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and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114.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 04/09/2026 has been entered.
Response to Amendment
Claims 1-19 remain pending within the application.
Claim Rejections - 35 USC § 101
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.
The claims 1-4, 6, and 8-19 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: Is the claim directed to a process, machine, manufacture, or composition of matter?
Claims 1-4, 6, and 8-14 are directed to a system, hence fall within the statutory category of a machine.
Claims 15-18 are directed to a method, hence fall within the statutory category of a process.
Claim 19 is directed to a non-transitory computer readable medium, hence fall within the statutory category of a machine.
Thus, each of the claims fall within one of the four statutory categories.
Claim 1 includes the steps of:
A system, comprising: at least one data processor; and at least one memory storing instructions which, when executed by the at least one data processor, result in operations comprising:
retrieving a set of authentic electronic medical records from a database;
converting the authentic set of electronic medical records to a set of numerical vectors;
creating a first training set based on a random noise generator sample;
training a first neural network using the first training set, the first neural network outputting synthetic electronic medical records;
creating a second training set based on the synthetic electronic medical records and the set of numerical vectors;
training a second neural network using the second training set, the second neural network outputting a loss distribution, the loss distribution indicating whether the output synthetic electronic medical records are classified as authentic or synthetic,
wherein training the first neural network further comprises updating a first gradient of the first neural network by descending the first gradient based on the loss distribution,
wherein training the second neural network further comprises updating a second gradient of the second neural network by ascending the second gradient based on the loss distribution.
Step 2A Prong 1: Does the claim recite an abstract idea, law of nature, or natural phenomenon?
The broadest reasonable interpretation of the following limitations falls within the mental process groupings of abstract ideas because they cover concepts performed in the human mind, including observation, evaluation, judgment, and opinion. See MPEP 2106.04(a)(2), subsection III. The claim(s) recite(s) in part:
“converting the authentic set of electronic medical records to a set of numerical vectors”. As drafted and under its broadest reasonable interpretation, this limitation recites an abstract idea of a mental process because converting medical records to numerical vectors encompasses mental evaluations that are practically performed in the human mind, but for the recitation of generic computer components. Even if most humans would use a physical aid, like a pen and paper or a calculator, to make such evaluations, the use of a physical aid would not negate the mental nature of this limitation. See MPEP 2106.04(a)(2), subsection III.B.
“creating a first training set based on a random noise generator sample” . As drafted and under its broadest reasonable interpretation, this limitation recites an abstract idea of a mental process because creating training sets based on random noise generator sample encompasses mental evaluations that are practically performed in the human mind, but for the recitation of generic computer components. Even if most humans would use a physical aid, like a pen and paper or a calculator, to make such evaluations, the use of a physical aid would not negate the mental nature of this limitation. See MPEP 2106.04(a)(2), subsection III.B. For example, one could reasonably alter training set data by hand based on random noise samples to create a first training set.
“creating a second training set based on the synthetic electronic medical records and the set of numerical vectors”. As drafted and under its broadest reasonable interpretation, this limitation recites an abstract idea of a mental process because creating training sets based on synthetic records and numerical vectors encompasses mental evaluations that are practically performed in the human mind, but for the recitation of generic computer components. Even if most humans would use a physical aid, like a pen and paper or a calculator, to make such evaluations, the use of a physical aid would not negate the mental nature of this limitation. See MPEP 2106.04(a)(2), subsection III.B. For example, one could reasonably alter training set data by combining synthetic records with numerical data.
“wherein training the first neural network further comprises updating a first gradient of the first neural network by descending the first gradient based on the loss distribution, wherein training the second neural network further comprises updating a second gradient of the second neural network by ascending the second gradient based on the loss distribution”. As drafted and under its broadest reasonable interpretation, this limitation recites mathematical concepts and calculations because gradient ascent and gradient descent with loss distribution are mathematical calculations comprising application of mathematical formulas i.e. applying loss functions, gradient ascent formulae, and gradient descent formulae. See MPEP 2106.04(a)(2), subsection I.
Step 2A Prong 2: Does the claim recite additional elements that integrate the judicial exception into a practical application?
