CTNF 18/885,949 CTNF 88095 DETAILED ACTION Notice of Pre-AIA or AIA Status 07-03-aia AIA 15-10-aia The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA. Priority 02-26 AIA Receipt is acknowledged of certified copies of papers required by 37 CFR 1.55. Claim Rejections - 35 USC § 101 07-04-01 AIA 07-04 35 U.S.C. 101 reads as follows: Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title. Claim 14 is rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter because, taken as a whole, this claim is directed to the description of a computer program. 35 U.S.C. 101 enumerates four categories of subject matter that Congress deemed to be appropriate subject matter for a patent: processes, machines, manufactures and compositions of matter. As explained by the courts, these "four categories together describe the exclusive reach of patentable subject matter. If a claim covers material not found in any of the four statutory categories, that claim falls outside the plainly expressed scope of § 101 even if the subject matter is otherwise new and useful." In re Nuijten , 500 F.3d 1346, 1354, 84 USPQ2d 1495, 1500 (Fed. Cir. 2007). Non-limiting examples of claims that are not directed to any of the statutory categories include products that do not have a physical or tangible form, such as information (often referred to as "data per se") or a computer program per se (often referred to as "software per se") when claimed as a product without any structural recitations. Software expressed as code or a set of instructions detached from any medium is an idea without physical embodiment. See Microsoft Corp. v. AT&T Corp ., 550 U.S. 437, 449, 82 USPQ2d 1400, 1407 (2007 ); see also Benson , 409 U.S. 67, 175 USPQ2d 675 (An "idea" is not patent eligible). Thus, a product claim to a software program that does not also contain at least one structural limitation (such as a "means plus function" limitation) has no physical or tangible form, and thus does not fall within any statutory category. Claim Rejections - 35 USC § 102 07-07-aia AIA 07-07 The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action: A person shall be entitled to a patent unless – 07-12-aia AIA (a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention. 07-15-03-aia AIA Claim s 1-3, 5 and 12-15 are rejected under 35 U.S.C. 102(a)(2) as being anticipated by Banerjee et al (US PG Pub. No. 2021/0201003) . With regards to claim 1, Banerjee discloses receiving an input training dataset comprising real training data (“’real world’ images”) corresponding to a real modality and augmented training data (“synthetic data”) at: ¶ [0025]; ¶ [0027]; ¶ [0036] Banerjee discloses iteratively (“retraining weights”) training an encoder (“generator”) of the machine-learning model (“generative adversarial network (GAN)”) to obtain a trained encoder at: ¶¶ [0023]-[0025], ¶ [0027]; ¶ [0030]; ¶¶ [0037]-[0038] and FIGS. 3B and 3C Banerjee further discloses the iterative training (“retraining weights”) comprises training a discriminator model (“discriminator neural network”) by minimizing a discriminator loss function using the input training dataset to obtain a trained discriminator, the trained discriminator being configured to discriminate whether input data is real data or augmented data at: ¶¶ [0023]- [0025], [0027](“The discriminator can be used to determine or ‘discriminate’ data. The discriminating includes predicting whether the data is real or synthetic.”); ¶¶ [0029]-[0031]; ¶¶ [0033]-[0035]; ¶ [0037](“Discriminator accuracy is determined based on the discriminator accurately predicting whether the vector is real. The loss is determined based on whether the determination or prediction by the discriminator is correct (no error) or incorrect (error). An error 366 is fed back to the discriminator 360. The fedback error can be used to adjust weights associated with nodes and layers within the neural network. The weights can be adjusted to minimize error or loss.”) and FIG. 3B. Banerjee further discloses the iterative training (“retraining weights”) comprises training the encoder (“generator”) by maximizing a discriminator error using the training dataset and the trained discriminator (“discriminator neural network”) to obtain a trained encoder, the trained encoder being configured to invariantly encode real data and augmented data to a real representation space, such that representations preserve information about the real modality at: ¶¶ [0037]-[0041](“The vector representations can include vectors encoded from facial images within a training dataset… An error 398 is fed back to the synthetic vector generator 372 . The fed back error can be used to adjust weights associated with nodes and layers within the neural network that comprises the synthetic vector generator. The weights can be adjusted to maximize error or loss. That is, the generator generates synthetic vectors that successfully fool the discriminator into determining that the vectors are real. Each incorrect determination is an error , thus maximizing error by the discriminator improves the “quality” of the synthetic vectors to fool the discriminator.”) See, also, FIGS. 3B and 3C; ¶ [0027]; ¶¶ [0029]-[0031]. With regards to claim 2, Banerjee discloses the machine-learning model comprises further downstream layers (e.g., “machine