CTNF 18/560,950 CTNF 87584 DETAILED ACTION This action is responsive to communications filed on November 15, 2023. This action is made Non-Final . Claims 1-20 are pending in the case. Claims 1 and 9 are independent claims. Claims 1-10, 13, and 15-20 are rejected. 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. Information Disclosure Statement The information disclosure statement (IDS(s)) submitted on 11/15/2023 is/are in compliance with the provisions of 37 C.F.R. 1.97. Accordingly, the IDS(s) is/are being considered by the examiner. Priority 02-26 AIA Receipt is acknowledged of papers submitted under 35 U.S.C. 119(a)-(d), which papers have been placed of record in the file. 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. Claims 1, 2, and 4-8 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception (i.e., a law of nature, a natural phenomenon, or an abstract idea) without significantly more. Step 1: Independent claims 1 and 9 are directed towards a non-transitory medium and a method, respectively. Therefore, these claims, as well as their dependent claims, are directed towards one of the four statutory categories (process, machine (i.e. apparatus), manufacture, or composition of matter. With further respect to claim 1: 2A Prong 1: Claim 1 recites the following judicial exceptions: perform a method of performing an analysis on digital information to be analyzed, the method comprising: ... performing the analysis on the digital information to be analyzed to generate an analysis result (mental process –can be performed in the human mind, or by a human using a pen and paper (e.g. a person may evaluate digital information and produce a corresponding result). 2A Prong 2: The additional elements recited in the claim do not integrate the judicial exception into a practical application. Additional elements: A non-transitory computer readable medium storing instructions readable and executable by at least one electronic processor ... receiving a cryptographic key; constructing an input dataset, the input dataset including both the digital information and the cryptographic key ... outputting the analysis result (mere instructions to apply the exception or implement the exception on a computer (e.g. using a computer to receive data, generate further data, and outputting; see MPEP §2106.05(f).). The additional elements do not effectively integrate the abstract idea into a practical application. 2B: revisiting the additional elements, the additional elements do not amount to significantly more than the judicial exception – recited high level of generality and corresponds to storing and retrieving information in memory and performing calculations). by applying an artificial neural network (ANN) to the input dataset (generally linking the use of a judicial exception to a particular technological environment or field of use (e.g. using a neural network on an input; see MPEP §2106.05(h).). The additional elements do not effectively integrate the abstract idea into a practical application. 2B: revisiting the additional elements, the additional elements do not amount to significantly more than the judicial exception – recited high level of generality and corresponds to storing and retrieving information and performing calculations.). With further respect to claim 2: 2A Prong 1: Claim 2 recites the following judicial exceptions: generate a correct analysis result for the digital information to be analyzed if the cryptographic key matches a valid cryptographic key; and generate an incorrect analysis result for the digital information to be analyzed if the cryptographic key does not match the valid cryptographic key (mental process –can be performed in the human mind, or by a human using a pen and paper (e.g. a person may evaluate digital information and produce a corresponding result). 2A Prong 2: The additional elements recited in the claim do not integrate the judicial exception into a practical application. Additional elements: wherein the ANN is trained to (generally linking the use of a judicial exception to a particular technological environment or field of use (e.g. using a neural network on an input; see MPEP §2106.05(h).). The additional elements do not effectively integrate the abstract idea into a practical application. 2B: revisiting the additional elements, the additional elements do not amount to significantly more than the judicial exception – recited high level of generality and corresponds to storing and retrieving information and performing calculations.). With further respect to claim 4: 2A Prong 1: Claim 4 recites the following judicial exceptions: wherein the analysis result includes a key validity output indicator indicating whether the cryptographic key matches the valid cryptographic key (mental process –can be performed in the human mind, or by a human using a pen and paper (e.g. a person may evaluate digital information and produce a corresponding result). With further respect to claim 5: 2A Prong 1: Claim 5 recites the following judicial exceptions: wherein the method of performing the analysis on the