The judicial exception is not integrated into a practical application. In particular, The claim(s) recite(s) in part:
“A system, comprising: at least one data processor; and at least one memory storing instructions which, when executed by the at least one data processor, result in operations comprising”. As drafted and under its broadest reasonable interpretation, this limitation recites additional elements which amount to generic computer components recited at a high level of generality, with merely the words “apply it” or an equivalent with the judicial exception, merely including instructions to implement an abstract idea on the additional elements, or merely using the additional elements as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f).
“retrieving a set of authentic electronic medical records from a database”. As drafted and under its broadest reasonable interpretation, this limitation recites retrieving data, which is mere data gathering and output recited at a high level of generality, and thus are insignificant extra-solution activity. See MPEP 2106.05(g) (“whether the limitation is significant”). In addition, all uses of the recited judicial exceptions require such data gathering and output, and, as such, these limitations do not impose any meaningful limits on the claim. These limitations amount to necessary data gathering and outputting. See MPEP 2106.05.
“training a first neural network using the first training set,” and “training a second neural network using the second training set”. As drafted and under its broadest reasonable interpretation, these limitations recite additional elements which amount to generic computer components recited at a high level of generality, with merely the words “apply it” or an equivalent with the judicial exception, merely including instructions to implement an abstract idea on the additional elements, or merely using the additional elements as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f).
“…the first neural network outputting synthetic electronic medical records” and “…the second neural network outputting a loss distribution, the loss distribution indicating whether the output synthetic electronic medical records are classified as authentic or synthetic”. As drafted and under its broadest reasonable interpretation, these limitations recite outputting data, which is mere data gathering and output recited at a high level of generality, and thus are insignificant extra-solution activity. See MPEP 2106.05(g) (“whether the limitation is significant”). In addition, all uses of the recited judicial exceptions require such data gathering and output, and, as such, these limitations do not impose any meaningful limits on the claim. These limitations amount to necessary data gathering and outputting. See MPEP 2106.05.
The additional elements have been considered both individually and as an ordered combination in order to determine whether they integrates the exception into a practical application. Therefore, no meaningful claim limits are imposed practicing the abstract idea. Accordingly, at Step 2A, prong two, the additional elements do not integrate the judicial exception into a practical application.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception?
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more that the judicial exception. As discussed, the claim limitations reciting generic computer elements amounts to no more than mere instructions to apply the exception using a generic computer. The claim reciting the additional elements of retrieving data amount to receiving/transmitting information.
“Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network); but see DDR Holdings, LLC v. Hotels.com, L.P., 773 F.3d 1245, 1258, 113 USPQ2d 1097, 1106 (Fed. Cir. 2014) ("Unlike the claims in Ultramercial, the claims at issue here specify how interactions with the Internet are manipulated to yield a desired result‐‐a result that overrides the routine and conventional sequence of events ordinarily triggered by the click of a hyperlink." (emphasis added)) MPEP § 2106.05(d)(II)(i).
The additional elements have been considered both individually and as an ordered combination in order to determine whether they warrant significantly more consideration. Thus, the claim does not provide an inventive concept.
The claim is ineligible.
Claim 2, which depends upon claim 1, recite(s) in part:
“wherein training the first neural network further comprises receiving a conditioning modifier, the conditioning modifier altering at least one characteristic of the synthetic electronic medical records”. As drafted and under its broadest reasonable interpretation, this limitation recites receiving conditional modifiers, which is mere data gathering and output recited at a high level of generality, and thus are insignificant extra-solution activity. See MPEP 2106.05(g) (“whether the limitation is significant”). In addition, all uses of the recited judicial exceptions require such data gathering and output, and, as such, these limitations do not impose any meaningful limits on the claim. These limitations amount to necessary data gathering and outputting. See MPEP 2106.05.
The claim reciting the additional elements of mere data gathering do not integrate the judicial exception into practical application. The additional elements have been considered both individually and as an ordered combination in order to determine whether they integrates the exception into a practical application.
The claim reciting the additional elements of “receiving” and/or “transmitting” amount to receiving/transmitting information. The additional elements have been considered both individually and as an ordered combination in order to determine whether they warrant significantly more consideration. The claim limitations do not recite additional elements that are sufficient to amount to significantly more than the judicial exception.
