Appl. No. WO 2010/134013 A1 to Von Berg et al. (hereinafter referred to as “VON BERG”).
With respect to claim 6, as discussed above with respect to claim 5, VON BERG teaches selecting a reference image from the one or more MRI images. The input images may be MRI images that are “two-dimensional slices or computed projections of three-dimensional” image. (p.5, lines 3-4). The user may indicate “a portion of at least one of the input images.” (p.5, line 11).
With respect to adjusting, in the reference image, a position of and/or a size of the contour of the anatomical region of interest, VON BERG teaches that the contour may be modified. “The segmenting means 2 may be arranged for performing model-based segmentation by deformable contours, which is a technique known in the art per se….Alternatively, region growing may be applied, starting from the user- indicated portion. By incorporating prior knowledge about the objects, such as the spatial relation to each other, the segmentation quality of the model-based segmentation may be improved.” (p.5, lines 20-26). See also: “The graphical user interface 1 may be arranged for receiving a further input indicating a further portion of at least one of the input images. The segmenting means 2 may be arranged for segmenting a further object represented by the input image based on the further portion.” (p.6, line 34 to p.7, line 2).
With respect to wherein the registering of the contour with each of the one or more MRI images comprises: registering the adjusted contour with each of the one or more MRI images based on the one or more registration parameters, VON BERG teaches that, after adjusting, the adjusted contour may be registered with the input images. “The segmenting means 2 may be arranged for segmenting a further object represented by the input image based on the further portion. The mask identifying means 3 may be arranged for identifying a further mask based on the further object. The registration means 4 may be arranged for registering the images taking into account the further mask, for obtaining at least two further registered images.” (p.7, lines 2-6).
It would have been obvious to one skilled in the art to modify AGARWAL-VON BERG to select a reference image and adjust a contour within the reference image using VON BERG’s method and then register the MRI images using the adjusted contour as taught in VON BERG. One would have been motivated to modify AGARWAL-VON BERG of claim 5 in this manner because VON BERG’s method improves the quality of segmentation. (p.5, lines 20-24). There would be a reasonable expectation of success because, as taught by VON BERG, computing devices can be configured to perform the method. (see, e.g., p.7, lines 7-8).
With respect to claim 8, AGARWAL teaches each and every limitation of claim 1 as discussed above. However, AGARWAL does not teach the limitations of claim 8.
In the same field of endeavor, VON BERG teaches “The specification, selection, and modification of the mask may be implemented as follows. There may be provided a navigation view that presents the image to be registered. The user may select image regions to be included or excluded. The mask may be visualized in the navigation view, e.g. by an overlay. There may be at least two working modes: The anatomical mode and the image mode. In the anatomical mode, selection of anatomical structures [i.e., landmarks] is possible, e.g. the clavicle, the ribs.” (emphasis added) (p.9, lines 4-9). Accordingly, VON BERG teaches wherein the determining of the control data comprises: determining, based on the MRI imaging data, one or more landmarks associated with the anatomical region of interest, the control data including the one or more landmarks.
With respect to wherein the masking of the one or more regions surrounding the anatomical region of interest is based on the one or more landmarks and/or a pre-defined model of the anatomical region of interest. As described immediately above, modification of the “mask” may be implemented using an “anatomical mode” where landmarks, such as “clavicle, the ribs” may be selected. As such, the masking would be based on the one or more landmarks.
It would have been obvious to one skilled in the art to modify AGARWAL to determine one or more landmarks using VON BERG’s method and then mask the MRI images using the one or more landmarks as taught in VON BERG. One would have been motivated to modify AGARWAL in this manner because selecting landmarks using the anatomical mode taught in VON BERG provides one easy option for designing the mask that is applied to the MRI images. (p.9, lines 8-9). There would be a reasonable expectation of success because, as taught by VON BERG, computing devices can be configured to perform the method. (see, e.g., p.7, lines 7-8).
With respect to claim 9, AGARWAL teaches each and every limitation of claim 1 as discussed above. However, AGARWAL does not teach wherein the anatomical region of interest comprises one or more target objects and the determining of the control data comprises determining a distance threshold value measured outwardly from an outer boundary of each respective one of the one or more target objects, the control data including the threshold.
