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 .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 05/29/2026 has been entered.
Priority
Acknowledgment is made of applicant’s claim for foreign priority under 35 U.S.C. 119 (a)-(d). The certified copy has been filed in parent Application No. JP2022-185040, filed on 11/18/2022.
Response to Amendment
Applicant’s Amendments filed 05/29/2026 have been entered and made of record.
Status of Claims
Currently pending Claim(s):
Amended claim(s):
Canceled claim(s):
New Claim(s):
1, 4, and 6-11
1, 7, and 8
2-3 and 5
9-11
Response to Arguments
This office action is responsive to Applicant’s Arguments/Remarks made in an Amendment received on 05/29/2026.
In view of applicant’s arguments, Remarks filed on 05/29/2026, with respect to independent claims 1, 7, and 8 under 35 U.S.C. 103, claim rejections have been fully considered but they are not persuasive. The Applicant argues on page 7:
PNG
media_image1.png
390
614
media_image1.png
Greyscale
The Examiner respectfully disagrees. Zhang discloses “the machine learning model is trained by using training data that includes third image data and fourth image data” at Training [section 2, page 1371 right column first full paragraph] "train a convolutional neural network for MAR. First, we generate metal-free, metal-inserted CT images to create a database." and Training [section 2, page 1371 right column paragraph 3] "...metal- inserted CT images, where beam hardening and Poisson noise are simulated. To ensure that the trained CNN works for real cases...we simulate the metal artifacts based on clinical CT images.". Further, Zhang discloses “the processing circuitry is further configured to reconstruct the third image data from the projection data and generate the fourth image data including a generated low-count artifact from the projection data” at page [1372 left column second to last paragraph] "The metal-free image is reconstructed using filtered backprojection (FBP), and the image is assumed as reference and denoted as xref.” and Experiments [section 4, page 1375 right column paragraph 1] "A 120 kVp x-ray source is simulated and each detector bin is expected to receive 2 x 107 photons in the case of blank scan [46]. There are 984 projection views over a rotation and 920 detector ins in a row. The distance between the x-ray source and the rotation center is 59.5 cm. The metal-free and metal-inserted images are reconstructed by the FBP from simulated sinograms and each image consists of 512 X 512 pixels.".
PNG
media_image2.png
269
586
media_image2.png
Greyscale
PNG
media_image3.png
66
612
media_image3.png
Greyscale
PNG
media_image4.png
104
604
media_image4.png
Greyscale
On page 7-8 of Remarks, the Applicant argues that:
Applicant’s arguments with respect to the pending claims have been considered but are moot because the new ground of rejection does not rely on any reference being used in the current rejection and the arguments are now rejected by newly cited art Tohme et al. (US 2021/0158486 A1) and Wang et al. (US 2017/0301066 A1).
Claim Objections
Claims 1, 7-8, and 10 objected to because of the following informalities:
Claim 1 line 11 “the projection data” lack antecedent basis. The Examiner suggests amending the phrase to “a projection data”.
Claim 7 lines 7-8, “the projection data” lack antecedent basis. The Examiner suggests amending the phrase to “a projection data”
Claim 8 lines 7-8, “the projection data” lack antecedent basis. The Examiner suggests amending the phrase to “a projection data”
Claim 10 line 3 seems to be missing a period at the end of the sentence. The Examiner suggest amending the claim to “…a zero clipping process with respect to a negative value of the projection data.”. Appropriate correction is required.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1 and 7-8 are rejected under 35 U.S.C. 103 as being unpatentable over Zhang et al. ("Convolutional neural network based metal artifact reduction in x-ray computed tomography." IEEE transactions on medical imaging 37.6 (2018): 1370-1381.) (hereinafter, “Zhang”) in view of Tohme et al. (US 2021/0158486 A1) (hereinafter, “Tohme”); and further in view of Wang et al. (US 2017/0301066 A1) (hereinafter, “Wang”).
