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 .
Response to Arguments
Applicant’s arguments, see Remarks at page 9, filed 2 December 2025, with respect to the rejections of claims 17, 90, and 20 under 35 U.S.C. 112(b) have been fully considered and are persuasive. The rejections have been withdrawn.
Applicant’s arguments, see Remarks at pages 9-13, filed 2 December 2025, with respect to the rejection of claims 1-3 and 5-21 under 35 U.S.C. 103 have been fully considered but they are not persuasive. Applicant argues Seth and Liu, whether alone or in combination, does not teach every feature of claims 1, 19, and 20. Examiner respectfully disagrees. As explained in the rejection of claim 5 under 35 U.S.C. 103 in the previous Office action, Liu discloses removing a feature of interest in an inpainting process. As explained in paragraph 30, Liu discloses using a lung image without COVID (i.e., lesion-free) and synthesizing abnormality patterns (lesions) into a removed (masked) region of the lung image to generate training data including a synthetic lung image showing a lesion. The combination of Seth and Liu, as provided below, teaches every feature of claims 1, 19, and 20. Accordingly, the rejection is maintained.
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.
Claim 5 is rejected under 35 U.S.C. 112(b) as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor regards as the invention.
Claim 5 recites, in part, “the patch of the medical image” (emphasis added). Claim 1 includes “a removed patch”, “a further patch of the medical image”, “a patch of a reference image”, “a lesion-free patch of the medical image” and “a lesion-free patch of the reference image”. Since claim 5 also recites “the patch of the reference image” and “the removed patch”, in context, “the patch” cannot refer to “the patch of the reference image” or “the removed patch”. However, it is unclear whether the antecedent basis of “the patch of the medical image” (emphasis added) is “a further patch of the medical image”, “a lesion-free patch of the medical image”, or “a lesion-free patch of the reference image”. For purposes of applying prior art, the Examiner assumes “the patch” in claim 5 refers to “a further patch of the medical image” in claim 1.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-3 and 5-21 are rejected under 35 U.S.C. 103 as being unpatentable over U.S. Pat. Appl. Pub. No. 20220167947 to Seth et al. (hereinafter “Seth”) in view of U.S. Pat. Appl. Pub. No. 20210327054 to Liu et al. (hereinafter “Liu”).
Regarding claim 1, Seth teaches a medical imaging system, comprising:
a data storage resource (Seth, par. [0138], “If a computer program is discussed above, it may be stored/distributed on a suitable medium, such as an optical storage medium or a solid-state medium supplied together with or as part of other hardware”) and probability data (Seth, par. [0112], “A probability map”) representing probability information for a location (Seth, par. [0112], “a given point”) of at least one feature of interest (Seth, par. [0112], “anatomical feature”) in an anatomical region (Seth, par. [0112], “anatomical feature occupying a given point within the first 3D ultrasound image and the one or more additional 3D ultrasound images.”); and
processing circuitry (Seth, par. [0093], “processor 30”) configured to:
receive medical image data (Seth, pars. [0094]-[0095], “FIG. 2 shows a method 100 for obtaining a composite 3D ultrasound image of a region of interest. The method begins in step 110 by obtaining preliminary ultrasound data from a region of interest of a subject.”) representing a medical image of the anatomical region (Seth, par. [0095], “The method begins in step 110 by obtaining preliminary ultrasound data from a region of interest of a subject.”; Scout data represents data obtained from subsequent scans for the same region.);
determine a presence of a particular feature of interest by processing the received medical image data (Seth, par. [0101], “selecting 2D imaging planes from the preliminary ultrasound data likely to contain an anatomical feature”; In step 130, a particular anatomical feature is detected.);
retrieve the probability data (Seth, par. [0107], “In step 150, spatial registration is performed between the first 3D ultrasound image and the one or more additional 3D ultrasound images based on the anatomical feature.”; par. [0120], “As discussed above, a probability map may be generated for the first 3D ultrasound image and the one or more additional 3D ultrasound images during the spatial registration step. This may be used as the basis for a 3D confidence map of the overlapping portion of 3D ultrasound images.”); and
process the medical image data to perform at least one image augmentation operation (Seth, par. [0121], “blending”) on the medical image represented by the received medical image data for the particular feature of interest (Seth, par. [0107], “anatomical feature”), but does not teach that which is explicitly taught by Liu.