In the same field of endeavor, VON BERG teaches that the anatomical region of interest may include target objects (i.e., the segmented objects). Moreover, VON BERG teaches that “[a] margin around the segmented object may also be excluded from the mask.” Here, the margin is a distance that separates an edge of the segmented object from a region that will be excluded by the mask (i.e., measured outwardly from an outer boundary). Accordingly, VON BERG teaches determining a threshold associated with a distance to a respective one of the one or more target objects.
It would have been obvious to one skilled in the art to modify AGARWAL to enable the user to provide a threshold distance (i.e., margin) that surrounds the target objects as taught in VON BERG. One would have been motivated to modify AGARWAL in this manner because providing a margin for a target object ensures that the anatomy surrounding the anatomical region of interest will be provided and/or the anatomical region of interest will not be inadvertently masked. There would be a reasonable expectation of success because, as taught by VON BERG, computing devices can be configured to perform the method. (see, e.g., p.7, lines 7-8).
Claims 7, 10, and 12 are rejected under 35 U.S.C. 103 as being unpatentable over U.S. Patent Appl. Publ. No. 2021/0319539 A1 (hereinafter referred to as “AGARWAL”) in view of U.S. Patent Appl. Publ. No. 2014/0314289 A1 (hereinafter referred to as “SPOTTISWOODE”) and U.S. Patent Appl. Publ. No. 2021/0216878 A1 (hereinafter referred to as “NORMAN”) as applied to claim 1 above, and further in view of U.S. Patent Appl. Publ. No. 2021/0295501 A1 to Hu et al. (hereinafter referred to as “HU”).
With respect to claim 7, AGARWAL teaches that the localizer or scout images (i.e., further imaging data) are “low resolution, larger field of view images.” ([0005]). However, AGARWAL does not explicitly teach wherein the further MRI imaging data are acquired using a different RF coil from that for obtaining of the MRI imaging data and/or a different MRI sequence from that for obtaining of the MRI imaging data.
In the same field of endeavor, HU teaches that a scout image (or localizer image) “may be acquired by an MR device via scanning the object according to a fast pre-scanning sequence (also referred to as a fast pre-scanning pulse sequence). Exemplary fast pre-scanning sequences may include a low-resolution magnetic resonance sequence, a fast pulse sequence, etc.” ([0089]).
It would have been obvious to one skilled in the art to acquire the localizer images (further MRI imaging data) using a different pulse sequence than that used for diagnostic imaging. One skilled in the art would have been motivated to use a “fast pre-scanning pulse sequence” because imaging with such pulse sequences can be performed more quickly and require less data. There would have been a reasonable expectation of success because, as taught in HU, such pulse sequences have been used in MRI imaging.
With respect to claim 10, while AGARWAL teaches using a trained machine-learning algorithm for analyzing localizer scans and generating binary masks, AGARWAL does not explicitly teach wherein the determining of the control data comprises: applying a trained machine-learning algorithm to the MRI imaging data to identify the anatomical region of interest in the MRI imaging data.
In the same field of endeavor, HU teaches a system and method for image reconstruction and processing. (Abstract). HU seeks to improve the efficiency and accuracy of image reconstruction by addressing reducing the amount of image data acquired during a scanning session. ([0003]). To this end, the region of interest (i.e., region to be imaged for diagnosis) is determined prior to the diagnostic imaging. “In some embodiments, the ROI may be determined based on a reference image….” ([0088]). This reference image may include a pre-scanning “scout image.” ([0089]).
In HU, the system processes “the scout image based on an image segmentation technique to segment the one or more ROIs.” ([0092]). For example, processing the scout image may be “based on a trained machine learning model (also referred to as a trained ROI determination model).” (Id). “[T]he trained ROI determination model may be determined by training a machine learning model offline based on a plurality of training samples using the processing device 120 or a processing device other than the processing device 120.” (Id).
It would have been obvious to one skilled in the art to apply a trained machine-learning algorithm to the MRI imaging data of AGARWAL (i.e., localizer images) to identify the anatomical region of interest therein when determining the control data. One would have been motivated to use a trained machine-learning model, as taught in HU, to determine the anatomical region of interest because machine-learning models can do so quickly and reliably. (see, e.g., [0032] and [0043] of AGARWAL). There would have been a reasonable expectation of success because, as taught in HU, trained machine learning models can be used to identify ROIs.