Regarding claim 1, Zhang discloses a medical image processing apparatus, comprising:
processing circuitry configured to: acquire second image data in which a low-count artifact is reduced (Introduction [section 1, page 1370 right column last full paragraph] “…convolutional neural network (CNN) has been applied to medical imaging for low dose CT reconstruction and artifacts reduction”; Introduction [section 1, page 1371 left column paragraph 1] “the information from different correction methods is captured and the merits of these methods are fused, leading to a higher quality image (i.e. second image)”), by applying a trained machine learning model to first image data that is obtained by an X-ray CT scan (Experiments [section 4 page 1376 left column first full paragraph] “scanned on a Siemens SOMATOM Sensation 16 CT scanner with 120 kVp and 496 mAs using the helical scanning geometry. (i.e. first image)”), and
output image data based on the second image data (See figure 3 [page 1373 left column last full paragraph continued to right column first full paragraph] “Fig. 3 depicts the workflow of our CNN, which is comprised of an input layer, an output layer and L = 5 convolutional layers…nl convolution kernel with a fixed size of cl × cl . Cl (u) generates new feature maps based on the (l − 1)th layer’s output. For the last layer, feature maps are used to generate an image that is close to the target (i.e. image data).”),
wherein the machine learning model is trained by using training data that includes third image data (Training [section 2, page 1371 right column first full paragraph] “train a convolutional neural network for MAR. First, we generate metal-free (i.e. third image data), metal-inserted … CT images to create a database.”) and fourth image data (Training [section 2, page 1371 right column paragraph 3] “…metal-inserted CT images (i.e. fourth image data), where beam hardening and Poisson noise are simulated. To ensure that the trained CNN works for real cases…we simulate the metal artifacts based on clinical CT images.”),
the processing circuitry is further configured to reconstruct the third image data from the projection data (page [1372 left column second to last paragraph] “The metal-free image (i.e. third image data) is reconstructed using filtered backprojection (FBP), and the image is assumed as reference and denoted as xref .”) and generate the fourth image data including a generated low-count artifact from the projection data (Experiments [section 4, page 1375 right column paragraph 1] “A 120 kVp x-ray source is simulated and each detector bin is expected to receive 2 × 107 photons in the case of blank scan [46]. There are 984 projection views over a rotation and 920 detector bins in a row. The distance between the x-ray source and the rotation center is 59.5 cm. The metal-free (i.e. third image data) and metal-inserted images (i.e. fourth image data) are reconstructed by the FBP from simulated sinograms and each image consists of 512 × 512 pixels.”), and
[the fourth image data is image data that is obtained by adding a low-count artifact image that is generated in advance to image data that is reconstructed from the projection data].
However, Zhang fails to teach the fourth image data is image data that is obtained by adding a [low-count] artifact image that is generated in advance to image data that is reconstructed from the projection data.
Tohme teaches the fourth image data is image data (noise-modulated image 1000 in Paragraph [0058] equates to the fourth image data) that is obtained by adding a [low-count] artifact image (noise image 900 equates to the artifact image) that is generated in advance to image data that is reconstructed from the projection data (smoothed image 800 in Paragraph [0058] equates to image data that is reconstructed) (Paragraph [0024] “To reduce the total scan time, a “helical” scan may be performed…The helix mapped out by the cone beam yields projection data from which images in each prescribed slice may be reconstructed.”; Paragraph [0058] “FIG. 10 shows an example noise-modulated image 1000 generated from the noise image 900 and the smoothed image 800 based on the noise modulation map 600…the noise filtered from the lung region may actually include valid signal rather than noise, and so the signal is added back to the lung region when multiplying the noise image 900 by the noise modulation map 600 and adding the modulated noise back to the smoothed image 800. Conversely, the noise filtered from the abdomen region likely corresponds to actual noise rather than x-ray signal, and so the lower SNR values of the noise modulation map 600 in the abdomen region results in less noise from the noise image 900 being added back to the smoothed image 800.”).