Liu teaches a data storage resource configured (Liu, par. [0048], “memory or storage”) to store probability data (Liu, par. [0039], “spatial probability map”) and retrieving the probability data from the data storage resource (Liu, par. [0039], “Lesion center locations are sampled from the spatial probability map and the sampled locations are mapped to the corresponding image space of the synthesized segmentation mask.”);
wherein the at least one image augmentation operation comprises:
removing the particular feature of interest from the medical image to form a removed patch (Liu, par. [0041], “At step 106, the input medical image is masked based on the synthesized segmentation mask. The masked input medical image includes one or more unmasked portions and one or more masked portions, as defined by the synthesized segmentation mask. The masked portions are filled with uniform noise with values between, e.g., [ -1, 1].”; Filling the mask portions of the input medical image removes the underlying image data, thereby forming a removed patch of image data ready for replacement by synthesized data.; par. [0053], “training images 802 may be of any suitable modality.”) and performing an image data reconstruction process on the removed patch using image data obtained from a further patch of the medical image (unmasked region) or image data from a patch of a reference image (The unmasked region is also a patch that serves as a reference for the boundaries of the region that is inpainted. See Liu at par. [0043], “The initial synthesized medical image includes a synthesized version of the unmasked portions of the masked input medical image and synthesized patterns (e.g., abnormality mappers associated with the disease) in the masked portions of the masked input medical image. The synthesized version of the unmasked portions of the masked input medical image may be synthesized by regenerating the unmasked portions or by copying imaging data of the unmasked portions from the masked input medical image.”),
wherein the image data reconstruction process includes a region inpainting process of inpainting the removed patch, and the region inpainting process uses image data representing a lesion-free patch of the medical image or a lesion-free patch of the reference image to inpaint the removed patch (Liu, par. 31, “Embodiments described herein provide for generating synthesized medical images depicting abnormality patterns associated with COVID-19. Such abnormality patterns associated with COVID-19 are inpainted on medical images of lungs of patients without COVID-19 using a machine learning based generator network”. The inpainting process described by Liu inserts image data representing a lesion into an image that is lesion-free (i.e., without features indicative of COVID-19). The processes uses both the image data of the original image and the image data of the inserted lesion. The original patch represents a lesion-free patch before it is replaced with a patch representative of a lesion.).
Seth discloses a technique of registering ultrasound images using a computer (Seth, par. [0040]) and a storage (Seth, par. [0138], “medium”) storing a program (Seth, par. [0138], “computer program”) to implement the technique. Seth further discloses generating a probability map representing anatomical features in a region of interest in the ultrasound images (See pars. [0064] and [0112]). Seth also discloses “a need for a means of acquiring an accurate and robust composite ultrasound image” due to “existing techniques often [resulting] in an inaccurate view of the region of interest” (pars. [0005]-[0006]). Thus, Seth shows that it was known in the art before the effective filing date of the claimed invention to composite medical images together to form a larger and more comprehensive data set depicting a region of interest, which is analogous to the claimed invention in that it is pertinent to the problem being solved by the claimed invention, generating datasets for machine learning training. Liu discloses a technique of blending (Liu, par. [0047]) ultrasound images (Liu, par. [0034]) using a computer (Liu, par. [0033], “computer 1702”) and storage (Liu, par. [0106], “storage devices”) storing a program to implement the technique. Liu further discloses generating a probability map (Liu, par. [0039], “spatial probability map”) representing anatomical features in the ultrasound images of a lesion (Liu, par. [0039], “Lesion center locations”) and synthesizing a masked region (Liu, par. [0043], “masked portions”) of a medical image for inserting a synthesized feature of interest (Liu, par. [0032]). Liu also discloses “Machine learning based systems for automatically assessing COVID-19 based on such features would be useful. However, due to the novelty of COVID-19, the availability of appropriate imaging data and annotations for training such machine learning based systems is limited.” (Liu, par. [0005]). Thus, Liu shows that it was known in the art before the effective filing date of the claimed invention to composite medical images together to form a larger and more comprehensive data set depicting a region of interest, which is analogous to the claimed invention in that it is pertinent to the problem being solved by the claimed invention, generating datasets for machine learning training.