digital information to be analyzed does not employ homomorphic encryption (mental process –can be performed in the human mind, or by a human using a pen and paper (e.g. a person may evaluate digital information and produce a corresponding result). With further respect to claim 6: 2A Prong 1: Claim 6 recites the following judicial exceptions: wherein the digital information to be analyzed comprises a digital image and the analysis is an image processing analysis (mental process –can be performed in the human mind, or by a human using a pen and paper (e.g. a person may evaluate digital information and produce a corresponding result). With further respect to claim 7: 2A Prong 1: Claim 7 recites the following judicial exceptions: wherein the digital information to be analyzed comprises medical information of a subject and the analysis is computer-aided diagnosis (CADx) analysis (mental process –can be performed in the human mind, or by a human using a pen and paper (e.g. a person may evaluate digital information and produce a corresponding result). With further respect to claim 8: 2A Prong 1: Claim 8 recites the following judicial exceptions: wherein the further includes analyzing log files of a medical imaging device (mental process –can be performed in the human mind, or by a human using a pen and paper (e.g. a person may evaluate digital information and produce a corresponding result). 2B continued: After considering all claim elements individually and as an ordered combination, it is determined that the claims do not include any additional elements that are sufficient to amount to significantly more than the judicial exception. 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 – 07-08-aia AIA (a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention. 07-15-aia AIA Claim(s) 1, 2, 4-6, 9, 10, 13, and 15-17 is/are rejected under 35 U.S.C. 102 (a)(1) as being anticipated by AprilPyone, MaungMaung, and Hitoshi Kiya. "Training DNN Model with Secret Key for Model Protection." arXiv preprint arXiv:2008.02450 (2020) . Claim 1: AprilPyone discloses a non-transitory computer readable medium storing instructions readable and executable by at least one electronic processor to perform a method of performing an analysis on digital information to be analyzed, the method comprising: receiving a cryptographic key (see Fig. 1-3; Abstract - performance of the protected model is close to that of non-protected models when the key is correct, while the accuracy is severely dropped when an incorrect key is given; §A. Overview - model (f) is trained by preprocessed images with a key (K) for the first time; §E. Requirements - rightful user with key K can access the model without any noticeable overhead in both training and inference time, and performance degradation; §B. Overview - When key K was given, the performance accuracy was closer to the baseline accuracy.); constructing an input dataset, the input dataset including both the digital information to analyzed (see Fig. 1-3; Table 1; §A. Overview - test images are also preprocessed with the same key K before testing; §B. Results – verify the effectiveness of the proposed model protection method, we tested the protected models against a wrong key K’, plain images (without preprocessing), and fine-tuning attacks to adapt a wrong key K’.); performing the analysis on the digital information to be analyzed to generate an analysis result by applying an artificial neural network (ANN) to the input dataset (see Fig1-3; Table 1; §A. Overview - test images are also preprocessed with the same key K before testing; §A. Experiment Set-up - used deep residual networks; §B. Results – verify the effectiveness of the proposed model protection method, we tested the protected models against a wrong key K’, plain images (without preprocessing), and fine-tuning attacks to adapt a wrong key K’.); and outputting the analysis result (see Fig1-3; Table 1; §A. Overview - test images are also preprocessed with the same key K before testing; §A. Experiment Set-up - used deep residual networks; §B. Results – verify the effectiveness of the proposed model protection method, we tested the protected models against a wrong key K’, plain images (without preprocessing), and fine-tuning attacks to adapt a wrong key K’. When key was given, the performance accuracy was closer to the baseline accuracy. When the key was not correct, the accuracy was drastically decreased.). Claim 2: AprilPyone further discloses wherein the ANN is trained to: generate a correct analysis result for the digital information to be analyzed if the cryptographic key matches a valid cryptographic key; and generate an incorrect analysis result for the digital information to be analyzed if the cryptographic key does not match the valid cryptographic key (see Fig1-3; Table 1; §A. Overview - test images are also preprocessed with the same key K before testing; §A. Experiment Set-up - used deep residual networks; §B. Results – trained three protected models by preprocessed images with a secret key in different block sizes (M 2 f2; 4; 8g), and also a non-protected