Therefore, it would have been obvious to one of ordinary skill of the art before the effective filing date to modify Zhang’s reference to include the fourth image data is image data that is obtained by adding a [low-count] artifact image that is generated in advance to image data that is reconstructed from the projection data taught by Tohme’s reference. The motivation for doing so would have been to reduce radiation dose during imaging and improve image quality as suggested by Tohme (see Tohme, Paragraph [0004]).
However, Zhang and Tohme both fail to teach a low-count [artifact] image.
Wang teaches a low-count [artifact] image (Paragraph [0146] “The image creation block 1303 may generate an artifact image (e.g., an image including streak artifact of an original image, or referred to as a streak artifact image for brevity). In some embodiments, the artifact image may be generated by filtering the artifacts from the image relating to the ROI. In some embodiments, the artifact image may be further processed.”).
Therefore, it would have been obvious to one of ordinary skill of the art before the effective filing date to modify Zhang in view of Tohme to include a low-count [artifact] image taught by Wang’s reference. The motivation for doing so would have been to generate an artifact image based on a detected artifact and correct an image based on the artifact image as suggested by Wang (see Wang, Paragraph [0026]).
Further, one skilled in the art could have combined the elements described above by known methods with no change to the respective functions, and the combination would have yielded nothing more than predictable results. Therefore, it would have been obvious to combine Tohme and Wang with Zhang to obtain the invention specified in claim 1.
Regarding claim 7, Zhang discloses a medical image processing method, comprising:
acquiring second image data in which a low-count artifact is reduced (Introduction [section 1, page 1370 right column last full paragraph] “…convolutional neural network (CNN) has been applied to medical imaging for low dose CT reconstruction and artifacts reduction”; Introduction [section 1, page 1371 left column paragraph 1] “the information from different correction methods is captured and the merits of these methods are fused, leading to a higher quality image (i.e. second image)”), by applying a trained machine learning model to first image data that is obtained by an X-ray CT scan (Experiments [section 4 page 1376 left column first full paragraph] “scanned on a Siemens SOMATOM Sensation 16 CT scanner with 120 kVp and 496 mAs using the helical scanning geometry. (i.e. first image)”); and
outputting image data based on the second image data (See figure 3 [page 1373 left column last full paragraph continued to right column first full paragraph] “Fig. 3 depicts the workflow of our CNN, which is comprised of an input layer, an output layer and L = 5 convolutional layers…nl convolution kernel with a fixed size of cl × cl . Cl (u) generates new feature maps based on the (l − 1)th layer’s output. For the last layer, feature maps are used to generate an image that is close to the target (i.e. image data).”), wherein
the machine learning model is trained by using training data that includes third image data (Training [section 2, page 1371 right column first full paragraph] “train a convolutional neural network for MAR. First, we generate metal-free (i.e. third image data), metal-inserted … CT images to create a database.”) and fourth image data (Training [section 2, page 1371 right column paragraph 3] “…metal-inserted CT images (i.e. fourth image data), where beam hardening and Poisson noise are simulated. To ensure that the trained CNN works for real cases…we simulate the metal artifacts based on clinical CT images.”),
the method further comprises reconstructing the third image data from the projection data(page [1372 left column second to last paragraph] “The metal-free image (i.e. third image data) is reconstructed using filtered backprojection (FBP), and the image is assumed as reference and denoted as xref .”) and generate the fourth image data including a generated low-count artifact from the projection data (Experiments [section 4, page 1375 right column paragraph 1] “A 120 kVp x-ray source is simulated and each detector bin is expected to receive 2 × 107 photons in the case of blank scan [46]. There are 984 projection views over a rotation and 920 detector bins in a row. The distance between the x-ray source and the rotation center is 59.5 cm. The metal-free (i.e. third image data) and metal-inserted images (i.e. fourth image data) are reconstructed by the FBP from simulated sinograms and each image consists of 512 × 512 pixels.”), and
[the fourth image data is image data that is obtained by adding a low-count artifact image that is generated in advance to image data that is reconstructed from the projection data].