learning neural network 170”), the further downstream layers with the trained encoder being configured by training to output analysis results (e.g., “drowsiness detection”) of input data at ¶¶ [0030]-[0031] and FIG. 1. With regards to claim 3, a multi-objective optimization problem is an optimization problem that involves multiple objective functions. Banerjee iterative training of the further downstream layers by a multi-objective optimization procedure using the training dataset and the trained encoder to obtain trained further downstream layers at: ¶¶ [0037]-[0038](“An error 366 is fed back to the discriminator 360. The fedback error can be used to adjust weights associated with nodes and layers within the neural network. The weights can be adjusted to minimize error or loss… The fed back error can be used to adjust weights associated with nodes and layers within the neural network that comprises the synthetic vector generator. The weights can be adjusted to maximize error or loss.”), ¶ [0041]. With regards to claim 5, Banerjee implicitly discloses receiving an input image and processing the input image using the trained machine-learning model to obtain analysis results of the input image when it discloses training a machine-learning model to obtain analysis results of an input image at ¶¶ [0030]-[0031] and FIG. 1. One of ordinary skill in the art would infer from this disclosure that a machine-learning model is trained with the intent that it be used to process input images in the manner it was trained. With regards to claim 12, Banerjee implicitly discloses receiving an input image and processing the input image using the trained machine-learning model to obtain analysis results of the input image when it discloses training a machine-learning model to obtain analysis results of an input image at ¶¶ [0030]-[0031] and FIG. 1. One of ordinary skill in the art would infer from this disclosure that a machine-learning model is trained with the intent that it be used to process input images in the manner it was trained. With respect to the training of the machine-learning model, the recited steps are anticipated by Banerjee for the same reasons as were provided in the rejection of claim 1 under 35 USC § 102 wherein these same steps were recited. With regards to claim 13, the steps performed by the apparatus of this claim are anticipated by Banerjee for the same reasons as were provided in the discussion of claim 1, which recites a method performing these same steps. With regards to claims 14 and 15, the steps implemented by computer executable instructions of these claims are anticipated by Banerjee for the same reasons as were provided in the discussion of claim 1, which recites a method performing these same steps . Claim Rejections - 35 USC § 103 07-20-aia AIA 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. 07-23-aia AIA 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. 07-21-aia AIA Claim 4 is rejected under 35 U.S.C. 103 as being unpatentable over Banerjee et al (US PG Pub. No. 2021/0201003) in view of Park et al (US PG Pub. No. 2024/0087265) . With regards to claim 4, Banerjee discloses the machine-learning model comprises further downstream layers (e.g., “machine learning neural network 170”), the further downstream layers with the trained encoder being configured by training to output analysis results (e.g., “drowsiness detection”) of input data at ¶¶ [0030]-[0031] and FIG. 1. Banerjee further discloses, “converting the additional synthetic vectors into image data 162” and “training the machine learning neural network … based on… converted synthetic images generated by the GAN generator…” However, Banerjee does not specify the further downstream layers comprise transformer encoders, reassembly modules, and fusion modules. However, this limitation was known in the art: Park discloses using a generative adversarial network to generate training vectors including further downstream layers comprising transformer encoders, reassembly modules, and fusion modules at: ¶ [0052], ¶ [0077], ¶ [0081], ¶ [0086]. At the time of filing of the present application, it would have been obvious to a person of ordinary skill in the art to use transformer encoders, reassembly modules, and fusion modules, as taught by Park, as a substitute for the unspecified manner of converting additional synthetic vectors into image data taught by Banerjee. This combination is a simple substitution of one known element for another to obtain predictable results. The prior art contained a method, taught by Banerjee, which differed from the claimed method by the substitution of the manner of converting synthetic vectors generated by a trained GAN into image data. The use of transformer encoders, reassembly modules, and fusion modules, and its functions were known in the art as evidenced by the Park reference. One of ordinary skill in the art could have substituted transformer encoders, reassembly modules, and fusion modules into the method taught by Banerjee and the results would have been predictable; to wit, the synthetic vectors generated by the encoder of Banerjee would be converted into image data . 