model (i.e., baseline model). verify the effectiveness of the proposed model protection method, we tested the protected models against a wrong key K’, plain images (without preprocessing), and fine-tuning attacks to adapt a wrong key K’. When key was given, the performance accuracy was closer to the baseline accuracy. When the key was not correct, the accuracy was drastically decreased.). Claim 4: AprilPyone further discloses wherein the analysis result includes a key validity output indicating whether the cryptographic key matches the valid cryptographic key (see Fig1-3; Table 1; Abstract - performance of the protected model is close to that of non-protected models when the key is correct, while the accuracy is severely dropped when an incorrect key is given; §A. Overview - model (f) is trained by preprocessed images with a key (K) for the first time; §E. Requirements - rightful user with key K can access the model without any noticeable overhead in both training and inference time, and performance degradation; §B. Overview - When key K was given, the performance accuracy was closer to the baseline accuracy; §A. Overview - test images are also preprocessed with the same key K before testing; §A. Experiment Set-up - used deep residual networks; §B. Results – trained three protected models by preprocessed images with a secret key in different block sizes (M 2 f2; 4; 8g), and also a non-protected model (i.e., baseline model). verify the effectiveness of the proposed model protection method, we tested the protected models against a wrong key K’, plain images (without preprocessing), and fine-tuning attacks to adapt a wrong key K’. When key was given, the performance accuracy was closer to the baseline accuracy. When the key was not correct, the accuracy was drastically decreased.). Claim 5: AprilPyone further discloses wherein the method of performing analysis on the digital information to be analyzed does not employ homomorphic encryption (see Fig 1-3; Table I; §I – model protection method with a key in such a way that a stolen model cannot be used without the key for the first time. Specifically, the proposed method preprocesses input images with a secret key and trains the model by using such preprocessed images.). Claim 6: AprilPyone wherein the digital information to be analyzed comprises a digital image and the analysis is an image processing analysis (see Fig1-3; Table 1; Abstract - performance of the protected model is close to that of non-protected models when the key is correct, while the accuracy is severely dropped when an incorrect key is given; §A. Overview - model (f) is trained by preprocessed images with a key (K) for the first time; §E. Requirements - rightful user with key K can access the model without any noticeable overhead in both training and inference time, and performance degradation. Unusability: Ideally, stolen models should not be usable in any case without key K. In addition, even when the adversary retrains a stolen model with a forged key, the performance of the model should be heavily dropped; §B. Overview - When key K was given, the performance accuracy was closer to the baseline accuracy; §A. Overview - test images are also preprocessed with the same key K before testing; §A. Experiment Set-up - used deep residual networks; §B. Results – trained three protected models by preprocessed images with a secret key in different block sizes (M 2 f2; 4; 8g), and also a non-protected model (i.e., baseline model). verify the effectiveness of the proposed model protection method, we tested the protected models against a wrong key K’, plain images (without preprocessing), and fine-tuning attacks to adapt a wrong key K’. When key was given, the performance accuracy was closer to the baseline accuracy. When the key was not correct, the accuracy was drastically decreased.). Claim 9: AprilPyone discloses a method of simultaneously training and securing an artificial neural network (ANN), the method comprising: generating a trained ANN for performing analysis, including: performing a plurality of valid-key training cycles on the ANN with datasets, each dataset including digital information to be analyzed and a valid cryptographic key, wherein the valid-key training cycles employ an analysis objective function that drives the valid-key training cycles to produce a correct analysis result for the digital information to be analyzed (see Fig1-3; Table 1; Abstract - performance of the protected model is close to that of non-protected models when the key is correct, while the accuracy is severely dropped when an incorrect key is given; §A. Overview - test images are also preprocessed with the same key K before testing; §E. Requirements - rightful user with key K can access the model without any noticeable overhead in both training and inference time, and performance degradation. Unusability: Ideally, stolen models should not be usable in any case without key K. In addition, even when the adversary retrains a stolen model with a forged key, the performance of the model should be heavily dropped; §A. Experiment Set-up - We used deep residual networks [20] with 18 layers (ResNet18) and trained for 200 epochs with cyclic learning rates [21] and mixed precision training; §B. Results – trained three protected models by preprocessed images with a secret key in different block sizes (M 2 f2; 4; 8g), and also a non-protected model (i.e., baseline model). verify the effectiveness of the proposed model protection method, we tested the protected models against a wrong key K’, plain images (without preprocessing), and fine-tuning attacks to adapt a wrong key K’. When key was given, the performance accuracy was closer to the baseline accuracy. When the key was not correct, the accuracy was drastically decreased.); performing a plurality of invalid-key training cycles on the ANN with datasets, each dataset including digital information to be analyzed and an invalid cryptographic key, wherein the invalid-key training cycles employ a security objective function that drives the invalid-key training cycles to produce an incorrect analysis result for the digital information to be analyzed (see Fig1-3; Table 1; Abstract - performance of the protected model is close to that of non-protected models when the key is correct, while the accuracy is severely dropped when an incorrect key is given; §A. Overview - test images are also preprocessed with the same key K before testing; §E. Requirements - rightful user with key K can access the model without any noticeable overhead in both training and inference time, and performance degradation. Unusability: Ideally, stolen models should not be usable in any case without key K. In addition, even when the adversary retrains a stolen model with a forged key, the performance of the model should be heavily dropped; §A. Experiment Set-up - We used deep residual networks [20] with 18 layers (ResNet18) and trained for 200 epochs with cyclic learning rates [21] and mixed precision training; §B. Results – trained three protected models by preprocessed images with a secret key in different block sizes (M 2 f2; 4; 8g), and also a non-protected model (i.e., baseline model). verify the effectiveness of the proposed model protection method, we tested the protected models against a wrong key K’, plain images (without preprocessing), and fine-tuning attacks to adapt a wrong key K’. When key was given, the performance accuracy was closer to the baseline accuracy. When the key was not correct, the accuracy was drastically decreased.); and storing the trained ANN on a non-transitory storage medium (see Fig1-3; Table 1; Abstract - performance of the protected model is close to that of non-protected models when the key is correct, while the accuracy is severely dropped when an incorrect key is given; §A. Overview - test images are also preprocessed with the same key K before testing; §E. Requirements - rightful user with key K can access the model without any noticeable overhead in both training and inference time, and performance degradation. Unusability: Ideally, stolen models should not be usable in any case without key K. In addition, even when the adversary retrains a stolen model with a forged key, the performance of the model should be heavily dropped; §A. Experiment Set-up - We used deep residual networks [20] with 18 layers (ResNet18) and trained for 200 epochs with cyclic learning rates [21] and mixed precision training; §B. Results – trained three protected models by preprocessed images with a secret key in different block sizes (M 2 f2; 4; 8g), and also a non-protected model (i.e., baseline model). verify the effectiveness of the proposed model protection method, we tested the protected models against a wrong key K’, plain images (without preprocessing), and fine-tuning attacks to adapt a wrong key K’. When key was given, the performance accuracy was closer to the baseline accuracy. When the key was not correct, the accuracy was drastically decreased.). Claim 10: Jackson discloses wherein a number of the plurality of valid-key training cycles is higher than a number of the plurality of invalid-key training cycles (see Fig1-3; Table 1; Abstract - performance of the protected model is close to that of non-protected models when the key is correct, while the accuracy is severely dropped when an incorrect key is given; §A. Overview - test images are also preprocessed with the same key K before testing; §E. Requirements - rightful user with key K can access the model without any noticeable overhead in both training and inference time, and performance degradation. Unusability: Ideally, stolen models should not be usable in any case without key K. In addition, even when the adversary retrains a stolen model with a forged key, the performance of the model should be heavily dropped; §A. Experiment Set-up - We used deep residual networks [20] with 18 layers (ResNet18) and trained for 200 epochs with cyclic learning rates [21] and mixed precision training; §B. Results – trained three protected models by preprocessed images with a secret key in different block sizes (M 2 f2; 4; 8g), and also a non-protected model (i.e., baseline model). verify