With respect to claim 12, AGARWAL does not teach wherein the determining of the control data is further based on imaging data associated with the anatomical region of interest that is acquired using one or more other imaging modalities.
In the same field of endeavor, HU teaches a system and method for image reconstruction and processing. (Abstract). HU seeks to improve the efficiency and accuracy of image reconstruction by addressing reducing the amount of image data acquired during a scanning session. ([0003]). To this end, the region of interest (i.e., region to be imaged for diagnosis) may be determined using an imaging device other than the imaging device (e.g., MRI) that will scan for diagnosis. (see, e.g., [0068] and [0088]). “The determination module 420 may determine the position of the ROI in the scanning region of the first imaging device based on a relationship between the coordinate system of the first imaging device and the coordinate system of the second imaging device.” ([0068]).
The scout image may be acquired by “an optical imaging device (e.g., a camera),” a PET device, CT device, or an X-ray device. ([0090]). “After obtaining the scout image, the one or more ROIs of the object may be determined manually or automatically,”(Id), or by using a trained machine-learning model ([0092]). The position of the ROI may then be determined by converting the coordinates of the other imaging device into the coordinates of the medical imaging device that is acquiring the relevant image. ([0116]).
It would have been obvious to one skilled in the art to determine the control data based on imaging data associated with the anatomical region of interest using a different imaging modality. One would have been motivated to use imaging data from the different modality because imaging data from a different modality may be used to, for example, more quickly and accurately position the patient for scanning by the primary modality. There would have been a reasonable expectation of success because, as taught in HU, other modalities may be used in conjunction with the primary modality to determine how to better and/or more quickly position the ROI within the primary modality’s scanning region.
Claim 14 is rejected under 35 U.S.C. 103 as being unpatentable over U.S. Patent Appl. Publ. No. 2021/0319539 A1 (hereinafter referred to as “AGARWAL”) in view of U.S. Patent Appl. Publ. No. 2014/0314289 A1 (hereinafter referred to as “SPOTTISWOODE”) and U.S. Patent Appl. Publ. No. 2021/0216878 A1 (hereinafter referred to as “NORMAN”) as applied to claim 1 above, and further in view of Zhang, Lu, et al. “Dynamic 3D images fusion of the temporomandibular joints: A novel technique.” Journal of Dentistry 126 (2022) to Zhang et al. (hereinafter referred to as “ZHANG”). NOTE: Zhang was publicly available as of 10 September 2022 (see bottom left corner of first page).
With respect to claim 14, AGARWAL does not teach the method further comprising capturing one or more optical images including the anatomical region of interest; and registering each of the one or more optical images with the MRI imaging data, the determining of the control data being further based on the one or more optical images.
In the same field of endeavor, ZHANG teaches “a procedure for fusing images from cone-beam computed tomography (CBCT), magnetic resonance imaging (MRI) and optical positioning tracking system to dynamically evaluate the relative motion of the temporomandibular joint (TMJ) for the diagnosis of temporomandibular disorders (TMD).” (Abstract). Specifically, ZHANG “combines the optical position tracking system with CBCT and MRI images to visually reproduce the fossa-disc-condyle real movement status.” (page 2, Introduction, top left corner). An optical tracking system was used “to record the mandibular movements in real time through the marker points attached to the mandibular fork.” (page 4, left column, Discussion, second paragraph). “In addition to the static fusion of 3D image data, the integration of dynamic mandibular movement trajectories into 3D virtual patients to transform into 4D virtual patients is the future direction for digital dentistry. This study proposes a method to reproduce the functional movements of TMJ by fusing mandibular movement trajectories with CBCT and MRI data, which can be used as a tool for dynamic diagnosis of TMJ disorders and provide new ideas for 4D virtual patients.” (page 4, left column, Discussion, first paragraph).
It would have been obvious to one skilled in the art to capture optical images including the anatomical region of interest and register the optical images with the MRI imaging data such that the control data is based on the optical images. One would have been motivated to implement a method like Zhang by fusing (i.e., registering) optical imaging with the MRI imaging to reproduce functional anatomical movements by joints because doing so, as taught in Zhang, provides a tool for dynamic diagnosis of anatomical disorders. (page 4, left column, Discussion, first paragraph). There would have been a reasonable expectation of success because, as taught in Zhang, optical imaging can be registered with MRI imaging.