07-21-aia AIA Claim s 6-8 are rejected under 35 U.S.C. 103 as being unpatentable over De-yi Ji et al (Chinese Pub. No. CN 113378668 A) in view of Banerjee et al (US PG Pub. No. 2021/0201003) . With regards to claim 6, De-yi Ji discloses a machine-learning model for water image segmentation, the input training dataset comprising water images in the English translation at pp.7-9, p.11, p.13. De-yi Ji does not specify the water images of the input training dataset comprise real training images and simulated training images. However, this limitation was known in the art as evidenced by Banerjee. Banerjee discloses images of an input training dataset comprising real training images and simulated training images at ¶ [0031] and ¶ [0040]. At the time of the filing of the present application, it would have been obvious to a person of ordinary skill in the art to use both real training images and simulated training images, as taught by Banerjee, when training a machine-learning model, as taught by De-yi Ji. The motivation for doing so comes from Banerjee, which discloses, “The example training dataset can be imbalanced, where the training dataset can include sufficient examples that include alert samples, while slightly drowsy samples and moderately/extremely drowsy samples can be sparse or insufficient. Discussed throughout, training, using the example dataset, can result in a biased classifier.” (¶ [0040]). One of ordinary skill in the art would infer from this disclosure, as a matter of common sense, that synthetic images could be used to enhance the data set of images when training a machine learning model. Therefore, it would have been obvious to combine Banerjee with De-yi Ji to obtain the invention specified in this claim. With regards to claim 7, De-yi Ji discloses receiving an input water image in real-time and processing the input water image using the trained machine-learning model to obtain water segmentation results of the input water image in the English translation at pp.7-9, p.11, p.13. The motivation for the combination is the same as previously presented. With regards to claim 8, De-yi Ji discloses responsive to determining that the water segmentation results exceed a threshold, outputting an alarm in the English translation at pp.7-9, p.11 (“In one specific example, the to-be-detected image in the area occupying the whole image area ratio is greater than the proportion threshold (such as more than 50 %),can generate alarm information to prompt the monitoring staff...”), p.13. The motivation for the combination is the same as previously presented . 07-21-aia AIA Claim s 9-11 are rejected under 35 U.S.C. 103 as being unpatentable over Kanan et al (US PG Pub. No. 2022/0051403) in view of Banerjee et al (US PG Pub. No. 2021/0201003) . With regards to claim 9, Kanan discloses a machine-learning model for medical image segmentation or classification, the input training dataset comprising training medical images and corresponding genomic training data at: ¶¶ [0083]-[0096]; in particular, see, ¶ [0084](“For training the machine learning system, one or more images may be paired with information about one or more continuous values… e.g., biomarkers from genomic testing…”); see, also, ¶ [0061], ¶ [0063], ¶ [0069]. Banerjee discloses images of an input training dataset comprising real training images and simulated training images at ¶ [0031] and ¶ [0040]. The motivation for the combination is the same as previously presented. With regards to claim 10, Kanan discloses receiving an input medical image in real-time and processing the input medical image using the trained machine-learning model to obtain medical segmentation or classification results of the input medical image at: ¶¶ [0083]-[0096]; see, also, ¶ [0061], ¶ [0063], ¶ [0069]. The motivation for the combination is the same as previously presented. With regards to claim 11, Kanan discloses, responsive to determining that the medical segmentation or classification results exceed a threshold, outputting an alarm at ¶¶ [0083]-[0096]; in particular, see, ¶ [0096](“In step 418, the method may include, upon determining a biomarker is present, outputting the prediction to a digital storage device. Optionally, the method may also include using a visual indicator to alert a user (e.g., a pathologist, histology technician, radiologist etc.) to the level of the continuous scores.”); see, also, ¶ [0061], ¶ [0063], ¶ [0069]. The motivation for the combination is the same as previously presented. Conclusion Any inquiry concerning this communication or earlier communications from the examiner should be directed to DAVID F DUNPHY whose telephone number is (571)270-1230. The examiner can normally be reached 9 am - 5 pm. 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, Chineyere Wills-Burns can be reached at (571) 272-9752. 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. /DAVID F DUNPHY/Primary Examiner, Art Unit 2673 Application/Control Number: 18/885,949 Page 2 Art Unit: 2673 Application/Control Number: 18/885,949 Page 3 Art Unit: 2673 Application/Control Number: 18/885,949 Page 4 Art Unit: 2673 Application/Control Number: 18/885,949 Page 5 Art Unit: 2673 Application/Control Number: 18/885,949 Page 6 Art Unit: 2673 Application/Control Number: 18/885,949 Page 7 Art Unit: 2673 Application/Control Number: 18/885,949 Page 8 Art Unit: 2673 Application/Control Number: 18/885,949 Page 9 Art Unit: 2673
Prosecution Insights