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 .
Priority
Receipt is acknowledged that application is a National Stage application of PCT/EP2021/083325. Receipt is acknowledged that application claims priority to foreign application with application number EP21161545.5 dated 03/09/2021 and application number EP21167803.2 dated 04/12/2021. Copies of certified papers required by 37 CFR 1.55 have been received. Priority is acknowledged under 35 USC 119(e) and 37 CFR 1.78. Claims 1-15 have been afforded the benefit of filing date 03/09/2021.
Response to Amendment
The amendment filed 02/25/2026 has been entered. Applicant’s amendments to the drawings and claims have overcome each and every objection previously set forth in the Non-Final Office Action mailed 11/26/2025. Applicant’s amendments to the claims have overcome each and every 112(b) rejection previously set forth in the Non-Final Office Action mailed 11/26/2025. The double patenting rejections previously set forth in the Non-Final Office Action mailed 11/26/2025 are withdrawn. Claims 1-13 and 15 remain pending in the application, with claim 14 having been cancelled.
Response to Arguments
In response to applicant's arguments against the references individually (arguments directed to Zaharchuk and Schlemper on pg. 11-13 of the Remarks filed 02/25/2026), one cannot show nonobviousness by attacking references individually where the rejections are based on combinations of references. See In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck & Co., 800 F.2d 1091, 231 USPQ 375 (Fed. Cir. 1986).
Regarding amended claim 1, applicant argues in pg. 13, ¶2 of the Remarks filed 02/25/2026, that Schlemper fails to cure the deficiencies of Zaharchuk. The examiner respectfully disagrees. The amendments to claim 1 include the following underlined limitations:
PNG
media_image1.png
308
764
media_image1.png
Greyscale
Zaharchuk teaches a deep learning model that generates one or more second reference representations which represent a contrast of the examination region during a second time span (full-dose images – FIG. 2-3, para 24-28). The first and second reference representations are indicative of the examination region with different contrast information (multi-contrast images – para 24). While Zaharchuk fails to explicitly teach wherein the first and second reference representations are in the frequency space, Schlemper cures this deficiency (see motivation to combine in the claim 1 rejection below). Schlemper teaches a neural network that processes data in the frequency domain (para 83-85). Accordingly, the combination of Zaharchuk in view of Schlemper teaches a prediction model that generates predicted representations using first and second reference representations in the frequency space.
The underlined amended portion above further limits that the first and second reference representations each represent a frequency region of the frequency space indicative of contrast information associated with the examination region during their respective time spans. As described above, and in further detail in the rejection below, Zaharchuk teaches wherein the first and second reference representations represent the examination region during different time spans, and representing different types of contrast. Schlemper is relied upon to teach wherein the reference representations represent a frequency region of the frequency space (k-space data in the spatial frequency domain, see para 419). Certain regions in the frequency space, or MRI k-space, correspond to the contrast of the image when transformed to the image domain, as demonstrated by ThoracicKey in pg. 2, FIG I7-1: “The peripheral portions of k-space (right two columns) primarily provide information about fine details and edges, while the overall image contrast information is contained in the central portions of k-space (left two columns)”. MRIquestions (see cited pertinent art), further demonstrates this concept when utilizing a contrast agent: “When gadolinium contrast passes through the anatomy of interest, only the general brightness or darkness change significantly. This dynamic information can be obtained from sampling the center of k-space only.” Therefore, by operating with MRI data in the frequency domain, as taught Schlemper, representations of information indicative of contrast information and representing a contrast of a region is disclosed.