The claim is ineligible.
Claim 3, which depends upon claim 2, recite(s) in part:
“wherein receiving the conditioning modifier comprises receiving the conditioning modifier via a user interface”. As drafted and under its broadest reasonable interpretation, this limitation recites receiving conditional modifiers via a user interface, which is mere data gathering and output recited at a high level of generality, and thus are insignificant extra-solution activity. See MPEP 2106.05(g) (“whether the limitation is significant”). In addition, all uses of the recited judicial exceptions require such data gathering and output, and, as such, these limitations do not impose any meaningful limits on the claim. These limitations amount to necessary data gathering and outputting. See MPEP 2106.05.
The claim reciting the additional elements of mere data gathering do not integrate the judicial exception into practical application. The additional elements have been considered both individually and as an ordered combination in order to determine whether they integrates the exception into a practical application.
The claim reciting the additional elements of “receiving” and/or “transmitting” amount to receiving/transmitting information. The additional elements have been considered both individually and as an ordered combination in order to determine whether they warrant significantly more consideration. The claim limitations do not recite additional elements that are sufficient to amount to significantly more than the judicial exception.
The claim is ineligible.
Claim 4, which depends upon claim 1, recite(s) in part:
“wherein training the first neural network is in response to receiving a request for synthetic electronic health records from a front end system”. As drafted and under its broadest reasonable interpretation, this limitation further clarifies the mental evaluation of training a neural network, recited at a high level of generality, based on random noise samples to output synthetic medical records recited in claim 1, by introducing doing so in response to receiving a request, and thus falls under the same analysis.
“receiving a request for synthetic electronic health records from a front end system”. As drafted and under its broadest reasonable interpretation, this limitation recites receiving a request, which is mere data gathering and output recited at a high level of generality, and thus are insignificant extra-solution activity. See MPEP 2106.05(g) (“whether the limitation is significant”). In addition, all uses of the recited judicial exceptions require such data gathering and output, and, as such, these limitations do not impose any meaningful limits on the claim. These limitations amount to necessary data gathering and outputting. See MPEP 2106.05.
The claim reciting the additional elements of mere data gathering do not integrate the judicial exception into practical application. The additional elements have been considered both individually and as an ordered combination in order to determine whether they integrates the exception into a practical application.
The claim reciting the additional elements of “receiving” and/or “transmitting” amount to receiving/transmitting information. The additional elements have been considered both individually and as an ordered combination in order to determine whether they warrant significantly more consideration. The claim limitations do not recite additional elements that are sufficient to amount to significantly more than the judicial exception.
The claim is ineligible.
Claim 6, which depends upon claim 4, recite(s) in part:
“wherein the first gradient comprises
PNG
media_image1.png
54
298
media_image1.png
Greyscale
” . As drafted and under its broadest reasonable interpretation, this limitation further clarifies the mental evaluation of updating gradients based on loss distribution recited in claim 1, by introducing
PNG
media_image1.png
54
298
media_image1.png
Greyscale
, and thus falls under the same analysis. The broadest reasonable interpretation also encompasses an abstract idea of mathematical concepts and calculations because using a gradient function is a mathematical calculation comprising application of mathematical formulas i.e. applying gradient functions. See MPEP 2106.04(a)(2), subsection I.
The claim does not integrate the judicial exception into practical application.
The claim limitations do not recite additional elements that are sufficient to amount to significantly more than the judicial exception.
The claim is ineligible.
Claim 8, which depends upon claim 6, recite(s) in part:
“wherein the second gradient comprises
PNG
media_image2.png
31
491
media_image2.png
Greyscale
” . As drafted and under its broadest reasonable interpretation, this limitation further clarifies the mental evaluation of updating gradients based on loss distribution recited in claim 1, by introducing
PNG
media_image2.png
31
491
media_image2.png
Greyscale
, and thus falls under the same analysis. The broadest reasonable interpretation also encompasses an abstract idea of mathematical concepts and calculations because using a gradient function is a mathematical calculation comprising application of mathematical formulas i.e. applying gradient functions. See MPEP 2106.04(a)(2), subsection I.
The claim does not integrate the judicial exception into practical application.
The claim limitations do not recite additional elements that are sufficient to amount to significantly more than the judicial exception.