However, Zhang fails to teach the fourth image data is image data that is obtained by adding a [low-count] artifact image that is generated in advance to image data that is reconstructed from the projection data.
Tohme teaches the fourth image data is image data (noise-modulated image 1000 in Paragraph [0058] equates to the fourth image data) that is obtained by adding a [low-count] artifact image (noise image 900 equates to the artifact image) that is generated in advance to image data that is reconstructed from the projection data (smoothed image 800 in Paragraph [0058] equates to image data that is reconstructed) (Paragraph [0024] “To reduce the total scan time, a “helical” scan may be performed…The helix mapped out by the cone beam yields projection data from which images in each prescribed slice may be reconstructed.”; Paragraph [0058] “FIG. 10 shows an example noise-modulated image 1000 generated from the noise image 900 and the smoothed image 800 based on the noise modulation map 600…the noise filtered from the lung region may actually include valid signal rather than noise, and so the signal is added back to the lung region when multiplying the noise image 900 by the noise modulation map 600 and adding the modulated noise back to the smoothed image 800. Conversely, the noise filtered from the abdomen region likely corresponds to actual noise rather than x-ray signal, and so the lower SNR values of the noise modulation map 600 in the abdomen region results in less noise from the noise image 900 being added back to the smoothed image 800.”).
Therefore, it would have been obvious to one of ordinary skill of the art before the effective filing date to modify Zhang’s reference to include the fourth image data is image data that is obtained by adding a [low-count] artifact image that is generated in advance to image data that is reconstructed from the projection data taught by Tohme’s reference. The motivation for doing so would have been to reduce radiation dose during imaging and improve image quality as suggested by Tohme (see Tohme, Paragraph [0004]).
However, Zhang and Tohme both fail to teach a low-count [artifact] image.
Wang teaches a low-count [artifact] image (Paragraph [0146] “The image creation block 1303 may generate an artifact image (e.g., an image including streak artifact of an original image, or referred to as a streak artifact image for brevity). In some embodiments, the artifact image may be generated by filtering the artifacts from the image relating to the ROI. In some embodiments, the artifact image may be further processed.”).
Therefore, it would have been obvious to one of ordinary skill of the art before the effective filing date to modify Zhang in view of Tohme to include low-count [artifact] image taught by Wang’s reference. The motivation for doing so would have been to generate an artifact image based on a detected artifact and correct an image based on the artifact image as suggested by Wang (see Wang, Paragraph [0026]).
Further, one skilled in the art could have combined the elements described above by known methods with no change to the respective functions, and the combination would have yielded nothing more than predictable results. Therefore, it would have been obvious to combine Tohme and Wang with Zhang to obtain the invention specified in claim 7.
Regarding claim 8, Zhang discloses a model generation method for generating a machine learning model that acquires second image data in which a low count artifact is reduced (Introduction [section 1, page 1370 right column last full paragraph] “…convolutional neural network (CNN) has been applied to medical imaging for low dose CT reconstruction and artifacts reduction”; Introduction [section 1, page 1371 left column paragraph 1] “the information from different correction methods is captured and the merits of these methods are fused, leading to a higher quality image (i.e. second image)”), by applying a trained machine learning model to first image data that is obtained by an X-ray CT scan (Experiments [section 4 page 1376 left column first full paragraph] “scanned on a Siemens SOMATOM Sensation 16 CT scanner with 120 kVp and 496 mAs using the helical scanning geometry. (i.e. first image)”), the model generation method comprising:
generating the machine learning model by training a model that is not yet trained, by using training data that includes third image data and fourth image data (Introduction [section 1, page 1371 left column, paragraph 1] “Specifically, before the MAR, we build a MAR database to generate training data for the CNN.”; Introduction [section 1, page 1371 right column, first full paragraph] “There are two phases to train a convolutional neural network for MAR. First, we generate metal-free, metal-inserted and MAR corrected CT images to create a database. Then, a CNN is constructed and the training data is collected from the established database and used to train the CNN”),
wherein the method further comprises reconstructing the third image data from the projection data (page [1372 left column second to last paragraph] “The metal-free image (i.e. third image data) is reconstructed using filtered backprojection (FBP), and the image is assumed as reference and denoted as xref .”) and generate the fourth image data including a generated low-count artifact from the projection data (Experiments [section 4, page 1375 right column paragraph 1] “A 120 kVp x-ray source is simulated and each detector bin is expected to receive 2 × 107 photons in the case of blank scan [46]. There are 984 projection views over a rotation and 920 detector bins in a row. The distance between the x-ray source and the rotation center is 59.5 cm. The metal-free (i.e. third image data) and metal-inserted images (i.e. fourth image data) are reconstructed by the FBP from simulated sinograms and each image consists of 512 × 512 pixels.”), and
[the fourth image data is image data that is obtained by adding a low-count artifact image that is generated in advance to image data that is reconstructed from the projection data].