Accordingly, since Schlemper teaches that the input and output of the prediction model may be in the spatial-frequency domain, each reference representation represents the frequency region of the frequency space indicative of the contrast information associated with the examination region during a respective time span. This is because the data in the frequency domain, corresponding to the MRI image, contains regions that indicate contrast information. Therefore, Zaharchuk in view of Schleper teaches the limitations of amended claim 1, described in further detail in the rejection below.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 2 and 3 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 2 recites the limitation “before the administration of the contrast agent” and claims 2 and 3 each recite the limitation “after the administration of the contrast agent”. Because claim 1 recites “one or more administrations of a respective contrast agent”, there may be more than one contrast agent administered and more than one administration actions. Thus, it is unclear if before and after the “administration of the contrast agent” in claims 2 and 3 is before/after the administration of one contrast agent or multiple contrast agent administrations. For examination purposes, the limitations will be interpreted to be before/after at least one administration of a contrast agent.
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, 6-8, and 11-13, and 15 are rejected under 35 U.S.C. 103 as being unpatentable over Zaharchuk et al. (U.S. Patent No. 2019/0108634 A1), hereinafter Zaharchuk, in view of Schlemper et al. (U.S. Patent No. 2020/0294287 A1), hereinafter Schlemper.
Regarding claim 1, Zaharchuk teaches a method comprising:
receiving a plurality of first representations of an examination region of an examination object (Zaharchuk, para 24: “The input to the deep learning network is a zero-contrast dose image 100 and low-contrast dose image 102”; region is an area of the body, shown here as the brain in para 24 and FIG 5; object is the subject, see last sentence of para 50), wherein at least some of the first representations represent the examination region during a first time span after one or more administrations of a respective contrast agent (Zaharchuk, para 24: “low-contrast dose image”; the first time span is defined as the time after administration of the low-dose contrast agent; FIG 2 demonstrates this workflow protocol in sequential order, see also para 25);
feeding the plurality of first representations to a prediction model (Zaharchuk, para 24: “deep learning network”), wherein the prediction model has been trained using first reference representations of the examination region of a multiplicity of examination objects (Zaharchuk, para 24: “A deep learning network is trained using multi-contrast images 100, 102, 104 acquired from scans of a multitude of subjects”) to generate, from the first reference representations, of which at least some represent the examination region during the first time span after an administration of a contrast agent (Zaharchuk, training images include low-contrast dose images, para 25: “FIG. 2 illustrates the workflow of the protocol and procedure for acquisition of images used for training”), one or more second reference representations which represent a contrast of the examination region (Zaharchuk, full-dose image 312 in FIG 3, para 28: “The resulting images then are used in a deep learning training stage 304 for training a deep learning network to synthesize a full-dose image 312”; full-dose image represents increased contrast of the examination region compared to no-/low-dose images) during a second time span (Zaharchuk, the second time span is defined as the time after administration of the full-dose contrast agent; see FIG 2 and para 25);
receiving one or more predicted representations of the examination region from the prediction model, wherein the one or more predicated representations represent the examination region during the second time span (Zaharchuk, para 24: “the output of the network is a synthesized prediction of a full-contrast dose image 116”); and
outputting the one or more representations of the examination region in real space (Zaharchuk, para 50: “the deep learning network (DLN) which then generates as output a synthesized full-dose contrast agent image of the subject”; see FIG 6 wherein the synthesized image is in the real image space).
Zaharchuk fails to explicitly teach: wherein the method is a computer-implemented method; wherein the plurality of first representations is in frequency space; at least some of the first reference representations are in frequency space; one or more second reference representations are in frequency space; wherein each first reference representation of the examination region in frequency space represents a frequency region of the frequency space indicative of contrast information associated with the examination region during the first time span, wherein each second reference representation represents the frequency region of the frequency space indicative of the contrast information associated with the examination region during the second time span; receiving one or more predicted representations of the examination region in frequency space (emphasis added); and transforming the one or more predicted representations into one or more representations of the examination region in real space.