A person of ordinary skill in the art would have been motivated to modify the image compositing method and storage of Seth to store probability data for retrieval in an inpainting-based machine learning process to remove image data of an area of one image and synthesize data from another image into the removed area as disclosed by Liu to thereby generate synthetic training images of lesions inpainted onto healthy images or conversely, generate synthetic healthy images with healthy patches inpainted onto lesion images, and expedite the training process by placing the probability data in an easily-accessible location for compositing medical images including a lesion region, a healthy region, or a particular pathology. Based on the foregoing, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have made such modification according to known methods to yield the predictable results to have the benefit of retaining the probability data for multiple iterations of a machine learning process and expediting the production of training data for improving a machine learning process.
Regarding claim 2, Seth in view of Liu teaches the system of claim 1, wherein the processing circuitry is further configured to determine an augmentation position (Seth - a region to be blended) in the medical image based on the probability data for the location and the augmentation operation comprises adding the particular feature of interest to the image at the augmentation position (Seth, par. [0103], “The acquisition of these 3D ultrasound images based on anatomical context provides for more accurate fusion of the first 3D ultrasound image and the one or more additional 3D ultrasound images to form the composite 3D ultrasound image.”; Poisson blending is a combination of image features from multiple images, where features (e.g., gradients) of one of image are added to another.).
Regarding claim 3, Seth in view of Liu teaches the system of claim 2, wherein the augmentation operation further comprises performing an image blending process (Seth, par. [0121], “The blending may be performed using Poisson blending, which is a gradient domain image processing technique.”) for the particular feature of interest at the determined augmentation position (Seth, par. [0095], “region of interest of a subject.”).
Regarding claim 5, Seth in view of Liu teaches the system according to claim 1,
wherein the patch of the medical image and/or the patch of the reference image is selected based on a spatial relationship (Seth, pars. [0107]-[0108], “In step 150, spatial registration is performed between the first 3D ultrasound image and the one or more additional 3D ultrasound images based on the anatomical feature. In other words, the location and/or orientation of the first 3D ultrasound image is determined relative to the one or more additional 3D ultrasound images and vice versa.”; Multiple images have corresponding patches of shared anatomical features.) between the removed patch (Liu, par. [0041], “one or more masked portions”), the patch of the medical image (Liu, par. [0041], “one or more unmasked portions and one or more masked portions”), and the patch of the reference image (Liu, par. [0043], “The initial synthesized medical image includes a synthesized version of the unmasked portions of the masked input medical image and synthesized patterns (e.g., abnormality mappers associated with the disease) in the masked portions of the masked input medical image. The synthesized version of the unmasked portions of the masked input medical image may be synthesized by regenerating the unmasked portions or by copying imaging data of the unmasked portions from the masked input medical image.”)
The rationale for obviousness is the same as provided for claim 1.
Regarding claim 6, Seth in view of Liu teaches the system according to claim 1, wherein the particular feature of interest comprises at least one of a lesion (Liu, par. [0039], “Lesion center locations are sampled from the spatial probability map and the sampled locations are mapped to the corresponding image space of the synthesized segmentation mask.”), a pathology (Liu, par. [0039], “Many diseases, such as, e.g., COVID-19, typically present with abnormality patterns in subpleural, peripheral, bilateral, and multilobar locations. Accordingly, in one embodiment, to simulate the spatial distribution of such abnormality patterns, a spatial probability map of the abnormality patterns is computed using aligned, manually annotated images”) or an artefact.
The rationale for obviousness is the same as provided in claim 1.
Regarding claim 7, Seth in view of Liu teaches the system according to claim 1, wherein at least one:
the probability information for location is represented by a probability map (Seth, par. [0120], “a probability map”);
the probability information is representative of a spatial probability density across the anatomical region (Seth, par. [0112], “A probability map may be generated based on the 3D point cloud, the probability map representing a confidence value of the anatomical feature occupying a given point within the first 3D ultrasound image and the one or more additional 3D ultrasound images.” A density, under the broadest reasonable interpretation, is a quantity of things in a given space. The probability map provides confidence values of three-dimensional points in a volume/space being anatomical features. Thus, the collection of confidence values of the three-dimensional points in the given space is a spatial probability density.); or
the probability information is representative of anatomical information and/or anatomical knowledge of the at least one feature of interest in the anatomical region (Seth, par. [0112], “the probability map representing a confidence value of the anatomical feature occupying a given point within the first 3D ultrasound image and the one or more additional 3D ultrasound images.”).