the effectiveness of the proposed model protection method, we tested the protected models against a wrong key K’, plain images (without preprocessing), and fine-tuning attacks to adapt a wrong key K’. When key was given, the performance accuracy was closer to the baseline accuracy. When the key was not correct, the accuracy was drastically decreased.). Claim 13: AprilPyone discloses generating an analysis result for input digital information to be analyzed by retrieving the trained ANN from the non-transitory storage medium and applying the trained ANN to a dataset that includes both the input digital information to be analyzed and an input cryptographic key; and displaying the result (see Fig1-3; Table 1; Abstract - performance of the protected model is close to that of non-protected models when the key is correct, while the accuracy is severely dropped when an incorrect key is given; §A. Overview - test images are also preprocessed with the same key K before testing; §E. Requirements - rightful user with key K can access the model without any noticeable overhead in both training and inference time, and performance degradation. Unusability: Ideally, stolen models should not be usable in any case without key K. In addition, even when the adversary retrains a stolen model with a forged key, the performance of the model should be heavily dropped; §A. Experiment Set-up - We used deep residual networks [20] with 18 layers (ResNet18) and trained for 200 epochs with cyclic learning rates [21] and mixed precision training; §B. Results – trained three protected models by preprocessed images with a secret key in different block sizes (M 2 f2; 4; 8g), and also a non-protected model (i.e., baseline model). verify the effectiveness of the proposed model protection method, we tested the protected models against a wrong key K’, plain images (without preprocessing), and fine-tuning attacks to adapt a wrong key K’. When key was given, the performance accuracy was closer to the baseline accuracy. When the key was not correct, the accuracy was drastically decreased.). Claim 15: AprilPyone discloses wherein digital information comprises images and the analysis is an image processing analysis (see Fig1-3; Table 1; Abstract - performance of the protected model is close to that of non-protected models when the key is correct, while the accuracy is severely dropped when an incorrect key is given; §A. Overview - test images are also preprocessed with the same key K before testing; §E. Requirements - rightful user with key K can access the model without any noticeable overhead in both training and inference time, and performance degradation. Unusability: Ideally, stolen models should not be usable in any case without key K. In addition, even when the adversary retrains a stolen model with a forged key, the performance of the model should be heavily dropped; §A. Experiment Set-up - We used deep residual networks [20] with 18 layers (ResNet18) and trained for 200 epochs with cyclic learning rates [21] and mixed precision training; §B. Results – trained three protected models by preprocessed images with a secret key in different block sizes (M 2 f2; 4; 8g), and also a non-protected model (i.e., baseline model). verify the effectiveness of the proposed model protection method, we tested the protected models against a wrong key K’, plain images (without preprocessing), and fine-tuning attacks to adapt a wrong key K’. When key was given, the performance accuracy was closer to the baseline accuracy. When the key was not correct, the accuracy was drastically decreased.). Claim 16: AprilPyone discloses wherein the trained ANN has a validity output indicating whether the input cryptographic key matches the valid cryptographic key (see Fig1-3; Table 1; Abstract - performance of the protected model is close to that of non-protected models when the key is correct, while the accuracy is severely dropped when an incorrect key is given; §A. Overview - test images are also preprocessed with the same key K before testing; §E. Requirements - rightful user with key K can access the model without any noticeable overhead in both training and inference time, and performance degradation. Unusability: Ideally, stolen models should not be usable in any case without key K. In addition, even when the adversary retrains a stolen model with a forged key, the performance of the model should be heavily dropped; §A. Experiment Set-up - We used deep residual networks [20] with 18 layers (ResNet18) and trained for 200 epochs with cyclic learning rates [21] and mixed precision training; §B. Results – trained three protected models by preprocessed images with a secret key in different block sizes (M 2 f2; 4; 8g), and also a non-protected model (i.e., baseline model). verify the effectiveness of the proposed model protection method, we tested the protected models against a wrong key K’, plain images (without preprocessing), and fine-tuning attacks to adapt a wrong key K’. When key was given, the performance accuracy was closer to the baseline accuracy. When the key was not correct, the