Schlemper teaches a similar system (Schlemper, abstract: “generating magnetic resonance (MR) images from MR data obtained by a magnetic resonance imaging (MRI) system”) wherein the method is a computer-implemented method (Schlemper, para 447: “FIG. 26 is a diagram of an illustrative computer system on which embodiments described herein may be implemented”); wherein the plurality of first representations (data input to the neural network) is in frequency space (Schlemper, para 83: “The neural networks described herein may be configured to operate on data in any suitable domain. For example, one or more of the neural networks described herein may be configured to receive as input, data in the “sensor domain”, “spatial-frequency domain” (also known as k-space)”); at least some of the first reference representations (training data input to the neural network) are in frequency space (Schlemper, see last citation); one or more second reference representations (data output from the neural network) are in frequency space (Schlemper, para 85: “In some embodiments, the output of a neural network may be in the same domain as its input”; see para 83 wherein input data is in the spatial-frequency domain); receiving one or more predicted representations of the examination region (data output from the neural network) in frequency space (Schlemper, see last citation); and transforming the one or more predicted representations into one or more representations of the examination region in real space (Schlemper, para 83: “Image-domain data may be obtained by applying an inverse Fourier transformation (e.g., an inverse fast Fourier transform if the samples fall on a grid) to k-space data”; para 84: “the neural network may be configured to receive as input data in the target domain, and after completing processing, the output may be transformed back to the source domain”).
Accordingly, since Schlemper teaches wherein the first and second reference representations represent the examination region in frequency space, Schlemper further teaches: wherein each first reference representation of the examination region in frequency space represents a frequency region of the frequency space indicative of contrast information associated with the examination region during the first time span, wherein each second reference representation represents the frequency region of the frequency space indicative of the contrast information associated with the examination region during the second time span (Schlemper, contrast information of the examination region is represented by regions of the k-space in the frequency domain, see above citations).
Schlemper teaches a similar method to Zaharchuk’s method wherein MRI images are input to a neural network and generated images are outputted. While Zaharchuk teaches processing steps of image acquisition and machine learning, implementation of the method on a computer is not explicitly taught. One of ordinary skill in the art, before the effective filing date of the claimed invention, could have combined the computer of Schlemper with the method of Zaharchuk using known methods. In doing so, each element merely would have performed the same functions as it did separately and would achieve the predictable results of increasing the efficiency of image processing while decreasing the manual burden of the operator.
Additionally, Zaharchuk teaches a method wherein images are processed, with a neural network, in the image domain. Schlemper applies a known technique of converting MRI image data to k-space frequency data, and vice versa, specifically teaching a neural network wherein the images can be processed in either the image or frequency domain. One of ordinary skill in the art, before the effective filing date of the claimed invention, would have recognized that applying the neural network processing of the representations in the frequency domain, as taught by Schlemper, with the method of Zaharchuk would have yielded predictable results and resulted in an improved generated image by allowing the model to learn how different image properties are represented in the k-space (Similar to k-space weighting during reconstruction, Schlemper, para 419: “The inventors have appreciated that the neural network techniques described herein may be improved if the input MR spatial frequency data were weighted in the spatial frequency domain (k-space). In particular, the inventors have appreciated that weighting input MR spatial frequency data in k-space prior to reconstruction may improve the quality of the reconstruction”; para 421: “training the neural network may include one or more terms to guide the type of weighting that is to be learned (e.g., to weight more near the k-space origin, away from the k-space origin, near a particular region of k-space, or in any other suitable way”). Since certain areas of the MRI k-space representation correspond to certain image properties, the trained model can learn and manipulate these properties in the frequency domain.
Regarding claim 2 (dependent on claim 1), Zaharchuk in view of Schlemper teaches wherein the plurality of first representations comprises: at least one representation of the examination region in frequency space (Representations in freq. space taught in combination with Schlemper, see claim 1) that represents the examination region before the administration of the contrast agent (Zaharchuk, para 24: “zero-contrast dose image 100”; see FIG 2 attached below, see also para 25); and
at least one representation of the examination region in frequency space that represents the examination region in the first time span after the administration of the contrast agent (Zaharchuk, para 24: “low-contrast dose image”; see FIG 2 attached below, see also para 25), wherein the second time span comes after the first time span (Zaharchuk, para 24: “the output of the network is a synthesized prediction of a full-contrast dose image 116”; predicted images are for a full-contrast dose image which occurs after the first time span, according to para 25 and FIG 2).