The claim is ineligible.
Claim 9, which depends upon claim 1, recite(s) in part:
“wherein the first neural network comprises a recurrent neural network”. As drafted and under its broadest reasonable interpretation, this limitation further clarifies the mental evaluation of training a neural network, recited at a high level of generality, based on random noise samples to output synthetic medical records as recited in claim 1, by introducing a recurrent neural network, and thus falls under the same analysis.
The claim does not integrate the judicial exception into practical application.
The claim limitations do not recite additional elements that are sufficient to amount to significantly more than the judicial exception.
The claim is ineligible.
Claim 10, which depends upon claim 9, recite(s) in part:
“wherein the recurrent neural network utilizes a time aware long short term memory”. As drafted and under its broadest reasonable interpretation, this limitation further clarifies the mental evaluation of training a neural network, recited at a high level of generality, based on random noise samples to output synthetic medical records as recited in claims 1 and 9, by introducing a time aware long short term memory, and thus falls under the same analysis.
The claim does not integrate the judicial exception into practical application.
The claim limitations do not recite additional elements that are sufficient to amount to significantly more than the judicial exception.
The claim is ineligible.
Claim 11, which depends upon claim 9, recite(s) in part:
“wherein the recurrent neural network utilizes a gated recurrent unit”. As drafted and under its broadest reasonable interpretation, this limitation further clarifies the mental evaluation of training a neural network, recited at a high level of generality, based on random noise samples to output synthetic medical records as recited in claims 1 and 9, by introducing a gated recurrent unit, and thus falls under the same analysis.
The claim does not integrate the judicial exception into practical application.
The claim limitations do not recite additional elements that are sufficient to amount to significantly more than the judicial exception.
The claim is ineligible.
Claim 12, which depends upon claim 1, recite(s) in part:
“wherein the operations further comprise: validating the synthetic medical records, wherein the validating comprises comparing a statistical distribution of the synthetic medical records to a statistical distribution of the authentic medical records”. As drafted and under its broadest reasonable interpretation, this limitation recites an abstract idea of a mental process because validating synthetic medical records by comparing statistical distributions encompasses mental evaluations that are practically performed in the human mind, but for the recitation of generic computer components. Even if most humans would use a physical aid, like a pen and paper or a calculator, to make such evaluations, the use of a physical aid would not negate the mental nature of this limitation. See MPEP 2106.04(a)(2), subsection III.B.
The claim does not integrate the judicial exception into practical application.
The claim limitations do not recite additional elements that are sufficient to amount to significantly more than the judicial exception.
The claim is ineligible.
Claim 13, which depends upon claim 12, recite(s) in part:
“wherein the operations further comprise: validating the synthetic medical records, wherein the validating comprises comparing a statistical distribution of the synthetic medical records to a statistical distribution of the authentic medical records”. As drafted and under its broadest reasonable interpretation, this limitation recites an abstract idea of a mental process because validating synthetic medical records by comparing statistical distributions encompasses mental evaluations that are practically performed in the human mind, but for the recitation of generic computer components. Even if most humans would use a physical aid, like a pen and paper or a calculator, to make such evaluations, the use of a physical aid would not negate the mental nature of this limitation. See MPEP 2106.04(a)(2), subsection III.B.
The claim does not integrate the judicial exception into practical application.
The claim limitations do not recite additional elements that are sufficient to amount to significantly more than the judicial exception.
The claim is ineligible.
Claim 14, which depends upon claim 1, recite(s) in part:
“wherein the second neural network is distributed across multiple devices in separate locations in a federated learning structure”. As drafted and under its broadest reasonable interpretation, this limitation further clarifies the mental evaluation of training a neural network recited at a high level of generality, to output a loss distribution recited in claim 1, by introducing the second neural network is distributed across multiple devices in separate locations in a federated learning structure, and thus falls under the same analysis.
The claim does not integrate the judicial exception into practical application.
The claim limitations do not recite additional elements that are sufficient to amount to significantly more than the judicial exception.
The claim is ineligible.
Claims 15 and 19 are substantially similar to claim 1, and thus are rejected on the same basis as claim 1.