However, Zhang fails to teach the fourth image data is image data that is obtained by adding a [low-count] artifact image that is generated in advance to image data that is reconstructed from the projection data.
Tohme teaches the fourth image data is image data (noise-modulated image 1000 in Paragraph [0058] equates to the fourth image data) that is obtained by adding a [low-count] artifact image (noise image 900 equates to the artifact image) that is generated in advance to image data that is reconstructed from the projection data (smoothed image 800 in Paragraph [0058] equates to image data that is reconstructed) (Paragraph [0024] “To reduce the total scan time, a “helical” scan may be performed…The helix mapped out by the cone beam yields projection data from which images in each prescribed slice may be reconstructed.”; Paragraph [0058] “FIG. 10 shows an example noise-modulated image 1000 generated from the noise image 900 and the smoothed image 800 based on the noise modulation map 600…the noise filtered from the lung region may actually include valid signal rather than noise, and so the signal is added back to the lung region when multiplying the noise image 900 by the noise modulation map 600 and adding the modulated noise back to the smoothed image 800. Conversely, the noise filtered from the abdomen region likely corresponds to actual noise rather than x-ray signal, and so the lower SNR values of the noise modulation map 600 in the abdomen region results in less noise from the noise image 900 being added back to the smoothed image 800.”).
Therefore, it would have been obvious to one of ordinary skill of the art before the effective filing date to modify Zhang’s reference to include the fourth image data is image data that is obtained by adding a [low-count] artifact image that is generated in advance to image data that is reconstructed from the projection data taught by Tohme’s reference. The motivation for doing so would have been to reduce radiation dose during imaging and improve image quality as suggested by Tohme (see Tohme, Paragraph [0004]).
However, Zhang and Tohme both fail to teach a low-count [artifact] image.
Wang teaches a low-count [artifact] image (Paragraph [0146] “The image creation block 1303 may generate an artifact image (e.g., an image including streak artifact of an original image, or referred to as a streak artifact image for brevity). In some embodiments, the artifact image may be generated by filtering the artifacts from the image relating to the ROI. In some embodiments, the artifact image may be further processed.”).
Therefore, it would have been obvious to one of ordinary skill of the art before the effective filing date to modify Zhang in view of Tohme to include low-count [artifact] image taught by Wang’s reference. The motivation for doing so would have been to generate an artifact image based on a detected artifact and correct an image based on the artifact image as suggested by Wang (see Wang, Paragraph [0026]).
Further, one skilled in the art could have combined the elements described above by known methods with no change to the respective functions, and the combination would have yielded nothing more than predictable results. Therefore, it would have been obvious to combine Tohme and Wang with Zhang to obtain the invention specified in claim 8.