accuracy was drastically decreased.). Claim 17: AprilPyone discloses wherein the ANN comprises a multilayer ANN (see Fig1-3; Table 1; Abstract - performance of the protected model is close to that of non-protected models when the key is correct, while the accuracy is severely dropped when an incorrect key is given; §A. Overview - test images are also preprocessed with the same key K before testing; §E. Requirements - rightful user with key K can access the model without any noticeable overhead in both training and inference time, and performance degradation. Unusability: Ideally, stolen models should not be usable in any case without key K. In addition, even when the adversary retrains a stolen model with a forged key, the performance of the model should be heavily dropped; §A. Experiment Set-up - We used deep residual networks [20] with 18 layers (ResNet18) and trained for 200 epochs with cyclic learning rates [21] and mixed precision training; §B. Results – trained three protected models by preprocessed images with a secret key in different block sizes (M 2 f2; 4; 8g), and also a non-protected model (i.e., baseline model). verify the effectiveness of the proposed model protection method, we tested the protected models against a wrong key K’, plain images (without preprocessing), and fine-tuning attacks to adapt a wrong key K’. When key was given, the performance accuracy was closer to the baseline accuracy. When the key was not correct, the accuracy was drastically decreased.) . 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. 07-21-aia AIA Claim (s) 3 and 19 is/are rejected under 35 U.S.C. 103 as being unpatentable over AprilPyone, and further in view of Furukawa et al., US Publication 2020/0293944 (“Furukawa”) . Claim 3: AprilPyone further teaches or suggests wherein the method further comprises: training the ANN on both (i) valid-key input datasets that include the valid cryptographic key and (ii) invalid-key input datasets ...; wherein the training employs an objective function that drives the training to generate correct analysis results for the valid-key input datasets and that drives the training to generate incorrect analysis results for the invalid-key input datasets (see Fig1-3; Table 1; §A. Overview - test images are also preprocessed with the same key K before testing; §A. Experiment Set-up - used deep residual networks; §B. Results – trained three protected models by preprocessed images with a secret key in different block sizes (M 2 f2; 4; 8g), and also a non-protected model (i.e., baseline model). verify the effectiveness of the proposed model protection method, we tested the protected models against a wrong key K’, plain images (without preprocessing), and fine-tuning attacks to adapt a wrong key K’. When key was given, the performance accuracy was closer to the baseline accuracy. When the key was not correct, the accuracy was drastically decreased; §C. Robustness – retrains the model with a forged key (K0) for 30 epochs. We ran an experiment with different sizes of the adversary’s dataset.; §D. Future - try many different keys or improve a key by observing the accuracy or the loss; §V - fine-tuning attacks considering the adversary has a small subset of training dataset to adapt a new forged key.). AprilPyone does not explicitly disclose that include randomly or pseudorandomly generated cryptographic keys . Furukawa teaches or suggests that include randomly or pseudorandomly generated cryptographic keys (see para. 0033 - cryptographic keys are generated, for example, randomly. The length of the string may be selected to be sufficiently long such that the probability of randomly computing the respective cryptographic key using realistically available computational sources is sufficiently low; para. 0035 - adjusting each of the data item instances according to a respective unique crypto graphic key. Each adjusted instance is inputted into a respect tive sub-classifier. The classification outputs and/or confidence levels outputted by the sub-classifiers are analyzed to compute a single classification outcome; para. 0037 - set of rules define a distribution of confidence levels outputted by the multiple trained subclassifiers indicative of fabricated input; para. 0081 – cryptographic keys may be generated, for example, by a random generation process, and/or based on a cryptographic process such as a hash of a data input such as randomly generated data.). Accordingly, it would have been obvious to one having ordinary skill before the effective filing date of the claimed invention to modify the system and method, taught in AprilPyone, to include that include randomly or pseudorandomly generated cryptographic keys for the purpose of efficiently securing classifiers using cryptographic keys generated using random generation processes, improving model security, as taught by Furukawa (0033, 0035, 0081). Claim 19: Furukawa further teaches or suggests wherein the valid cryptographic key comprises one of: a binary mask, a vector, a two-dimensional array, or a three-dimensional array (see para. 0017 - and the