Claims 15-18 are rejected under 35 U.S.C. 103 as being unpatentable over U.S. Patent Appl. Publ. No. 2021/0319539 A1 (hereinafter referred to as “AGARWAL”) in view of U.S. Patent Appl. Publ. No. 2014/0314289 A1 (hereinafter referred to as “SPOTTISWOODE”) and U.S. Patent Appl. Publ. No. 2021/0216878 A1 (hereinafter referred to as “NORMAN”) as applied to claim 1 above, and further in view of SyMRI® SyntheticMR AB (2019-12-16), User Manual Version 11.2.2 (hereinafter “SYMRI”).
With respect to claim 15, AGARWAL does not teach obtaining a graphical representation associated with the anatomical region of interest based on the one or more masked MRI images.
In the same field of endeavor, SYMRI is a user manual for imaging visualization software. “SyMRI is a post-processing software medical device intended for use in visualization of the brain.” (p.2, Intended Use, 1.2.1 Indications for use). “SyMRI automatically segments several tissue types….The automatic segmentation is done on all voxels contained within the [intracranial volume (ICV)], defined by the intracranial mask.” (p.44, 6.2 Tissue Segmentation, first paragraph). The ICV is a masked MRI image. “The intracranial mask defines the intracranial volume in SyMRI NEURO. All segmentation and volume calculations are applied to the voxels included in the intracranial mask.” (p.42, 6.1 The Intracranial Mask, first paragraph after Figure 6.2). Results of the tissue segmentation can be presented as “segmentation maps.” (p.44, 6.2 Tissue Segmentation, first paragraph). “Segmentation maps are used to visualize the tissue content in each voxel within the ICV. As default, segmentation maps are displayed as overlays, colored images that are displayed over contrast weighted images.” (p.47, 6.3 Segmentation Maps, first paragraph). “The overlays have different color depending on which tissue they represent.” (emphasis added) (Id). Figure 6.4 illustrates distinct overlays for the different tissue.
It would have been obvious to one skilled in the art to program the computing device to obtain a graphical representation associated with the anatomical region of interest based on the one or more masked MRI images. More specifically, one skilled in the art would have configured the system to provide a graphic/overlay that identifies a location of the ROI (of the measurement volume) within the image as taught in SYMRI. One skilled in the art would be motivated to make this modification because it is important for physicians and researchers to be able to quickly and reliably distinguish the region of interest from other regions of the measurement volume. (see, e.g., p.2, 1.2 Intended Use, 1.2.1 Indications for use, third paragraph: “When interpreted by a trained physician, SyMRI images can provide information useful in determining diagnosis.”) There would be a reasonable expectation of success because, as taught in SYMRI, imaging systems can be configured to display graphics to the user along with the medical images.
With respect to claim 16 (depending from claim 15), SYMRI teaches storing the control data together with the graphical representation associated with the anatomical region of interest. “Various types of images and image stacks can be saved from SyMRI to your workstation or PACS.” (p. 37, Section 5.5 Save and Load, first paragraph). SYMRI provides quantification maps called SyMaps. (p.32, Section 5.2 SyMaps). “The quantification maps in SyMRI display the measured values for tissue properties per voxel in a dataset.” (Id). “When saving SyMaps the intracranial mask is saved to the saved SyMaps dataset.” (p. 37, 6.1.2 Additional Functions for the Intracranial Mask, paragraph with “StandAlone” graphic). As shown in Figure 5.5, the SyMaps include graphical overlays that identify different tissues. (p.32, Section 5.2 SyMaps). Accordingly, SYMRI teaches storing the control data (i.e., intracranial mask) together with the graphical representation (i.e., graphical overlays identifying different tissues associated with the anatomical region of interest).