Claims 4 and 6 are rejected under 35 U.S.C. 103 as being unpatentable over Zhang et al ("Convolutional neural network based metal artifact reduction in x-ray computed tomography." IEEE transactions on medical imaging 37.6 (2018): 1370-1381.) (hereinafter, “Zhang”) ”) in view of Tohme et al. (US 2021/0158486 A1) (hereinafter, “Tohme”) and Wang et al. (US 2017/0301066 A1) (hereinafter, “Wang”) as applied to claim 1 above; and further in view of Ziabari et al (US 2022/0035961 A1) (hereinafter, “Ziabari”).
Regarding claim 4, which claim 1 is incorporated, Zhang discloses wherein the processing circuitry is further configured to acquire, by the machine learning model, the second image data in which the low count artifact is reduced (Introduction [section 1, page 1370 right column last full paragraph] “…convolutional neural network (CNN) has been applied to medical imaging for low dose CT reconstruction and artifacts reduction”).
However, Zhang, Tohme and Wang fail to teach wherein the processing circuitry acquires, by the machine learning model, the second image data in which noise is reduced.
Ziabari teaches wherein the processing circuitry acquires, by the machine learning model, the second image data in which noise is reduced (paragraph [0040] “trained model can be used to rapidly process new, non-simulated data and produce reconstructions by effectively suppressing artifacts the model was trained to reduce, such as detector noise”, paragraph [0050] “… occurs in the projection or sinogram domain, the training skips reconstruction and trains directly on the projection volumes to reduce artifacts, in this case beam hardening and detector noise artifacts. Removing beam hardening artifacts from the projection or sinogram helps to avoid reliance on an initial reconstruction for image domain data can result in a higher quality reconstruction.).
Therefore, it would have been obvious to one of ordinary skill of the art before the effective filing date to modify Zhang in view of Tohme; and further in view of Wang to include noise reduction taught by Ziabari. The motivation for doing so would have been to reduce the amount of artifact and thereby increasing the resolution of the imaging system as suggested by Ziabari (see Ziabari paragraph [0048]).
Further, one skilled in the art could have combined the elements described above by known methods with no change to the respective functions, and the combination would have yielded nothing more that predictable results Therefore, it would have been obvious to combine Ziabari with Zhang, Tohme, and Wang to obtain the invention specified in claim 4.
Regarding claim 6, which claim 1 is incorporated, Zhang, Tohme and Wang fail to teach wherein the processing circuitry is further configured to acquire processed image data in which noise is reduced, by applying a machine learning model that is trained for at least reducing noise to the second image data, and output image data based on the processed image data.
Ziabari teaches the processing circuitry is further configured to acquire processed image data in which noise is reduced, by applying a machine learning model that is trained for at least reducing noise to the second image data (paragraph [0040] “trained model can be used to rapidly process new, non-simulated data and produce reconstructions by effectively suppressing artifacts the model was trained to reduce, such as detector noise”), and output image data based on the processed image data (paragraph [0042] “output of a conventional reconstruction algorithm or in the form of a complete reconstructed image with artifact reduction.”).
Therefore, it would have been obvious to one of ordinary skill of the art before the effective filing date to modify Zhang in view of Tohme; and further in view of Wang to include acquiring image data in which noise is reduced by applying a machine learning model trained to reduce noise taught by Ziabari. The motivation for doing so would have been to reduce the amount of artifact and thereby increasing the resolution of the imaging system as suggested by Ziabari (see Ziabari, paragraph [0048]).
Further, one skilled in the art could have combined the elements described above by known methods with no change to the respective functions, and the combination would have yielded nothing more that predictable results Therefore, it would have been obvious to combine Ziabari with Zhang, Tohme, and Wang to obtain the invention specified in claim 6.
Claim 9 is rejected under 35 U.S.C. 103 as being unpatentable over Zhang et al ("Convolutional neural network based metal artifact reduction in x-ray computed tomography." IEEE transactions on medical imaging 37.6 (2018): 1370-1381.) (hereinafter, “Zhang”) ”) in view of Tohme et al. (US 20210158486 A1) (hereinafter, “Tohme”) and Wang et al. (US 2017/0301066 A1) (hereinafter, “Wang”) as applied to claim 1 above; and further in view of Zeng et al. ("A simple low-dose x-ray CT simulation from high-dose scan." IEEE transactions on nuclear science 62.5 (2015): 2226-2233.) (hereinafter, “Zeng”).