cryptographic key denotes a condition vector having a size according to the size of the direction vector and the parameter vector, wherein a product of the two dimensional matrix and the parameter vector is equal to the condition vector when the condition function computed according to the respective cryptographic key and the parameter vector is equal to zero; para. 0033 - cryptographic keys are generated, for example, randomly. The length of the string may be selected to be sufficiently long such that the probability of randomly computing the respective cryptographic key using realistically available computational sources is sufficiently low; para. 0035 - adjusting each of the data item instances according to a respective unique crypto graphic key. Each adjusted instance is inputted into a respect tive sub-classifier. The classification outputs and/or confidence levels outputted by the sub-classifiers are analyzed to compute a single classification outcome; para. 0037 - set of rules define a distribution of confidence levels outputted by the multiple trained subclassifiers indicative of fabricated input; para. 0081 – cryptographic keys may be generated, for example, by a random generation process, and/or based on a cryptographic process such as a hash of a data input such as randomly generated data; para. 0113 – cryptographic key denotes a condition vector having a size according to the size of the direction vector and the parameter vector.). Accordingly, it would have been obvious to one having ordinary skill before the effective filing date of the claimed invention to modify the system and method, taught in AprilPyone, to include wherein the valid cryptographic key comprises one of: a binary mask, a vector, a two-dimensional array, or a three-dimensional array for the purpose of efficiently securing classifiers using cryptographic keys generated using random generation processes, improving model security, as taught by Furukawa (0033, 0035, 0081) . 07-21-aia AIA Claim (s) 7, 8, and 18 is/are rejected under 35 U.S.C. 103 as being unpatentable over AprilPyone, and further in view of Jackson et al., US Patent 12,249,063 (“Jackson”) . Claim 7: Jackson further teaches or suggests wherein the digital information to be analyzed comprises medical information of a subject and the analysis is a computer-aided diagnosis (CADx) analysis (see Fig. 1-3; col. 4, lines 25-55 - providing a computerized, rapid, virtual diagnostic companion, used by practitioners in analyzing and diagnosing conditions; convolutional neural network CNN, running on a processor, is trained based upon images of a plurality of the tissue sample slides; col. 5, lines 23-33 - can be instantiated using any acceptable data handling modality-for example, a PC, laptop, tablet or smartphone. In illustrative embodiments, the results can include at least one of color-coding, graphics, statistical, and textual information. secure channels, encryption, etc.; col. 6, lines 42-48 - slide imager 120 is used in the field to image patient slides for diagnosis in runtime, and this data 130 is thereby presented to the processor 110. The other slide 45 imager 122 can be part of a system that produces a large volume of slide image data based upon various types of cells and/or conditions. This data 132 is part of a training set that is input to the processor 110 for use in construction a CNN; col. 7, lines 54-56 - CNN and any associated image data is stored (step 220) with respect to the process(or) 110 for use in 55 follow-on runtime operations; col. 15, lines 41-43 - download such data (e.g. via the Internet or physical storage media-thumbdrives, etc.) to a subscription site that performs the analysis.). Accordingly, it would have been obvious to one having ordinary skill before the effective filing date of the claimed invention to modify the system and method, taught in AprilPyone, to include wherein the digital information to be analyzed comprises medical information of a subject and the analysis is a computer-aided diagnosis (CADx) analysis for the purpose of efficiently providing a virtual diagnostic companion for compuiterized and rapid analysis, improving diagnosis of conditions, as taught by Jakson (col. 4, 7, 15). Claim 8: Jackson further teaches or suggests wherein the method further includes analyzing log files of a medical imaging device (see Fig. 1-3; col. 4, lines 25-55 - providing a computerized, rapid, virtual diagnostic companion, used by practitioners in analyzing and diagnosing conditions; convolutional neural network CNN, running on a processor, is trained based upon images of a plurality of the tissue sample slides; col. 5, lines 23-33 - can be instantiated using any acceptable data handling modality-for example, a PC, laptop, tablet or smartphone. In illustrative embodiments, the results can include at least one of color-coding, graphics, statistical, and textual information. secure channels, encryption, etc.; col. 6, lines 42-48 - slide imager 120 is used in the field to image patient slides for diagnosis in runtime, and this data 130 is thereby presented to the