Regarding claim 8, Seth in view of Liu teaches the system according to claim 1, wherein the method further comprises determining an augmentation feature parameter for the at least one feature of interest, wherein the augmentation feature parameter comprises at least one of: size, scale, type (Seth, par. [0110], “The 3D point cloud generation may be performed as follows. A specific anatomy model, such as a cardiac model or a fetal model, may be used to drive a classification module for identifying slices/volumes of interest within the 3D ultrasound images.”; par. [0111], “A segmentation module may then be used to generate an anatomy driven heat map on the slices/volumes selected by the classification module to produce a 3D point cloud within the 3D ultrasound images. This bi-directional approach reduces the search space needed for point cloud generation, thereby increasing the computational efficiency of the method.”; par. [0112], “A probability map may be generated based on the 3D point cloud, the probability map representing a confidence value of the anatomical feature occupying a given point within the first 3D ultrasound image and the one or more additional 3D ultrasound images.”; par. [0113], “Confidence values derived from the above mentioned process may be used to create a 3D confidence map of the anatomical feature, which may be used as a deciding factor to blend the volumes.”; A 3D confidence map includes parameters of augmentation for image features of a type of anatomy or specific anatomy model.) or subtype, wherein the augmentation feature parameter is selected using the probability information associated with the augmentation feature parameter (Seth, par. [0113], “Confidence values derived from the above mentioned process may be used to create a 3D confidence map of the anatomical feature, which may be used as a deciding factor to blend the volumes.”; Confidence values are information indicating the probability/likelihood of a correct feature identification. The 3D confidence map is selected/generated using the confidence values.).
Regarding claim 9, Seth in view of Liu teaches the system according to claim 1, wherein the processing circuitry is further configured to obtain reference image data representing one or more images of at least a part of the anatomical region (Seth, par. [0112], “anatomical feature occupying a given point within the first 3D ultrasound image and the one or more additional 3D ultrasound images.”) and determine a geometrical mapping between the part of the anatomical region and a corresponding part of the medical image (Seth, par. [0064], “A first 3D ultrasound image is obtained from the first imaging position and one or more additional 3D ultrasound images are obtained from the one or more additional imaging positions, wherein a portion of the first 3D ultrasound image overlaps a portion of the one or more additional 3D ultrasound images, thereby forming an overlapping portion comprising the anatomical feature.”); and
wherein the at least one image augmentation operation is based at least one on the determined geometrical mapping (Seth, par. [0064], “Spatial registration is performed between the first 3D ultrasound image and the one or more additional 3D ultrasound images based on the anatomical feature the 3D ultrasound images are then blended based on the spatial registration, thereby generating a composite 3D ultrasound image.”).
Regarding claim 10, Seth in view of Liu teaches the system according to claim 9, wherein at least one of:
the reference image data represents one or more anatomical atlases (Seth, par. [0064], “A first 3D ultrasound image is obtained from the first imaging position and one or more additional 3D ultrasound images are obtained from the one or more additional imaging positions, wherein a portion of the first 3D ultrasound image overlaps a portion of the one or more additional 3D ultrasound images, thereby forming an overlapping portion comprising the anatomical feature.”; An anatomical atlas, under the broadest reasonable interpretation, is a map of anatomical features. The ultrasound images use the overlap as a guide to form a composite image. Thus, an ultrasound image including an anatomical portion overlapping with another ultrasound image represents an anatomical atlas of the same modality that allows a geometrical mapping between the images.);
the reference image data represents one or more further reference images in the same modality (Seth, par. [0064], “A first 3D ultrasound image is obtained from the first imaging position and one or more additional 3D ultrasound images are obtained from the one or more additional imaging positions, wherein a portion of the first 3D ultrasound image overlaps a portion of the one or more additional 3D ultrasound images, thereby forming an overlapping portion comprising the anatomical feature.”);
the reference image data comprises a tissue volume or a set of tissue volumes (Seth, par. [0110], “The 3D point cloud generation may be performed as follows. A specific anatomy model, such as a cardiac model or a fetal model, may be used to drive a classification module for identifying slices/volumes of interest within the 3D ultrasound images.”); or
the reference image data are representative of a particular part of the anatomical region that is free of the at least one feature of interest (Seth at Fig. 3 shows how the images include the anatomical feature of interest surrounded by areas not including the anatomical feature of interest.).