Regarding claim 9, which claim 1 is incorporated, Zhang discloses wherein the fourth image data includes image data reconstructed from data output from a low-count simulation process applied to the projection data (Training [section 2, page 1371 right column paragraph 1] ”noisy polychromatic projection is obtained, and then the image containing artifacts is reconstructed.”; Training [section 2, page 1371 right column paragraph 3] “…metal-inserted CT images (i.e. fourth image data), where beam hardening and Poisson noise are simulated. To ensure that the trained CNN works for real cases…we simulate the metal artifacts based on clinical CT images.”), and
[the low-count simulation process is a process of simulating a low-dose count by multiplying a photon count of the projection data by a coefficient].
However, Zhang, Tohme, and Wang fail to teach the low-count simulation process is a process of simulating a low-dose count by multiplying a photon count of the projection data by a coefficient.
PNG
media_image5.png
50
390
media_image5.png
Greyscale
Zeng teaches the low-count simulation process is a process of simulating a low-dose count by multiplying a photon count of the projection data by a coefficient (Methods and Material, page 2227 [left column, first paragraph] “The associative formula can be expressed as follows:
where is the measured noisy transmission datum and is the mean number of photon passing though the patient, and the magnitude of is primarily determined by the mAs value”; page 2227 [right column, first paragraph and step 4] “Based on the theoretical statistical model of CT transmission data in (1), in this paper we propose a simple low-dose CT simulation strategy…Multiply the transmission data Tnd by the simulated low-dose scan incident flux I--old,sim to produce the simulated low-dose transmission data I--old,sim, i.e., I--old,sim (s) = I--old,sim (s)Tnd (s), wherein the factor I--old,sim is determined by a relationship between the incident fluxes of low- and high-dose scans”).
Therefore, it would have been obvious to one of ordinary skill of the art before the effective filing date to modify Zhang in view of Tohme; and further in view of Wang to include the low-count simulation process is a process of simulating a low-dose count by multiplying a photon count of the projection data by a coefficient taught by Zeng’s reference. The motivation for doing so would have been to provide medical professionals the ability to study the effects of lower dose on image quality as suggested by Zeng (see Zeng, page 2232 Discussion and Conclusion [right column, first paragraph]).
Further, one skilled in the art could have combined the elements described above by known methods with no change to the respective functions, and the combination would have yielded nothing more that predictable results. Therefore, it would have been obvious to combine Zeng with Zhang, Tohme, and Wang to obtain the invention specified in claim 9.
Claim 11 is rejected under 35 U.S.C. 103 as being unpatentable over Zhang et al ("Convolutional neural network based metal artifact reduction in x-ray computed tomography." IEEE transactions on medical imaging 37.6 (2018): 1370-1381.) (hereinafter, “Zhang”) ”) in view of Tohme et al. (US 20210158486 A1) (hereinafter, “Tohme”) and Wang et al. (US 2017/0301066 A1) (hereinafter, “Wang”) as applied to claim 1 above; and further in view of Ikuta et al. (US 12,433,560 B2) (hereinafter, “Ikuta”).
Regarding claim 11, which claim 1 is incorporated, Zhang fails to teach wherein the processing circuitry is further configured to: select one of a plurality of pre-generated, low-count artifact images; and generate the fourth image data by adding the selected low-count artifact image to the reconstructed third image data.
