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and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114.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 1/06/2026 has been entered.
Claim Interpretation
In regards to claim 7, due to the amendments added to claim 1, the term “applying” is being interpreted in light of those amendments. So, the term “applying” is being interpreted as involving concatenation, element-wise multiplication, or some combination thereof. Further, the term, “smooth scaling map” is being interpreted as some form of continuous map with continuous values. This definition excludes the cited idea in applicant’s definition of a form of gradual change as that could lead to further 112(b) issues. If applicant has an issue with this interpretation, they may respond in a future filing in response to this rejection.
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claims 1-2, 5-11 and 22-23 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ),
first paragraph, as failing to comply with the written description requirement. The claims contain
subject matter which was not described in the specification in such a way as to reasonably
convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications
subject to pre-AIA 35 U.S.C. 112, the inventors, at the time the application was filed, had possession of the claimed invention.
In regards to claim 1, claim 1 lacks any compensation for “domain gap” or “domain shift” as it is known in the art. It is common knowledge in the art that data collected with different imaging equipment and scanning protocols can be quite different in many aspects.
Therefore, if a network is trained with a particular type of data, it cannot address the issues
related to a different type of data. In this case, the desired effect cannot be achieved without
some form of domain adaptation. However, no domain adaptation is disclosed in the claim or in
the description. Further, and the reason for claim 2 to be similarly rejected, is that claim 2
discloses that no such transformation is performed on the training data which could address this
problem.
In regards to claims 5-11, and 22-23 is that they inherit the deficiency of claim 1, and they are similarly rejected.
Claims 1-2, 5-11 and 22-24 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the enablement requirement. The claims contain
subject matter which was not described in the specification in such a way as to enable one skilled
in the art to which it pertains, or with which it is most nearly connected, to make and/or use the
invention.
In regards to claim 1, claim 1 lacks any compensation for “domain gap” or “domain
shift” as it is known in the art. It is common knowledge in the art that data collected with
different imaging equipment and scanning protocols can be quite different in many aspects.
Therefore, if a network is trained with a particular type of data, it cannot address the issues
related to a different type of data. In this case, the desired effect cannot be achieved without
factoring in domain shift. However, no way to compensate for domain shift is disclosed in the
claim or in the description. Further, and the reason for claim 2 to be similarly rejected, is that
claim 2 discloses that no such transformation is performed on the training data which could
address this problem. As such, claim 1 and 2 would require undue experimentation on part of the
applicant due to the nature of the invention and the amount of direction provided by the inventor
would likely lead to one of ordinary skill in the art to practice the invention without factoring in
the domain shift and thus lead them to undue experimentation to factor in the domain shift for
claim 1. Claim 2 limits the invention to not factor in the domain shift at all, and it would require
a person of ordinary skill to further modify the invention to make it perform properly.
In regards to claims 5-11, and 22-23 is that they inherit the deficiency of claim 1, and they are similarly rejected.
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 text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
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.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claims 1-2, 5, 11, and 22-24 are rejected under 35 U.S.C. 103 as being unpatentable over Mizobe et al. (US 20210104313 A1), hereinafter referred to as Mizobe, in view of Souza et al. (US-8824753-B2), hereinafter referred to as Souza, and Koch et al. (“Accurate Segmentation of Dental Panoramic Radiographs with U-Nets”), hereinafter referred to as Koch, and Li et al. (CN 111145177 A), hereinafter referred to as Li, and Schildkraut et al. (US 20210153823 A1), hereinafter referred to as Schildkraut, and Gallagher et al. (US 20030039401 A1), hereinafter referred to as Gallagher.