Claims 16-18 are substantially similar to claims 2-4, and thus are rejected on the same basis as claims 2-4.
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 1-19 are rejected under 35 U.S.C. 103 as being unpatentable over Truong et al. (Pub. No.: US 10460235 B1) as cited in the IDS dated 06/03/2022, hereafter Truong, in view of Guan et al. ("Generation of Synthetic Electronic Medical Record Text "), hereafter Guan, in further view of Liu et al. (Pub.. No.: 2018/0247201 A1), hereafter Liu.
Regarding claim 1, Truong discloses:
A system, comprising: at least one data processor; and at least one memory storing instructions which, when executed by the at least one data processor, result in operations comprising (Truong, Fig. 1 and column 3, lines 31-39),
retrieving a set of authentic electronic medical records from a database (Truong, Fig. 8, column 3, lines 49-50 “In some aspects, the operations can further include retrieving a reference dataset from a database” and column 25, lines 49-51 “an exemplary dataset can be an access log for patient medical records that tracks the job title of the employee accessing a patient medical record” teaches retrieving patient medical records as a set of authentic electronic medical records from a database),
converting the authentic set of electronic medical records to a set of numerical vectors (Truong, Fig. 8, elements 801-805 teach converting the authentic dataset to a set of numerical values within a predetermined range as numerical vectors),
creating a first training set based on a random noise generator sample (Truong, column 13, line 55-58 “training parameters from model optimizer 107 … input noise dimension” and column 21, lines 21-25 “… a mapping from a sample space (e.g., a random number or vector) to a data space”, column 9, lines 61-67 and column 10, lines 1-15 teaches a first training set based on random noise generator sample in the sample space),
training a first neural network using the first training set, the first neural network outputting synthetic electronic medical records (Truong, column 21, lines 21-25 “The generative adversarial network can include a generator network and a discriminator network. The generator network can be configured to learn a mapping from a sample space (e.g., a random number or vector) to a data space”, column 13, lines 1-5 “dataset generator 103 can be configured to generate synthetic data from sample values. For example, dataset generator 103 can be configured to use the generative network of a generative adversarial network to generate data samples from random-valued vectors”, column 13, line 55-58 “training parameters from model optimizer 107 … input noise dimension”, and column 25, lines 48-54 “an exemplary dataset can be an access log for patient medical records that tracks the job title of the employee accessing a patient medical record. The job title "Administrator" may be a rare value of job title and appear in 3% of the log entries. System 100 can be configured to generate synthetic log data based on the actual dataset,” teaches training a generator network in a generative adversarial network as a first neural network to output synthetic medical records based on using the noisy random data an input training data),
creating a second training set based on the synthetic electronic medical records and the set of numerical vectors (Truong, Fig. 8, column 22, lines 19-22 “ Process 800 can then proceed to step 807. In step 807, 20 system 100 (e.g., dataset generator 103) can train the generative network using the normalized dataset, consistent with disclosed embodiments.” and column 21, lines 21-30 “The discriminator can be configured to determine, when presented with either an actual data sample or a sample of synthetic data generated by the generator network” creating a second training set based on a normalized dataset comprising the synthetic electronic medical records and the set of numerical vectors in Fig. 8),
training a second neural network using the second training set, the second neural network outputting a loss distribution, the loss distribution indicating whether the output synthetic electronic medical records are classified as authentic or synthetic (Truong, Fig. 8, column 21, lines 21-30 “The generative adversarial network can include a generator network and a discriminator network…The discriminator can be configured to determine, when presented with either an actual data sample or a sample of synthetic data generated by the generator network, whether the sample was generated by the generator network or was a sample of actual data.”, column 2, lines 33-42 “The generative network can be trained to generate the output data with an output data schema matching a schema of the reference dataset. The method can further comprise training the generative adversarial network using a loss function that penalizes generation of data differing from the reference dataset by less than the predetermined amount.” Teaches training a discriminator in a generative adversarial network as a second neural network that outputs a loss distribution indicating whether the output synthetic electronic medical records are classified as authentic or synthetic using the second training set disclosed in Fig. 8),
wherein training the first neural network further comprises updating … first neural network … based on the loss distribution (Truong, column 2, lines 57-62 “A loss function can be updated that penalizes generation of data differing from the reference dataset by less than the predetermined amount using the similarity metric value. The generative adversarial network can be trained using the normalized reference dataset, the synthetic training dataset, and the updated loss function.” Teaches training the generative network to comprise updating the network based on the loss distribution),
wherein training the second neural network further comprises updating … the second neural network … based on the loss distribution (Truong, column 2, lines 57-62 “A loss function can be updated that penalizes generation of data differing from the reference dataset by less than the predetermined amount using the similarity metric value. The generative adversarial network can be trained using the normalized reference dataset, the synthetic training dataset, and the updated loss function.” And column 21, lines 30-33 “As training progresses, the generator can improve at generating the synthetic data and the discriminator can improve at determining whether a sample is actual or synthetic data.” teaches updating a discriminator based on the loss distribution).