Tohme discloses wherein the processing circuitry is further configured to:
[select one of a plurality of pre-generated, low-count artifact images]; and
generate the fourth image data by adding the [selected low-count] artifact image to the reconstructed third image data (Paragraph [0024] “To reduce the total scan time, a “helical” scan may be performed…The helix mapped out by the cone beam yields projection data from which images in each prescribed slice may be reconstructed.”; Paragraph [0058] “FIG. 10 shows an example noise-modulated image 1000 generated from the noise image 900 and the smoothed image 800 based on the noise modulation map 600…the noise filtered from the lung region may actually include valid signal rather than noise, and so the signal is added back to the lung region when multiplying the noise image 900 by the noise modulation map 600 and adding the modulated noise back to the smoothed image 800. Conversely, the noise filtered from the abdomen region likely corresponds to actual noise rather than x-ray signal, and so the lower SNR values of the noise modulation map 600 in the abdomen region results in less noise from the noise image 900 being added back to the smoothed image 800.”).
Therefore, it would have been obvious to one of ordinary skill of the art before the effective filing date to modify Zhang’s reference to include generate the fourth image data by adding the [selected low-count] artifact image to the reconstructed third image data taught by Tohme’s reference. The motivation for doing so would have been to reduce radiation dose during imaging and improve image quality as suggested by Tohme (see Tohme, Paragraph [0004]).
However, Zhang and Tohme both fail to teach select one of a plurality of pre-generated, low-count artifact images.
Wang teaches low-count [artifact] image (Paragraph [0146] “The image creation block 1303 may generate an artifact image (e.g., an image including streak artifact of an original image, or referred to as a streak artifact image for brevity). In some embodiments, the artifact image may be generated by filtering the artifacts from the image relating to the ROI. In some embodiments, the artifact image may be further processed.”).
Therefore, it would have been obvious to one of ordinary skill of the art before the effective filing date to modify Zhang in view of Tohme to include low-count [artifact] image taught by Wang’s reference. The motivation for doing so would have been to generate an artifact image based on a detected artifact and correct an image based on the artifact image as suggested by Wang (see Wang, Paragraph [0026]).
However, Zhang, Tohme, and Wang fail to teach select one of a plurality of pre-generated, [low-count artifact images].
Ikuta teaches select one of a plurality of pre-generated, [low-count artifact images] (Column 16 [lines 36-40] “the artifact selection component 1102 can facilitate selecting specific pre-generated synthetic artifacts included in the synthetic artifact data 118 to add to one or more CT images (e.g., included in the CT image data 112) as described with reference to FIG. 5.”).
Therefore, it would have been obvious to one of ordinary skill of the art before the effective filing date to modify Zhang in view of Tohme; and further in view of Wang to include select one of a plurality of pre-generated, [low-count artifact images] taught by Ikuta’s reference. The motivation for doing so would have been to enable rapid and efficient integration of different types of synthetic artifacts as suggested by Ikuta (see Ikuta, Column 4 [lines 51-56]).
Further, one skilled in the art could have combined the elements described above by known methods with no change to the respective functions, and the combination would have yielded nothing more that predictable results. Therefore, it would have been obvious to combine Ikuta with Zhang, Tohme, and Wang to obtain the invention specified in claim 11.
Allowable Subject Matter
Claim 10 is objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
Claim 10 contains subject matter that is not disclosed or made obvious in the cited art.
In regards to claim 10, when considering claim 10 as a whole prior art fails to disclose or render obvious, alone or in combination:
“[…] wherein the low-count simulation process includes a noise addition process and a zero clipping process with respect to a negative value of the projection data.”
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Takahashi et al. (US 2015/0010224 A1) discloses an image processing method for generating a target image which edges of a structure are upheld and streaking artifacts are removed.
Goossens et al. (US 2013/0051674 A1) discloses estimating noise in a reconstructed image by determining first and second noise and generate a denoised image with visually improved image quality.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to UROOJ FATIMA whose telephone number is (571)272-2096. The examiner can normally be reached M-F 8:00-5:00.
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, Henok Shiferaw can be reached at (571) 272-4637. 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.
/UROOJ FATIMA/Examiner, Art Unit 2676
/Henok Shiferaw/Supervisory Patent Examiner, Art Unit 2676
Prosecution Insights