PNG
media_image2.png
323
498
media_image2.png
Greyscale
Regarding claim 6 (dependent on claim 1), Zaharchuk in view of Schlemper teaches wherein the prediction model comprises an artificial neural network (Zaharchuk, para 24: “convolutional neural network”).
Regarding claim 7 (dependent on claim 1), Zaharchuk in view of Schlemper teaches wherein the first representations of the examination region in frequency space (Representations in freq. space taught in combination with Schlemper, see claim 1) are k-space data of a magnetic resonance imaging examination (Schlemper, para 79: “MRI system”; para 83: “input, data in the “sensor domain”, “spatial-frequency domain” (also known as k-space)”).
Regarding claim 8 (dependent on claim 1), Zaharchuk in view of Schlemper teaches comprising: receiving a plurality of radiological images of the examination region in real space (Zaharchuk, see pre-contrast and low-dose images in the real space in FIG 6); and converting the received plurality of radiological images of the examination in real space into the first representations of the examination region in frequency space using Fourier transforms (Schlemper, para 146: “transforming the input image represented by xi to the spatial frequency domain using a non-uniform Fourier transformation”; see combination with the images of Zaharchuk in claim 1).
Regarding claim 11, Zaharchuk teaches a system (Zaharchuk, FIG 3-5, para 24-25) configured to perform the claimed steps. All claim limitations are met and rendered obvious by Zaharchuk in view of Schlemper because the steps of claim 11 are the same as the method steps of claim 1. Schlemper further teaches one or more processors to perform the method steps (Schlemper, para 447: “The computer system 2600 may include one or more processors 2610”).
Regarding claim 12, all claim limitations are met and rendered obvious by Zaharchuk in view of Schlemper because the executed steps of claim 12 are the same as the method steps of claim 1. Schlemper further teaches a non-transitory computer readable storage medium storing instructions that, when executed by one or more processors of a computer system, cause the computer system to perform the method steps (Schlemper, para 447: “the processor 2610 may execute one or more processor-executable instructions stored in one or more non-transitory computer-readable storage media (e.g., the memory 2620), which may serve as non-transitory computer-readable storage media storing processor-executable instructions for execution by the processor 2610”).
Regarding claim 13, Zaharchuk teaches a method for predicting at least one radiological image with use of a contrast agent (Zaharchuk, para 24: “synthesized prediction of a full-contrast dose image 116”; contrast agent is used to obtain low-dose image, para 25: “a low dose (e.g., 10%) of contrast is administered and a low-dose image 202 is acquired”), wherein the method comprises: administering the contrast agent, wherein the contrast agent spreads in an examination region of an examination object (Zaharchuk, see para 25 citation above and FIG 2). All further claim limitations are met and rendered obvious by Zaharchuk in view of Schlemper because the method steps of claim 13 are the same as the method steps of claim 1.
Regarding claim 15 (dependent on claim 12), Zaharchuk in view of Schlemper teaches a kit (Elements to perform the method steps taught by Zaharchuk in view of Schlemper) comprising the contrast agent (Zaharchuk, administered contrast agent, see para 25) and the non-transitory computer readable storage medium of claim 12 (See claim 12 rejection).
Claims 3-4 are rejected under 35 U.S.C. 103 as being unpatentable over Zaharchuk in view of Schlemper, in further view of Kudo et al. (U.S. Patent No. 2022/0198734 A1), hereinafter Kudo.
Regarding claim 3 (dependent on claim 1), Zaharchuk in view of Schlemper fails to teach wherein: the plurality of first representations comprises at least two representations of the examination region in frequency space that represent the examination region in the first time span after the administration of the contrast agent; and the second time span comes before the first time span.