Truong teaches training the first neural network further comprises updating … first neural network … based on the loss distribution, and training the second neural network further comprises updating … the second neural network …based on the loss distribution, but does not explicitly teach updating a gradient in order to do so.
Guan discloses:
updating a first gradient of the first neural network based on … loss (Guan, Fig. 1, Algorithm 1, lines 10 and 12 teaches updating the parameter gradients of the generator neural network using loss based on reward),
updating a second gradient of the second neural network based on … loss (Guan, Fig. 1, Algorithm 1, lines 10 and 16 teaches updating the parameter gradients of the discriminator neural network using loss based on reward),
Truong and Guan are analogous art because they are from the same field of endeavor, generative adversarial networks and medical data.
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Truong to include updating a first gradient of the first neural network based on … loss and updating a second gradient of the second neural network based on … loss, based on the teachings of Guan. One of ordinary skill in the art would have been motivated to make this modification in order to generate more realistic medical data, as suggested by Guan (Guan, page 380, left column, last paragraph, lines 5-6 “generate more realistic EMR texts”).
Trung, in view of Guan teaches training the first neural network further comprises updating a first gradient of the first neural network … based on the loss distribution, and wherein training the second neural network further comprises updating a second gradient of the second neural network … based on the loss distribution, but does not disclose descending the first gradient and ascending the second gradient in order to do so.
Liu discloses:
updating a first gradient of the first neural network by descending the first gradient based on the loss distribution (Liu, Fig. 2C and ¶[0055] teaches descending a first gradient based on loss distribution to update a gradient of generator neural networks),
wherein training the second neural network further comprises updating a second gradient of the second neural network by ascending the second gradient based on the loss distribution (Liu, Fig. 2C and ¶[0055] teaches ascending a second gradient based on loss distribution to update a gradient of discriminator neural networks).
Truong, Guan, Liu are analogous art because they are from the same field of endeavor, generative adversarial networks.
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Truong, in view of Guan, to include descending the first gradient and ascending the second gradient, based on the teachings of Liu. One of ordinary skill in the art would have been motivated to make this modification in order to improve accuracy during training, as suggested by Liu ([0031]).
Regarding claim 2, Truong, in view of Guan, in further view of Liu, discloses the system of claim 1. Truong further discloses:
wherein training the first neural network further comprises receiving a conditioning modifier, the conditioning modifier altering at least one characteristic of the synthetic electronic medical records (Truong, column 12, lines 39-49 “In some embodiments, system 100 can be configured to provide instructions for improving the quality of the synthetic data model. If a user requires synthetic data reflecting less correlation or similarity with the original data, the use can change the models' parameters to make them perform worse (e.g., by decreasing number of layers in GAN models,” teaches receiving conditional modifiers from users to alter at least one characteristic of the synthetic electronic medical records).
Regarding claim 3, Truong, in view of Guan, in further view of Liu, discloses the system of claim 2. Truong further discloses:
wherein receiving the conditioning modifier comprises receiving the conditioning modifier via a user interface (Truong, Fig. 1, element 113, and column 12, lines 39-49 teaches receiving the modifier from a user via an interface).
Regarding claim 4, Truong, in view of Guan, in further view of Liu, discloses the system of claim 1. Truong further discloses:
wherein training the first neural network is in response to receiving a request for synthetic electronic health records from a front end system (Truong, column 10, lines 23-27 “Model optimizer 107 can be configured to generate model based on instructions received from a user or another system. These instructions can be received through interface 113.").