In regards to claim 1, Mizobe discloses a computer-implemented method of denoising images, comprising (Mizobe in the last sentence of par. 436 and paragraph 503, paragraph 79 and 80, discloses image quality improvement and denotes denoising specifically in paragraph 503 along with showing that their system in paragraphs 79 and 80 is also focused on denoising images): obtaining a training set comprising two-dimensional digital images of a first type, wherein the images of the first type are acquired by a medical imaging modality (Mizobe in par. 438 in lines 11- 13 and paragraph 439 covers the training set of the first group is used to train the first training engine of the system with paragraph 84 covering two-dimensional images and paragraph 636 would allow for the images and embodiments to be on a computer which would make the images digital images), and the training set comprises a plurality of data pairs, where each data pair comprises a noisy image and a true image, where the noisy image is the true image with generated noise (Mizobe in paragraph 442 and paragraph 406, this refers to their 18th embodiment as described more thoroughly in paragraph 406 which includes direct reference to a pair of images one with additional noise and one true image); training at least one neural network on the training set, wherein input for the at least one neural network is a plurality of the noisy images, and a target of the at least one neural network is a plurality of the true images that correspond to the plurality of the noisy images, such that inputting a noisy image into a trained neural network results in output of a denoised image similar to a true image corresponding to the input noisy image (Mizobe paragraphs 411 and 412, Mizobe discloses a first and second noise component in 411 and in 412 discloses that the learning method is not limited to innately those two components as such, it would cover the above); obtaining a two-dimensional digital image of a second type; and denoising the image of the second type by using the image of the second type as input for the trained neural network (Mizobe in par. 448, an image of the second group density is inserted as input to the system and the system is able to denoise it using the trained engine for the second group with paragraph 84 covering two-dimensional images and paragraph 636 would allow for the images and embodiments to be on a computer which would make the images digital images).
However, Mizobe does not disclose that the image of the second type must be a dental image.
Souza does disclose using dental images in place of medical images (Column 1, Lines 12 – 18, Souza discloses how noise impacts medical and dental images and that they have shared applications).
Therefore it would have been a prima facie obvious case of simple substitution. As Souza establishes, it was well-known in the art that similar processes could be performed on medical images and dental images. Further, substituting the medical images of Mizobe for the dental images of Souza would lead to a predictable result as dental images are merely a subcategory of medical images, and thus, similar processes can be performed on both. Therefore, this acts as a simple substitution and is rejected under 35 U.S.C. 103.
Neither Mizobe or Souza explicitly disclose that wherein; at least one of the data pairs is generated by generating a noise matrix, and the noisy image is generated by combining the true image with the noise matrix.
However, Li does disclose wherein; at least one of the data pairs by: generating a noise matrix; and generating the noisy image by combining the true image with the noise matrix (First new paragraph of page 7, Li describes generating a noise matrix and then using it to add noise to the original image matrix).
It would have been prima facie obvious to combine the teachings of Mizobe and Souza with the teachings of Li. Utilizing a noise matrix would allow for an individual to utilize various probabilistic functions as noted by Li with Gaussian Covariance which would allow for greater accuracy with noise generation as it would allow a user to probabilistically determine the best areas to add noise to.
Souza, Mizobe, and Li fail to explicitly disclose wherein the dental image of the second type is a radiographic image, a fluorescence image, and/or an infrared image.
However, Koch does disclose wherein the dental image of the second type is a radiographic image, a fluorescence image, and/or an infrared image (Title, Koch’s title introduces Dental Panoramic Radiographs as a form of dental image as such).
Therefore, it would have been prima facie obvious to combine the references. Combining the teachings of Koch with the teachings of Souza, Mizobe, and Li would have led to a predictable increase in accuracy. As the dental images would be limited to certain explicit types, it would allow for any denoising process to be further trained to work with specific kinds of image allowing it to be more accurate and useful for future denoising.
Mizobe, Souza, Li, and Koch does not explicitly discloses that the first and second type of images are obtained at different wavelengths of light.
However, Schildkraut does disclose that the first and second type of images are obtained at different wavelengths of light (Paragraph 92, this paragraph discloses that there can be a first and second wavelength for the creation of x-ray images and this would split the images into those of a type that consists of one wavelength and to those of a type that consists of a different wavelength).