It would have been obvious to one skilled in the art to program the computing device to store the control data (i.e., a mask) together with the graphical representation (i.e., graphical overlays). More specifically, one skilled in the art would program the computing device to allow a user to save an image dataset (i.e., masked MRI images) along with one or more masks that was applied to the image dataset and the graphical representation that is applied to the images after masking. One skilled in the art would be motivated to make this modification to load the saved image dataset more quickly. (see, e.g., p. 38, Section 5.5.1.1 Save Time, only paragraph: “Before SyMRI is closed, saving the quantified SyMaps is recommended if you want to load the dataset at a later time. This is recommended because it is faster to load quantified SyMaps than starting from unquantified raw data (MDME).”) There would be a reasonable expectation of success because, as taught in SYMRI, masked MRI images can be saved with the corresponding masks and graphical objects.
With respect to claim 17 (depending from claim 15), SYMRI teaches adjusting imaging parameters associated with the MRI exam based on the graphical representation associated with the anatomical region of interest. SYMRI provides maps/images based on imaging parameters. “The parametric maps are R1, R2 relaxation rates, and proton density (PD). The inverse relaxation parameters, T1 relaxation time (1/R1), and T2 relaxation time (1/R2) are also provided as parametric maps.” (emphasis added) (p.1, Section 1.1 Device Description). These parametric maps correspond to the SyMaps discussed above in claim 16. “SyMRI generates all the different image contrasts from the same parametric maps, derived from the same acquisition…SyMRI provides the user the ability to change the contrast of the images after the acquisition. This is performed by adjusting the TE, TR, and/or TI parameters post-acquisition, to generate the specific contrast desired.” (emphasis added) (p.1, Section 1.1 Device Description). Accordingly, one is able to view the parametric maps (i.e., SyMaps) and can change the contrast of those images, thereby effectively changing the imaging parameters post-acquisition.
It would have been obvious to one skilled in the art to program the computing device to enable a user to adjust imaging parameters associated with the MRI exam based on the graphical representation. More specifically, one skilled in the art would program the computing device to allow a user to “generate multiple image contrasts from a single acquisition scan.” (Id). One skilled in the art would be motivated to make this modification because the ability to view multiple image contrasts and change those contrasts after acquisition provides information that is useful in determining a diagnosis. (see, e.g., p.2, Section 1.2 Intended Use). There would be a reasonable expectation of success because, as taught in SYMRI, a computing device can be programmed to present multiple image contrasts and enable a user to change those contrasts.
With respect to claim 18 (depending from claim 17), SYMRI teaches wherein the imaging parameters comprise parameters associated with the measurement volume. As discussed above with respect to claim 17, SYMRI teaches that the imaging parameters may “provide information useful in determining diagnosis.” (see, e.g., p.2, Section 1.2 Intended Use). “SyMRI is indicated for head imaging.” (Id). “SyMRI provides the user the ability to change the contrast of the images after the acquisition. This is performed by adjusting the TE, TR, and/or TI parameters post-acquisition, to generate the specific contrast desired.” (emphasis added) (p.1, Section 1.1 Device Description). Accordingly, SYMRI teaches that the imaging parameters are associated with the measurement volume (i.e., head) because changing those imaging parameters can provide better information for determining a diagnosis.
Claims 19 and 20 are rejected under 35 U.S.C. 103 as being unpatentable over U.S. Patent Appl. Publ. No. 2021/0319539 A1 (hereinafter referred to as “AGARWAL”) in view of U.S. Patent Appl. Publ. No. 2014/0314289 A1 (hereinafter referred to as “SPOTTISWOODE”) and U.S. Patent Appl. Publ. No. 2021/0216878 A1 (hereinafter referred to as “NORMAN”) as applied to claim 1 above, and further in view of U.S. Patent Appl. Publ. No. 2013/0252196 A1 to Rasche et al. (hereinafter referred to as “RASCHE”).