However, Kudo teaches a similar method (Kudo, para 58: “image generation model”) wherein: the plurality of first representations comprises at least two representations of the examination region in frequency space that represent the examination region in the first time span after the administration of the contrast agent (Kudo, input two images to model, para 106: “it is possible to increase the information of the image that is the source for deriving the virtual image V0 by inputting the plurality of target images Gi to the image generation device 1”; para 89: “in a case in which at least one target image Gi having any representation format and the target information A0 representing the representation format of the presence or absence of the contrast medium are input to the image generation device 1, the virtual image V0 having the representation format of contrast or non-contrast is generated”; see para 87 wherein a post-contrast image is a representation format); and the second time span comes before the first time span (para 87: “virtual images V0 having the converted representation format…a T1 non-contrast image”; see the top row of FIG 9 where a contrast image is input to the model and a T1 non-contrast image is output).
While Zaharchuk teaches wherein multiple images after the administration of a contrast agent may be used to train the prediction model (Zaharchuk, para 49: “It is also possible to base the training on more than two images having different dose levels, and it is not necessary to include a zero-dose image”), it is not explicitly disclosed wherein these images are fed to the prediction model as the first representations. Zaharchuk teaches wherein the first time span is defined by after administration of the contrast agent, and the absence of a contrast agent occurs before the first time span. Kudo teaches wherein input representations corresponding to the first time span may result in an output of non-contrast predicted representations, which correspond to before the first time span. It would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to have combined the output of non-contrast images using post-contrast input images, taught by Kudo, with the method of Zaharchuk in order to shorten the time of the imaging protocol the patient has to endure, by decreasing the number of images physically captured (Kudo, “The image generation device 1 including the learning device according to the present embodiment is a computer…The computer may be a workstation or a personal computer directly operated by a doctor who makes a diagnosis, or a server computer connected to the workstation or the personal computer via the network”). If a medical patient is undergoing contrast imaging, this method reduces the need to capture any non-contrast images.
Regarding claim 4 (dependent on claim 1), Zaharchuk in view of Schlemper fails to teach wherein the plurality of first representations comprises: at least one representation of the examination region in frequency space that represents the examination region in the first time span after an administration of a first contrast agent; and at least one representation of the examination region in frequency space that represents the examination region in the first time span after an administration of a second contrast agent, wherein the second contrast agent was administered after the first contrast agent, wherein the second time span comes before the first time span.
However, Kudo teaches a similar method (Kudo, para 58: “image generation model”) comprising: at least two representations of the examination region in frequency space (emphasis added; Representations in freq. space taught in combination with Schlemper, see claim 1) that represents the examination region in the first time span after an administration of a first contrast agent (Kudo, input two images to model, para 106: “it is possible to increase the information of the image that is the source for deriving the virtual image V0 by inputting the plurality of target images Gi to the image generation device 1”; para 89: “in a case in which at least one target image Gi having any representation format and the target information A0 representing the representation format of the presence or absence of the contrast medium are input to the image generation device 1, the virtual image V0 having the representation format of contrast or non-contrast is generated”; see para 87 wherein a post-contrast image is a representation format); and wherein the second time span comes before the first time span (para 87: “virtual images V0 having the converted representation format…a T1 non-contrast image”; see the top row of FIG 9 where a contrast image is input to the model and a T1 non-contrast image is output).
While Zaharchuk teaches wherein multiple images after the administration of a contrast agent may be used to train the prediction model (Zaharchuk, para 49: “It is also possible to base the training on more than two images having different dose levels, and it is not necessary to include a zero-dose image”), it is not explicitly disclosed wherein these images are fed to the prediction model as the first representations. Zaharchuk teaches wherein the first time span is defined by after administration of the contrast agent, and the absence of a contrast agent occurs before the first time span. Kudo teaches wherein input representations corresponding to the first time span may result in an output of non-contrast predicted representations, which correspond to before the first time span. It would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to have combined the output of non-contrast images using post-contrast input images, taught by Kudo, with the method of Zaharchuk in order to shorten the time of the imaging protocol the patient has to endure, by decreasing the number of images physically captured (Kudo, “The image generation device 1 including the learning device according to the present embodiment is a computer…The computer may be a workstation or a personal computer directly operated by a doctor who makes a diagnosis, or a server computer connected to the workstation or the personal computer via the network”). If a medical patient is undergoing contrast imaging, this method reduces the need to capture any non-contrast images.