It would have been prima facie obvious to combine the teachings of Schildkraut with those of Mizobe, Souza, Li, and Koch as it would have been simple substitution. Mizobe already disclosed that there can be two types utilized but that these would merely be images of different scanning densities. Koch discloses that radiographic dental images may be used. Schildkraut merely changes the type disclosed to be those of radiographs of different wavelengths. As such, it functions as a simple substitution of the types of Mizobe for the types disclosed by Schildkraut.
None of the previously recited arts disclose that the noise is created via a concatenation and/or an elementwise multiplication process.
However, Gallagher does disclose that noise can be created via a concatenation and/or an elementwise multiplication process (Paragraph 44, Gallagher discloses that the recited noise table requires a concatenation process of various other tables to create it)
It would be prima facie obvious to combine the teachings of these arts as it would be simple substitution to replace the noise disclosed by the previous arts with one generated via some form of concatenation. There is no major difference between noise generated in the claimed manner and the noise generated by some other means. Both affect images in the same manner. As such, one could take the noise created by Gallagher, and substitute it in for the noise used by the previous prior arts. As such, it would be prima facie obvious to combine these arts.
In regards to claim 2, Mizobe discloses a method according to claim 1, wherein the images of the first type include images that are directly used as true images in the data pair (Mizobe, par. 448: an image of the second group density is inserted as input to the system and the system is able to denoise it using the trained engine for the second group).
In regards to claim 5, Mizobe, Souza, Koch, and Schildkraut do not disclose generating an empty matrix; generating values from a probability distribution; and generating the noise matrix by assigning one of the generated values to each element of the empty matrix.
However, Li does teach generating an empty matrix(First new paragraph of page 7, this is inherent to generating a matrix as an empty matrix is merely a matrix with no values assigned); generating values from a probability distribution(First new paragraph of page 7, Li utilizes a Gaussian Covariance in the generation of values which is a form of probability distribution); and generating the noise matrix by assigning one of the generated values to each element of the empty matrix(First new paragraph of page 7, Li assigns the values to the matrix after generating them to create a noise matrix).
It would have been prima facie obvious to combine the teachings of Mizobe, Souza, Koch, and Schildkraut with the teachings of Li. Utilizing a noise matrix would allow for an individual to utilize various probabilistic functions as noted by Li with Gaussian Covariance which would allow for greater accuracy with noise generation as it would allow a user to probabilistically determine the best areas to add noise to.
In regards to claim 11, Mizobe, Souza, Koch, and Schildkraut fail to disclose generating values from a normal distribution for each line in the noise matrix; and applying one of the generated values to each element of the line.
However, Li does disclose that generating values from a normal distribution for each line in the noise matrix (First new paragraph of page 7, Li’s usage of a noise matrix and Gaussian Covariance is found applicable as it involves a normal distribution generating values for a noise matrix which would include all of the rows and columns); and applying one of the generated values to each element of the line (First new paragraph of page 7, Li’s application of generated values to the noise matrix along with the image as described in the paragraph, would cover this).
It would have been prima facie obvious to combine the teachings of Mizobe, Souza, Koch, and Schildkraut with the teachings of Li. Utilizing a noise matrix would allow for an individual to utilize various probabilistic functions as noted by Li with Gaussian Covariance which would allow for greater accuracy with noise generation as it would allow a user to probabilistically determine the best areas to add noise to.
In regards to claim 22, Mizobe discloses a non-transitory medium including computer program product configured, when run, to execute the method of claim 1 (Paragraph 68, Mizobe discloses that the process can be done by a software module that is executed by a processor).
In regards to claim 23, Mizobe discloses a data processing system comprising a processor configured to perform the steps of the method of claim 1 (Paragraph 68, Mizobe discloses that the process can be done by a software module that is executed by a processor).
In regards to claim 24, Mizobe discloses a computer-readable data carrier having stored thereon the computer program product of claim 22 (Paragraph 68, Mizobe discloses that the process can be done by a software module, within the BRI of a computer-readable data carrier, that is executed by a processor).