With respect to claim 19, AGARWAL teaches a computing device (“The image reconstruction module (120) may be a processor, or a computer-based device….” (emphasis added) ([0030])) that includes a processor; and a memory storing program code, that when executed by the processor, configure the computing device to. “Therefore, the image reconstruction module (120) may contain encoded and executable instructions embodied in a tangible, non-transitory, computer readable medium, including, without limitation, a storage device and/or a memory device. Such instructions, when executed by a processor, cause the processor to perform at least a portion of the methods described herein.” (emphasis added) (Id).
acquire, using a magnetic resonance (MR) scanner, MRI imaging data for a measurement volume including the anatomical region of interest. AGARWAL discloses obtaining high resolution MRI images of a measurement volume after localizer scans. For example, “[t]ask (Task 1) of acquiring images through a localizer scan precedes any other tasks (tasks 2 to task n) for acquired other type of high resolution images.…” ([0030]). “The image acquisition system (110) in the present example may be a magnetic resonance imaging (MRI) device.” ([0029]). “The image acquisition system (110) of a medical imaging device acquires the data in k-space that is reconstructed into an image by image reconstruction module (120) which is also referred as image reconstruction pipeline. Image reconstruction module typically involves reconstruction of image from k-space data and processing it further to get human readable, diagnostic quality DICOM images.” ([0030]).
acquire, using the MR scanner, further MRI imaging data associated with the anatomical region of interest. Applicant’s disclosure describes that the “further MRI imaging data” may include “localizer data and/or one or more localizer images associated with the anatomical region of interest.” ([0069] of Applicant’s disclosure). AGARWAL teaches acquiring localizer images prior to the high-resolution images. “Task (Task 1) of acquiring images through a localizer scan precedes any other tasks (tasks 2 to task n) for acquired other type of high resolution images.…” ([0030]). “Localizer images are pilot images or scouts that are obtained from same imaging system. The localizer image is an image of a portion of the body that includes the area from which the image series is taken but localizer images are generated for a larger area than the area of actual scan from where the image series is generated. Localizer images are also used to identify the location of the area of interest.” ([0054]; see also [0055]-[0056] and Figures 4 and 5).
determine, based on the further MRI imaging data, control data for identifying the anatomical region of interest in the MRI imaging data by…generating a masking image…. AGARWAL teaches determining control data by generating a binary mask to identify the anatomical region of interest. “Localizer images are reconstructed from k-space and during this reconstruction, a foreground mask of the deep learning module (425) is used to mask the foreground pixels and suppress the background pixels. Before applying the mask, morphological operations are performed on the mask to recover any potential loss of anatomical regions and images are returned to ongoing localizer reconstruction process (420).” ([0055]). AGARWAL discloses generating a binary mask “that is used by image reconstruction module (120) to carry out suppressing of background pixels by setting the value of background pixels to zero and not processing them from that point onwards in each subsequent reconstruction operation thereby suppressing the background pixels.” ([0031]; see also [0028] and [0055]). Applicant’s disclosure describes control data as including, among other things, “binary masks.” (see, e.g., p.6, line 20).
mask, based on the control data, one or more regions surrounding the anatomical region of interest for each of one or more MRI images obtained from the MRI imaging data… AGARWAL discloses using the binary mask “to carry out suppressing of background pixels by setting the value of background pixels to zero and not processing them from that point onwards in each subsequent reconstruction operation thereby suppressing the background pixels.” ([0031]). “Localizer images are reconstructed from k-space and during this reconstruction, a foreground mask of the deep learning module (425) is used to mask the foreground pixels and suppress the background pixels… During all subsequent scan tasks, same mask is loaded and applied to (setting background pixels value to zero) to get background suppressed images during reconstruction and the identified background pixels will be skipped for remaining stages of reconstruction for Digital Imaging and Communications in Medicine (DICOM) image generation.” ([0055]; see generally [0053]-[0056]).
generating one or more masked MRI images by performing element-wise multiplication between the masking image and each of the one or more MRI images, AGARWAL teaches that a “masked image is obtained by performing pixel wise intensity multiplication of mask pixels and pixels in the original image.” ([0052]; see also [0053]).
However, AGARWAL does not teach determining the contour of the anatomical region based on the further MRI data and then generating the masking image based on the contour. However, AGARWAL does teach generating a mask based on localizer images in which the mask is used for subsequent scans. ([0030] and [0053]).