Additionally, Zaharchuk teaches the acquisition of image representations of the examination region after the administration of a first contrast agent and image representations of the examination region after the administration of a second contrast agent, wherein the second contrast agent was administered after the first contrast agent (Zaharchuk, para 28: “In acquisition stage 300, scans are performed as described above to acquire a zero-contrast image 306, low-contrast image 308, and full-contrast image 310”; see FIG 2 wherein the contrast agents are different and the full-contrast is administered after the low-contrast). Although Zaharchuk teaches the acquisition of multi-contrast images to be used for training the model, the use of multiple contrast agents to generate different MRI images is a known technique. It would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to have utilized the multi-contrast images as the target images in the teachings of Kudo in order to improve the image predication output by increasing the amount of relevant input images.
Claim 5 is rejected under 35 U.S.C. 103 as being unpatentable over Zaharchuk in view of Schlemper, in further view of Ichihara et al. (U.S. Patent No. 2006/0241402 A1), hereinafter Ichihara.
Regarding claim 5 (dependent on claim 1), Zaharchuk in view of Schlemper fails to teach wherein the one or more predicted representations represent the examination region in the second time span with a contrast enhancement that is constant over time.
However, Ichihara teaches an image generation system and method (Ichihara, para 2) wherein the one or more predicted representations represent the examination region in the second time span with a contrast enhancement that is constant over time (Ichihara, para 45: “in a state where it can be considered that a concentration Ca(t) of the contrast medium within the coronary artery and a concentration Cmyo(t) of the contrast medium within a myocardial region are saturated in a constant value”). It would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to have combined the above teaching of Ichihara with the method of Zaharchuk in order to realistically model the results of contrast injection (Ichihara, see all of para 44, specifically, para 44: “the condition of injecting the contrast medium is adjusted using the contrast-medium injector 8 such that change-of-time of the concentration Ca(t) of the contrast medium within the coronary artery becomes constant”).
Claims 9-10 are rejected under 35 U.S.C. 103 as being unpatentable over Zaharchuk in view of Schlemper, in further view of Galea et al. (Galea, R. R., Diosan, L., Andreica, A., Popa, L., Manole, S., & Balint, Z. (2021). Region-of-interest-based cardiac image segmentation with deep learning. Applied Sciences, 11(4), 1965.), hereinafter Galea, and Thoracic Key, Posted by drzezo, More k-Space: The Relationship between k-Space and the Image. 7 June 2016 [online], [retrieved on 2025-11-20]. Retrieved from the Internet <URL: https://thoracickey.com/more-k-space-the-relationship-between-k-space-and-the-image/ >, hereinafter ThoracicKey.
Regarding claim 9 (dependent on claim 1), Zaharchuk in view of Schlemper teaches wherein the one or more predicted representations represent the examination region during the second time span; transforming the one or more predicted representations into the one or more representations of the examination region in real space; and outputting the one or more representations of the examination region in real space (See claim 1 rejection).
Zaharchuk in view of Schlemper fails to teach comprising: specifying a region in the plurality of first representations of the examination region, wherein the specified region comprises the center of the frequency space; reducing the first representations to the specified region; feeding the plurality of reduced first representations to the prediction model; receiving the one or more predicted representations of the examination region in frequency space from the prediction model, wherein the one or more predicted representations represent the examination region during the second time span; supplementing the one or more predicted representations by one or more regions of the received first representations that lie outside the specified region; transforming the one or more supplemented predicted representations into the one or more representations of the examination region in real space; and outputting the one or more representations of the examination region in real space (emphasis added).