Claim 6 is rejected under 35 U.S.C. 103 as being unpatentable over Mizobe et al. (US 20210104313 A1), hereinafter referred to as Mizobe, in view of Souza et al. (US-8824753-B2), hereinafter referred to as Souza, and Koch et al. (“Accurate Segmentation of Dental Panoramic Radiographs with U-Nets”), hereinafter referred to as Koch, and Li et al. (CN 111145177 A), hereinafter referred to as Li, and Schildkraut et al. (US 20210153823 A1), hereinafter referred to as Schildkraut, and Gallagher et al. (US 20030039401 A1), hereinafter referred to as Gallagher, as applied to claims 1-2, 5, 11, and 22-24 above, and further in view of Cao et al. (CN 111598808 A), hereinafter referred to as Cao.
In regards to claim 6, Mizobe, Souza, Koch, Li, Gallagher, and Schildkraut fail to disclose the aspects of this claim.
Cao does disclose generating a blurring kernel (Eighth new paragraph on page 6, this describes blur kernel generation ), comprising a matrix of dimensions of at least two by one (Eighth new paragraph on page 6, as a kernel is a matrix, it would encompass all sizes of a matrix including those greater than 2x1 since it encompasses all blurring kernels), where each element of the blurring kernel has a weight value (eighth new paragraph of page 6, this is inherent to a blur kernel as a matrix/kernel is a grouping of values, there must be some value/weight in the matrix); selecting an element of the noise matrix (eighth new paragraph of page 6, it describes how the noise and blur kernel interact); and altering the value of the selected element based on the combined value of at least one neighboring element and the weight values of at least one corresponding element of the blurring kernel (eighth new paragraph of page 6, last paragraph of page 7, and second new paragraph of page 8, it describes how each blur kernel and noise are added or combined into the degradation pool).
It would have been prima facie obvious to combine the teachings of all of these inventors as factoring in the blurring along with the noise would have allowed for more effective denoising. This additional factor of blurring would have led to predictable results of increasing the effectiveness as such, it is prima facie obvious to combine the various teachings.
Response to Amendment
The amendments, entered on 1/06/2026, have been considered in full. These amendments have overcome the 112 rejections based around claim 7 with claim interpretation along with some of the 112(a) rejections. They did not overcome the 112(a) arguments around domain gap. Amendments to the specification have overcome all objections therein.
Response to Arguments
Applicant's arguments filed 1/06/2026 have been fully considered but they are not persuasive.
In regards to the arguments based around domain adaptation, the arguments to the “domain adaptation” are not persuasive. If the noise were sufficient to overcome this rejection, then this rejection would not be maintained in the final rejection. As such, merely further specifying how the noise itself is created does not address the issue of how the domain is adapted to the changes within the images. To advance prosecution, further detailing on the kernel process or how the neural networks function may move the application further to allowance. Further, the granted European application that this document lays priority to may help in overcoming this rejection.
In regards to the arguments directed to the Schildkraut reference; it is in a clearly related field of x-ray tomography. Even if the reference does not explicitly use the words train or denoise, in paragraphs 99 and 73 recite the issues of noise and an iterative process respectively. As such, it is a related document that mentions these issues. Secondly, Schildkraut is being used to merely substitute two kinds of images. The usage of Schildkraut does not require it to mention training or noise as it is being used merely to show that one could substitute images with those of different wavelengths. As such, the argument against Schildkraut is not persuasive.
Applicant’s arguments with respect to claim 1 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
The newly introduced Gallagher reference covers the concept of concatenation being used to generate noise, so this 103 argument is rendered moot.
Other arguments presented in regards to other 112 issues were found to be persuasive, and as such, those rejections have been withdrawn with the related interpretations being put in the claim interpretations above.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to CONOR AIDAN O'MALLEY whose telephone number is (571)272-0226. The examiner can normally be reached Monday - Friday 9:00 am. - 5:00 pm. 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 Moyer can be reached at 5722729523. 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.
CONOR AIDAN. O'MALLEY
Examiner
Art Unit 2675
/CONOR A O'MALLEY/Examiner, Art Unit 2675
/ANDREW M MOYER/Supervisory Patent Examiner, Art Unit 2675