In the same field of endeavor, SPOTTISWOODE teaches determining patient-specific anatomical regions from MRI localizer images. SPOTTISWOODE teaches a computer-implemented method in which “[a]n image mask is created based on predetermined identification information about the tissue of interest. Next, an updated image mask is created based on a largest connected region in the image mask. The updated image mask is applied to the T1-mapping image to yield a masked image.” (Abstract). SPOTTISWOODE teaches that “a-priori information about the location” of patient-specific anatomy may be determined by automatically fitting a model of a sub-region to MRI localizer images. ([0034]; see also [0031]). This model is then used to determine contours of the anatomical region. “Once the T1 image is spatially aligned with the stack of localizer images, a re-slicing process may be applied to the model to estimate a left ventricular epicardial border, a left ventricular endocardial border, and the whereabouts of myocardium using the projection of the fitted mesh polygons. FIG. 6 shows three images 605, 610, 615 which depict the estimating of epicardial and endocardial contours for the T1-map, generated according to some embodiments of the present invention.” ([0037]).
SPOTTISWOODE teaches generating an image mask (e.g., “determining the image mask based on the first sub-region of the anatomical area and the one or more additional sub-regions of the anatomical area.”) ([0008]; see also [0041]: “Next, a-priori information about structure and anatomy in the area of interest are used to create a mask, as shown in image 915.”)
It would have been obvious to one having ordinary skill in the art at the time of filing to configure the AGARWAL method to determine the anatomical contour from the further MRI data, as taught in SPOTTISWOODE, and generate an image bask based on the contour. One of ordinary skill in the art would modify AGARWAL’s localizer-based identification of an anatomical region of interest to delineate a desired sub-region of the anatomical region of interest using SPOTTISWOODE’s fitted anatomical model. As such, the regions retained would be defined by anatomical boundaries within the image and not simply defined by non-anatomical background. There would have been a reasonable expectation of success as SPOTTISWOODE teaches a process that can determine the anatomical contour based on scout images.
However, AGARWAL does not explicitly teach determining one or more registration parameters for registering the further MRI imaging data with the MRI imaging data or, when masking the one or more regions, registering the contour with each of the one or more MRI images based on the one or more registration parameters. However, AGARAWAL teaches a desire to mask foreground pixels and suppress background in the subsequent MRI images using a binary mask generated by localizers. (see, e.g., [0031] and [0055]). Moreover, one having ordinary skill in the art would understand that, in order for the masking to be effective in the subsequent scans, the mask should be appropriately positioned for the subsequent scans.
SPOTTISWOODE, on the other hand, teaches that localizer-derived anatomical locations can become spatially misaligned with subsequent MRI scans. “The localizer images and T1-mapping images in this image were obtained during separate breath holds at the same diastolic phase. The two arrows show misregistration due to slight differences in breath hold position and cardiac phase.” ([0035]). SPOTTISWOODE further teaches that “the stack of localizer images can be re-sliced/interpolated to create a new image at the same plane as the T1-mapping image, and two-dimensional non-linear image coregistration can be used to correct for mismatches in the breath hold position and diastolic phase of the two scans.” ([0036]).
In the same field of endeavor, NORMAN teaches deep learning based coregistration of medical images. With respect to Figure 3, NORMAN teaches registering an image that has a corresponding segmentation mask with another medical image (see, e.g., “a pair of 2D cardiac SAX MR images (source image 302 and target image 303) from a database of medical images (301), where one of the images has a corresponding segmentation mask (304) of ventricular contours, for instance, to include the left ventricular endocardium (LV endo), left ventricular epicardium (LV epi), and/or right ventricular endocardium (RV endo), for example.”) ([0033]). “In at least some of the implementations, the segmentation mask (304) may correspond to the target image (303).” ([0033]). “Using the trained DeformationNet (305), a DDF (306) with respect to the source image is inferred. This DDF (306) is applied to the segmentation mask (304) corresponding to the target image (303) creating a warped segmentation mask (307) that has been warped onto the source image.” ([0033]). Notably, DDF is a dense deformation field (i.e., one or more registration parameters). (see also claim 36: “wherein the warped target image is obtained by applying the dense deformation field to an original target image.”).