Regarding claim 11, Seth in view of Liu teaches the system according to claim 1, wherein the at least one augmentation operation further comprises adding one or more features of interest to the medical image (Seth, par. [0103]; Poisson blending is a combination of image features, where features one of image are added to another.).
Regarding claim 12, Seth in view of Liu teaches the system according to claim 1, wherein the processing circuitry is further configured to obtain image data (Seth, par. [0115], “acquired 3D ultrasound images”) representative of the particular feature of interest and perform a transformation on the obtained image data to produce a transformed feature of interest for augmenting to the medical image (Seth, par. [0115], “a machine learning (and in particular, a deep learning) driven, anatomical probability based, 3D point cloud formation on the acquired 3D ultrasound images may be used to perform anatomy specific stitching using non-rigid registration techniques.”; Non-rigid registration transforms the structure of the image.).
Regarding claim 13, Seth in view of Liu teaches the system according to claim 1, wherein the at least one image augmentation operation comprises the image data reconstruction process and/or an image blending process (Seth, par. [0121], “The blending may be performed using Poisson blending, which is a gradient domain image processing technique.”), wherein the image data reconstruction process and/or the image blending process is performed in accordance with a pre-determined image data reconstruction procedure (Seth, par. [0121], “Poisson blending, which is a gradient domain image processing technique.”) and/or image blending procedure, selected based on at least one of a modality (Seth, par. [0121]; Seth discloses using Poisson blending for ultrasound image data. Thus, the Poisson blending procedure is selected for a modality.) and/or a type of the particular feature of interest and/or an anatomical location.
Regarding claim 14, Seth in view of Liu teaches system of claim 13, wherein the pre-determined procedure comprises a Poisson blending (Seth, par. [0121], “Poisson blending”) or other inpainting procedure, but does not teach that which is explicitly further taught by Liu.
Liu further teaches a procedure based on local blending (Liu, par. [0009], “the synthesized patterns are fused with the input medical image by blending the initial synthesized medical image with the input medical image to generate a blended image, smoothing boundaries of the synthesized segmentation mask to generate a smooth synthesized segmentation mask, cropping masked portions of the smooth synthesized segmentation mask from the blended image to extract the synthesized patterns, cropping unmasked portions of the smooth synthesized segmentation mask from the input medical image to extract remaining regions of the input medical image, and combining the extracted synthesized patterns and the extracted remaining regions.”; Blending in a masked area is local blending.).
The rationale for obviousness is the same as provided for claim 1.
Regarding claim 15, Seth in view of Liu teaches the system according to claim 1, wherein the processing circuitry is further configured to perform a transformation of the medical image based on at least one of a manual, rigid, or non-rigid registration (Seth, par. [0115], “a machine learning (and in particular, a deep learning) driven, anatomical probability based, 3D point cloud formation on the acquired 3D ultrasound images may be used to perform anatomy specific stitching using non-rigid registration techniques.”; Non-rigid registration transforms the structure of the image.) or based on a plurality of landmarks (Seth, par. [0106], “Common anatomical landmarks in the neighboring 3D volumes from overlapping regions are identified, which may then be used for translation correction in the spatial registration algorithm, based on the motion sensor data.”).
Regarding claim 16, Seth in view of Liu teaches the system according to claim 1, wherein the at least one image augmentation operation comprises applying one or more distortions to the particular feature of interest (Seth, par. [0121], “The blending may be performed using Poisson blending, which is a gradient domain image processing technique. Gradient domain image processing techniques operate on the differences between neighboring pixels, rather than on the pixel values directly”; Gradients of pixel values are changed, thereby distorted.), wherein the one or more distortions comprise geometrical distortions (Seth, par. [0115], “anatomy specific stitching using non-rigid registration techniques.”) or appearance distortions (Poisson blending changes the appearance of the image) or wherein the one or more distortions is based on the anatomical region (Seth, par. [0115], “anatomy specific stitching using non-rigid registration techniques.”), one or more properties of the medical image, a type, or a further property of the particular feature of interest.