However, Galea teaches a method for specifying a region in the plurality of first representations of the examination region; reducing the first representations to the specified region (Galea, extracted ROI, pg. 5, section 4.3.1; see extracted ROI in FIG 1 attached below); feeding the plurality of reduced first representations to the prediction model (Galea, ROI is fed to a segmentation model, pg. 5, FIG 1: “a ROI is cropped (by using the corresponding ground-truth) and resized to 224 × 224. The segmentation model is trained on these processed images and a set of masks are predicted for each input”); receiving one or more predicted representations of the examination region from the prediction model (Galea, final prediction image, pg. 5, FIG 1); and supplementing the one or more predicted representations by one or more regions of the received first representations that lie outside the specified region (Galea, pg. 5, FIG 1: “Finally, the masks are re-scaled to the original dimensions and re-inserted in the original image”).
PNG
media_image3.png
364
686
media_image3.png
Greyscale
It would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to have combined the specified region image processing of Galea with the method of Zaharchuk in order to improve the performance of the machine learned model (Galea, pg. 9, section 6: “using ROI localization improved performance”).
Further, ThoracicKey teaches a specified region comprising the center of the frequency space (ThoracicKey, pg. 2, FIG I7-1 attached below). It would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to have combined specified region image processing of Galea, wherein the specified region comprises the center of the frequency space, as taught by ThoracicKey, with the method of Zaharchuk in order to individually alter the contrast of the image (ThoracicKey, pg. 2, FIG I7-1: “The peripheral portions of k-space (right two columns) primarily provide information about fine details and edges, while the overall image contrast information is contained in the central portions of k-space (left two columns)”).
PNG
media_image4.png
586
908
media_image4.png
Greyscale
Regarding claim 10 (dependent on claim 9), Zaharchuk in view of Schlemper, Galea, and ThoracicKey teaches wherein the one or more supplemented predicted representations are transformed into the one or more representations of the examination region in real space using inverse Fourier transforms (Schlemper, para 83: “Image-domain data may be obtained by applying an inverse Fourier transformation”; see also para 84).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
MRIquestions.com. Dynamic Imaging. 23 May 2020, Internet Archive [online], [retrieved on 2026-04-27]. Retrieved from the Internet <URL: https://web.archive.org/web/20200523182612/https:/mriquestions.com/dynamic-ce-imaging.html>
PNG
media_image5.png
765
711
media_image5.png
Greyscale
Loecher, M., Wieben, O. (2015). k-Space. In: Syed, M., Raman, S., Simonetti, O. (eds) Basic Principles of Cardiovascular MRI. Springer, Cham. https://doi.org/10.1007/978-3-319-22141-0_2
Bustamante, M., Viola, F., Carlhall, C. J., & Ebbers, T. (2021). Using deep learning to emulate the use of an external contrast agent in cardiovascular 4D flow MRI. Journal of Magnetic Resonance Imaging, 54(3), 777-786.
Gong, E., Pauly, J. M., Wintermark, M., & Zaharchuk, G. (2018). Deep learning enables reduced gadolinium dose for contrast-enhanced brain MRI. Journal of magnetic resonance imaging, 48(2), 330-340.
Narayana, P. A., Coronado, I., Sujit, S. J., Wolinsky, J. S., Lublin, F. D., & Gabr, R. E. (2020). Deep learning for predicting enhancing lesions in multiple sclerosis from noncontrast MRI. Radiology, 294(2), 398-404.
Montalt-Tordera, J., Quail, M., Steeden, J. A., & Muthurangu, V. (2021). Reducing contrast agent dose in cardiovascular MR angiography with deep learning. Journal of Magnetic Resonance Imaging, 54(3), 795-805.
U.S. Patent No. 2014/0376794 A1
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to whose telephone number is (571)272-1179. The examiner can normally be reached M-F 9-5 EST.
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, ANDREW BEE can be reached at (571) 270-5183. 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.
/EMMA E DRYDEN/Examiner, Art Unit 2677 /ANDREW W BEE/Supervisory Patent Examiner, Art Unit 2677
Prosecution Insights