It would have been obvious to one having ordinary skill in the art at the time of filing to modify the AGARWAL process to include, prior to masking the MRI images, determining registration parameters for registering the further MRI imaging data with the MRI imaging data and then registering the contour with each of the one or more MRI images based on the one or more registration parameters. While SPOTTISWOODE teaches a method for determining an anatomical contour and generating a mask based on the contour, SPOTTISWOODE also acknowledges that localizer-derived image data can be misaligned with subsequent scans. To avoid misalignment, one having ordinary skill in the art would have been motivated to determine registration parameters (i.e., DDF) for registering the further MRI imaging data with the MRI imaging data and then register the contour with each of the one or more MRI images based on the one or more registration parameters in order to implement ARGAWAL’s method of applying masks to MRI images. There would have been a reasonable expectation of success as SPOTTISWOODE teaches that masks can be based on contours and NORMAN teaches that contours may be registered to MRI images.
However, AGARAWAL does not teach that the computing device can ascertain an anatomical region of interest associated with a magnetic resonance imaging (MRI) exam of a patient. But AGARWAL teaches that the MRI images include an area of interest or “useful portion” that includes internal anatomical structures of a patient (i.e., images were acquired with a purpose). ([0028]).
In the same field of endeavor, RASHCHE teaches a method that includes automatically selecting at least one recording sequence based on at least one of a particular diagnostic application of interest and FoV information. (Abstract). “Before a diagnostic inquiry is performed and a corresponding diagnostic image is recorded in modern clinical practice, a conventional MRI system used to acquire the diagnostic image can require meticulous preparation...” ([0004]). “[I]t would be useful to provide an imaging system that enables an operator to quickly perform a diagnostic examination, without requiring the operator to spend time generating a recording sequence that is used by the imaging system to perform the diagnostic examination or for manual, interactive target volume definition.” ([0016]).
The recording sequences can be determined based on the selected diagnostic application. RASCHE uses a “HF coil unit 210 [that] can recognize the type HF coil 114 that is currently inserted into the receptacle…Each of the different types of HF coils 114 is used for a particular application. When the HF coil unit 210 recognizes a particular type of HF coil 114 (e.g., a type used to conduct investigations of caries), the HF coil unit 210 sends the computer processor 222 an indication of the recognized type of the HF coil 114, which the computer processor 222 uses to automatically select a particular application (e.g., the Caries Application), without requiring a user to manually select that application via the input unit 230.” ([0096]). Applications are directed to particular regions of interest and require particular configurations or recording sequences. For example, “[t]he Caries Application is used to investigate one or more of the patient's teeth using a recording sequence that is appropriate for measuring hard substances (e.g., dentin and enamel) and generating a corresponding image with a high contrast.” ([0087]).
It would have been obvious to one skilled in the art to modify the computing device to ascertain an anatomical region of interest associated with a magnetic resonance imaging (MRI) exam of a patient. More specifically, one skilled in the art would program the computing device to automatically determine a recording sequence, as taught in RASCHE, based on a diagnostic application ([0087]) or other indicator ([0096]) that identifies the anatomical region of interest. One would have been motivated to modify the computing device in this manner because doing so would reduce the time to prepare the MRI imaging system for the diagnostic examination, which RASCHE notes would be useful. ([0016]). There would be a reasonable expectation of success because, as taught by RASCHE, computing devices can be programmed to automatically determine recording sequences based on the diagnostic application and/or other indicators.
With respect to claim 20 (depending from claim 19), AGARWAL teaches a magnetic resonance imaging (MRI) scanner comprising the computing device of claim 19. “FIG. 2 shows an example system (100) according to one aspect of the disclosure. The system (100) includes an image acquisition system (110) configured to acquire images of a subject (not shown). The image acquisition system (110) in the present example may be a magnetic resonance imaging (MRI) device.” ([0029]).
RESPONSE TO APPLICANT’S ARGUMENTS
Applicant’s arguments with respect to claims 1, 2, 4-10, and 12-22 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument. More specifically, Examiner is now relying upon SPOTTISWOODE and NORMAN.
Prior Art Made of Record
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Hoffmann, Malte, et al. "Rapid head‐pose detection for automated slice prescription of fetal‐brain MRI." International journal of imaging systems and technology 31.3 (2021): 1136-1154.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JASON P GROSS whose telephone number is (571)272-1386. The examiner can normally be reached Monday-Friday 9:00-5:00CT.
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, Anne M. Kozak can be reached at (571) 270-0552. 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.
/JASON P GROSS/Examiner, Art Unit 3797
/SERKAN AKAR/ Primary Examiner, Art Unit 3797