Regarding claim 5, Truong, in view of Guan, in further view of Liu, discloses the system of claim 1. Liu further discloses:
wherein updating the first gradient comprises descending the first gradient until the loss distribution satisfies a first threshold for the first neural network (Liu, equations 1-7, ¶[0052] and ¶[0055] teaches descending the first gradient until loss distribution satisfies the variational upper bound of an objective function as a threshold for generator neural networks),
updating the second gradient comprises ascending the second gradient until the loss distribution satisfies a second threshold for the second neural network (Liu, equations 1-7, ¶[0052] and ¶[0055] teaches ascending the second gradient until loss distribution satisfies the variational upper bound of an objective function as a threshold for discriminator neural networks),
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Truong, in view of Guan, to include wherein updating the first gradient comprises descending the first gradient until the loss distribution satisfies a first threshold for the first neural network, and updating the second gradient comprises ascending the second gradient until the loss distribution satisfies a second threshold for the second neural network, based on the teachings of Liu. One of ordinary skill in the art would have been motivated to make this modification in order to improve accuracy during training, as suggested by Liu ([0031]).
Regarding claim 6, Truong, in view of Guan, in further view of Liu, discloses the system of claim 4. Guan further discloses:
wherein the first gradient comprises
PNG
media_image3.png
33
178
media_image3.png
Greyscale
(Guan, Equations 4, 5, 6, 7, 8, and 9 teaches the embodiments of the first gradient, more specifically present in equations 4 and 6).
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Truong to include the first gradient comprises
PNG
media_image3.png
33
178
media_image3.png
Greyscale
, based on the teachings of Guan. One of ordinary skill in the art would have been motivated to make this modification in order to generate more realistic medical data, as suggested by Guan (Guan, page 380, left column, last paragraph, lines 5-6 “generate more realistic EMR texts”).
Regarding claim 7, Truong, in view of Guan, in further view of Liu, discloses the system of claim 1. Liu further discloses:
wherein the second neural network generates feedback comprising gradient updates for updating the first gradient and the second gradient (Liu, ¶[0055] discloses adversarial discriminator neural networks that generates feedback comprising gradient updates for updating the first gradient and the second gradient).
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Truong, in view of Guan to include wherein the second neural network generates feedback comprising gradient updates for updating the first gradient and the second gradient based on the teachings of Liu. One of ordinary skill in the art would have been motivated to make this modification in order to improve accuracy during training, as suggested by Liu ([0031]).
Regarding claim 8, Truong, in view of Guan, in further view of Liu, discloses the system of claim 6. Guan further discloses:
wherein the second gradient comprises
PNG
media_image4.png
16
311
media_image4.png
Greyscale
(Guan, Equations 4, 5, 6, 7, 8, and 9 teaches the embodiments of the second gradient, more specifically present in equations 5 and 6).
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Truong to include - wherein the second gradient comprises
PNG
media_image4.png
16
311
media_image4.png
Greyscale
, based on the teachings of Guan. One of ordinary skill in the art would have been motivated to make this modification in order to generate more realistic medical data, as suggested by Guan (Guan, page 380, left column, last paragraph, lines 5-6 “generate more realistic EMR texts”).
Regarding claim 9, Truong, in view of Guan, in further view of Liu, discloses the system of claim 1. Truong further discloses:
wherein the first neural network comprises a recurrent neural network (Truong, column 4, lines 44-47 “Generating the synthetic dataset using the synthetic dataset model and the training dataset can include identifying a sensitive portion of the training dataset using a recurrent neural network.” Teaches the first neural network to comprise a recurrent neural network).
Regarding claim 10, Truong, in view of Guan, in further view of Liu, discloses the system of claim 8. Truong further discloses:
wherein the recurrent neural network utilizes a time aware long short term memory (Truong, column 11, lines 53-55 “In some embodiments, a recurrent neural network can include long short term memory modules (LSTM units)”).