Regarding claim 17, Seth in view of Liu teaches the system according to claim 1, wherein the processing circuitry is further configured to determine the probability information (Seth, par. [0112], “probability map”) for the location of the at least one feature of interest by processing medical image data representing a plurality of medical images of the anatomical region (Seth, par. [0112], “anatomical feature occupying a given point within the first 3D ultrasound image and the one or more additional 3D ultrasound images.”; The ultrasound images are medical images, for example as shown in FIG. 3.), wherein at least one of the plurality of medical images comprise the at least one feature of interest in the anatomical region (Seth, par. [0110], “The 3D point cloud generation may be performed as follows. A specific anatomy model, such as a cardiac model or a fetal model, may be used to drive a classification module for identifying slices/volumes of interest within the 3D ultrasound images.”).
Regarding claim 18, Seth in view of Liu teaches the system according to claim 1, wherein the method further comprises obtaining annotation data (Seth, par. [0101], “annotated model of a heart”) for the medical image associated with the at least one feature of interest (Seth, pars. [0100]-[0101], “In other words, the preliminary ultrasound data obtained during the scouting scan is used to derive anatomical information from the region of interest, which may then be used to guide the movement of an ultrasonic probe for capturing 3D, or 4D, ultrasound data based on the identified anatomical feature. This may, for example, be performed by selecting 2D imaging planes from the preliminary ultrasound data likely to contain an anatomical feature and selecting an anatomical feature that overlaps between two or more ultrasound probe positions for centering the 3D volumes to be acquired. This selection may be performed manually, by way of receiving a user input, or automatically through model-based best plane identification in 3D mode, for example, using an annotated model of a heart.”).
Claim 19 substantially corresponds to claim 1, reciting a method (Seth, par. [0064], “The invention provides for a method of obtaining a composite 3D ultrasound image of a region of interest.”) of augmenting an image, the method comprising steps corresponding to the functions of the “processing circuitry” (Seth, par. [0093], “processor 30”) recited in claim 1. Claim 19 is rejected for the same rationale for obviousness provided for claim 1.
Claim 20 substantially corresponds to claim 1, reciting a non-transitory computer-readable medium storing instructions (Seth, par. [0138], “If a computer program is discussed above, it may be stored/distributed on a suitable medium, such as an optical storage medium or a solid-state medium supplied together with or as part of other hardware”) that, when executed by processing circuitry (Seth, par. [0093], “processor 30”), cause the processing circuitry to perform the steps corresponding to the functions of the “processing circuitry” recited in claim 1. Claim 20 is rejected for the same rationale for obviousness provided for claim 1.
Regarding claim 21, Seth in view of Liu teaches the system of claim 3, wherein the image blending process comprises an inpainting process (Liu, par. [0051], “The synthesized images (in columns 708-714) in row 704 were generated by inpainting synthesized abnormality patterns on to lungs of a patient without COVID-19 using a synthesized segmentation mask.”).
The rationale for obviousness is the same as provided for claim 1.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Brain MR Atlas Construction Using Symmetric Deep Neural Inpainting to Xing et al. is pertinent to the inpainting recited in the independent claims because it discloses inpainting to replace tumor regions with normal values. See Abstract; and
U.S. Pat. Appl. Pub. No. 20150045651 to Crainiceanu et al. discloses “volumes are inpainted to fill the places where lesions were removed with the values expected in this area if it were occupied by normal, healthy tissue” in paragraph 81, which is pertinent to “the region inpainting process” that “uses image data representing a lesion-free patch” recited in the independent claims.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to RYAN P POTTS whose telephone number is (571)272-6351. The examiner can normally be reached M-F, 9am-5pm 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, Sumati Lefkowitz can be reached at 571-272-3638. 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.
/RYAN P POTTS/Examiner, Art Unit 2672
/SUMATI LEFKOWITZ/Supervisory Patent Examiner, Art Unit 2672