processor 110. The other slide 45 imager 122 can be part of a system that produces a large volume of slide image data based upon various types of cells and/or conditions. This data 132 is part of a training set that is input to the processor 110 for use in construction a CNN; col. 7, lines 54-56 - CNN and any associated image data is stored (step 220) with respect to the process(or) 110 for use in 55 follow-on runtime operations; col. 15, lines 41-43 - download such data (e.g. via the Internet or physical storage media-thumbdrives, etc.) to a subscription site that performs the analysis.). Accordingly, it would have been obvious to one having ordinary skill before the effective filing date of the claimed invention to modify the system and method, taught in AprilPyone, to include wherein the method further includes analyzing log files of a medical imaging device for the purpose of efficiently providing a virtual diagnostic companion for compuiterized and rapid analysis, improving diagnosis of conditions, as taught by Jakson (col. 4, 7, 15). Claim 18: Jackson further teaches or suggests wherein the ANN comprises a convolutional NN (CNN) (see Fig. 1-3; col. 4, lines 25-55 - providing a computerized, rapid, virtual diagnostic companion, used by practitioners in analyzing and diagnosing conditions; convolutional neural network CNN, running on a processor, is trained based upon images of a plurality of the tissue sample slides; col. 5, lines 23-33 - can be instantiated using any acceptable data handling modality-for example, a PC, laptop, tablet or smartphone. In illustrative embodiments, the results can include at least one of color-coding, graphics, statistical, and textual information. secure channels, encryption, etc.; col. 6, lines 42-48 - slide imager 120 is used in the field to image patient slides for diagnosis in runtime, and this data 130 is thereby presented to the processor 110. The other slide 45 imager 122 can be part of a system that produces a large volume of slide image data based upon various types of cells and/or conditions. This data 132 is part of a training set that is input to the processor 110 for use in construction a CNN; col. 7, lines 54-56 - CNN and any associated image data is stored (step 220) with respect to the process(or) 110 for use in 55 follow-on runtime operations; col. 15, lines 41-43 - download such data (e.g. via the Internet or physical storage media-thumbdrives, etc.) to a subscription site that performs the analysis.). Accordingly, it would have been obvious to one having ordinary skill before the effective filing date of the claimed invention to modify the system and method, taught in AprilPyone, to include wherein the ANN comprises a convolutional NN (CNN) for the purpose of efficiently providing a virtual diagnostic companion for compuiterized and rapid analysis, improving diagnosis of conditions, as taught by Jakson (col. 4, 7, 15) . Allowable Subject Matter 12-151-08 AIA 07-43 12-51-08 Claim s 11, 12, 14, and 20 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. Conclusion Any inquiry concerning this communication or earlier communications from the examiner should be directed to Andrew T McIntosh whose telephone number is (571)270-7790. The examiner can normally be reached M-Th 8:00am-5:30pm. 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, Tamara Kyle can be reached at 571-272-4241. 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. /ANDREW T MCINTOSH/Primary Examiner, Art Unit 2144 Application/Control Number: 18/560,950 Page 2 Art Unit: 2144 Application/Control Number: 18/560,950 Page 3 Art Unit: 2144 Application/Control Number: 18/560,950 Page 4 Art Unit: 2144 Application/Control Number: 18/560,950 Page 5 Art Unit: 2144 Application/Control Number: 18/560,950 Page 6 Art Unit: 2144 Application/Control Number: 18/560,950 Page 7 Art Unit: 2144 Application/Control Number: 18/560,950 Page 8 Art Unit: 2144 Application/Control Number: 18/560,950 Page 9 Art Unit: 2144 Application/Control Number: 18/560,950 Page 10 Art Unit: 2144 Application/Control Number: 18/560,950 Page 11 Art Unit: 2144 Application/Control Number: 18/560,950 Page 12 Art Unit: 2144 Application/Control Number: 18/560,950 Page 13 Art Unit: 2144 Application/Control Number: 18/560,950 Page 14 Art Unit: 2144 Application/Control Number: 18/560,950 Page 15 Art Unit: 2144 Application/Control Number: 18/560,950 Page 16 Art Unit: 2144 Application/Control Number: 18/560,950 Page 17 Art Unit: 2144 Application/Control Number: 18/560,950 Page 18 Art Unit: 2144 Application/Control Number: 18/560,950 Page 19 Art Unit: 2144 Application/Control Number: 18/560,950 Page 20 Art Unit: 2144 Application/Control Number: 18/560,950 Page 21 Art Unit: 2144 Application/Control Number: 18/560,950 Page 22 Art Unit: 2144 Application/Control Number: 18/560,950 Page 23 Art Unit: 2144 Application/Control Number: 18/560,950 Page 24 Art Unit: 2144 Application/Control Number: 18/560,950 Page 25 Art Unit: 2144 Application/Control Number: 18/560,950 Page 26 Art Unit: 2144 Application/Control Number: 18/560,950 Page 27 Art Unit: 2144
Prosecution Insights