Regarding claim 11, Truong, in view of Guan, in further view of Liu, discloses the system of claim 8. Truong further discloses:
wherein the recurrent neural network utilizes a gated recurrent unit (Truong, column 11, lines 53-55 “In some embodiments, a recurrent neural network can include long short term memory modules (LSTM units)”, and column 15, lines 30-31 “selection can be based on model performance feedback received from development environment” teaches an LSTM recurrent neural network with gated recurrent units that perform selection via feedback).
Regarding claim 12, Truong, in view of Guan, in further view of Liu, discloses the system of claim 1. Truong further discloses:
wherein the operations further comprise: validating the synthetic medical records, wherein the validating comprises comparing a statistical distribution of the synthetic medical records to a statistical distribution of the authentic medical records (Truong, Fig. 14, element 1407, and column 31, lines 35-41 “the performance criteria can include a similarity metric (e.g., a statistical correlation score, data similarity score, or data quality score, as described herein). For example, model optimizer 1303 can be configured to compare the covariances or univariate distributions of a synthetic dataset generated by the new synthetic data model and a reference data stream dataset.” teaches validating the synthetic data by comparing a statistical distribution of the synthetic medical records to a statistical distribution of the authentic medical records).
Regarding claim 13, Truong, in view of Guan, in further view of Liu, discloses the system of claim 11. Truong further discloses:
wherein the validating further comprises comparing a predictive model performance of the synthetic medical records to a predictive model performance of the authentic medical records (Truong, column 2, lines 45-51 “ Training the data model using the synthetic dataset can include determining that the synthetic dataset satisfies a criterion concerning the at least one of the statistical correlation score between the synthetic dataset and the reference dataset, the data similarity score between the synthetic dataset and the reference dataset, and the data quality score for the synthetic dataset.” Teaches comparing a predictive model performance of the synthetic medical records to a predictive model performance of the authentic medical records for validation).
Regarding claim 14, Truong, in view of Guan, in further view of Liu, discloses the system of claim 1. Truong further discloses:
wherein the second neural network is distributed across multiple devices in separate locations in a federated learning structure (Truong, Fig 3, Fig. 4, and column 14, lines 28-33 “system 100 can be implemented in a stable and scalable fashion using a distributed computing environment, such as a public cloud-computing environment, a private cloud computing environment, a hybrid cloud computing environment, a computing cluster or grid, or the like.” teaches the second neural network to be distributed across multiple devices in separate locations in a federated learning structure).
Claims 15 and 19 are substantially similar to claim 1, and thus are rejected on the same basis as claim 1.
Claims 16-18 are substantially similar to claims 2-4, and thus are rejected on the same basis as claims 2-4.
Response to Arguments
Applicant's arguments filed 04/09/2026 have been fully considered with regards to the 35 U.S.C. 101 rejection, and are not found persuasive.
The applicant asserts on pages 7-9 of the remarks that the amended independent claims improve the technical field of machine learning in electronic health record systems and ”improving the generated synthetic patient health data such that machine learning models can be trained using records that are ‘purely synthetic and may be mathematically shown to not be re-identifiable’”. The Examiner respectfully disagrees, as this improvement of data is not shown in the claims. If the specification sets forth an improvement in technology, the claim must be evaluated to ensure that the claim itself reflects the disclosed improvement. That is, the claim must include the components or steps of the invention that provide the improvement described in the specification. Furthermore, the claims do not recite an improvement to a model such that the abstract idea is integrated into a practical application. As drafted and under its broadest reasonable interpretation, the amended implemented claims recite mathematical concepts and calculations because gradient ascent and gradient descent with loss distribution are mathematical calculations comprising application of mathematical formulas i.e. applying loss functions, gradient ascent formulae, and gradient descent formulae. See MPEP 2106.04(a)(2), subsection I.
Applicant's arguments filed 04/09/2026 have been fully considered with regards to the 35 U.S.C. 102/103 rejection.
Applicant’s arguments with respect to claim(s) 1-19 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.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to HUMAIRA ZAHIN MAUNI whose telephone number is (703)756-5654. The examiner can normally be reached Monday - Friday, 9 am - 5 pm (ET).
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, MATT ELL can be reached at (571) 270-3264. 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.
/H.Z.M./Examiner, Art Unit 2141
/MATTHEW ELL/Supervisory Patent Examiner, Art Unit 2141
Prosecution Insights