DETAILED ACTION
Applicant’s response filed 11/07/2025 has been fully considered. The following rejections and/or objections are either reiterated or newly applied.
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 .
Claim Status
Claims 2, 11, 13 and 20 are cancelled by Applicant.
Claims 1, 3-10, 12 and 14-19 are currently pending and are herein under examination.
Claims 1, 3-10, 12 and 14-19 are rejected.
Claims 1, 9, 12 and 19 are objected.
Priority
The instant application claims domestic benefit to U.S. Provisional Application No. 62/071,487 (hereinafter “App. ‘487”) filed 08/28/2020. However, App. ‘487 does not provide disclosure for an “electronic controller”, which is recited in claim 18. Thus, the claim to domestic benefit for claims 1, 3-10, 12, 14-17 and 19 is acknowledged, but is not acknowledged for claim 18. As such, the effective filing date for claims 1, 3-10, 12, 14-17 and 19 is 08/28/2020. The effective filing date for claim 18 is 08/27/2021, which is the date of original disclosure for an “electronic controller”.
Withdrawn Rejections
35 USC 112(a)
The rejection of claims 12 and 14-20 under 35 U.S.C. 112(a) is withdrawn in view of Applicant replacing “controller” with “processor” in claims 12, 15 and 19.
35 USC 112(b)
The rejection of claims 1, 3-12 and 14-20 are withdrawn in view of Applicant’s claim amendments.
35 USC 103
The rejection of claims 12, 14-15 and 17 under 35 USC 103 as being unpatentable over Kowsari in view of Bastarache and AMCI is withdrawn in view of Applicant removing the limitation of the “controller”.
The rejection of claims 16 and 18 under 35 USC 103 as being unpatentable over Kowsari in view of Bastarache and AMCI and in further view of Banda is withdrawn in view of Applicant removing the limitation of the “controller”.
The rejection of claim 19 under 35 USC 103 as being unpatentable over Kowsari in view of Bastarache and AMCI and in further view of Wu is withdrawn in view of Applicant removing the limitation of the “controller”.
The rejection of claims 9 and 10 under 35 USC 103 as being unpatentable over Kowsari in view of Bastarache and Wu is withdrawn in view of Applicant’s claim amendments.
Claim Objections
Claims 1, 9, 12 and 19 are objected to because of the following informalities:
Claim 1, second to last line, recites the initialism “ICD” which should be spelled out first.
Claim 1, last 2 lines, recites the phrase “the electronic health record before a recorded date of a genetic test in the electronic health record” which should be “the stored electronic health record before a recorded date of a genetic test in the stored electronic health record”.
Claims 1, 9 and 12 should all have a phrase similar to “are kept” at the end of the last line to correct the grammar of the clause that beings with “only phecodes corresponding to ICD codes.”
Claim 9, last 2 lines, recites the phrase “the electronic health record before a recorded date of a genetic test in the electronic health record” which should be “the stored electronic health record before a recorded date of a genetic test in the stored electronic health record”.
Claim 12, second to last line, recites the initialism “ICD” which should be spelled out first.
Claim 12, last 2 lines, recites the phrase “the electronic health record before a recorded date of a genetic test in the electronic health record” which should be “the stored electronic health record before a recorded date of a genetic test in the stored electronic health record”.
Claim 19, line 3, should contain “:” after “by”.
Appropriate correction is required.
Claim Rejections - 35 USC § 112
35 USC 112(b)
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1, 3-10, 12 and 14-19 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
These rejections are either newly recited as necessitated by claim amendment or are newly recited in view of further consideration of the claims.
Claim 1, lines 12-14, recites the phrase “wherein the trained artificial intelligence model is trained to produce an output indicating whether a training data patient is a candidate for genetic testing based on a training data phenotypic representation” which renders the claim indefinite. It is unclear if this phrase requires an active step of training an already trained AI model as indicated by “is trained”, or if it is a product by process limitation that defines the training process previously performed to acquire the trained AI model. To overcome this rejection, it is suggested to clarify the interpretation of this phrase. For examination purposes, this phrase is being interpreted as a product by process limitation.
Claim 1, last 3 lines, recites the phrase “wherein the trained artificial intelligence model is trained using a training data set including a plurality of stored electronic health records, wherein for each stored electronic health record, only phecodes corresponding to ICD codes added to the electronic health record before a recorded date of a genetic test in the electronic health record” which renders the claim indefinite. It is unclear if this phrase requires an active step of training an already trained AI model as indicated by “is trained”, or if it is a product by process limitation that defines the training process previously performed to acquire the trained AI model. To overcome this rejection, it is suggested to clarify the interpretation of this phrase. For examination purposes, this phrase is being interpreted as a product by process limitation.
Furthermore, claims 3-8 are also rejected because they depend on claim 1, which is rejected, and because they do not resolve the issue of indefiniteness.
Claim 6 recites “wherein the trained artificial intelligence model is trained to produce the output … by producing a numeric …” which renders the claim indefinite. It is unclear if this phrase requires an active step of training an already trained AI model as indicated by “is trained”, or if it is a product by process limitation that defines the training process previously performed to acquire the trained AI model. To overcome this rejection, it is suggested to clarify the interpretation of this phrase. For examination purposes, this phrase is being interpreted as a product by process limitation.
Claim 6 recites the phrase “the output indicating whether the patient is the candidate for the genetic testing” which lacks antecedent basis. Claim 1, lines 12-13, recites the phrase “an output indicating whether a training data patient is a candidate for genetic testing”, but claim 1 does not recite an output indicating “the patient” being a candidate.
Claim 7 recites the phrase “wherein the trained artificial intelligence model is trained to produce the output … by producing a first … and a second …” which renders the claim indefinite. It is unclear if this phrase requires an active step of training an already trained AI model as indicated by “is trained”, or if it is a product by process limitation that defines the training process previously performed to acquire the trained AI model. To overcome this rejection, it is suggested to clarify the interpretation of this phrase. For examination purposes, this phrase is being interpreted as a product by process limitation.
Claim 7 recites the phrase “the output indicating whether the patient is the candidate for the genetic testing” which lacks antecedent basis. Claim 1, lines 12-13, recites the phrase “an output indicating whether a training data patient is a candidate for genetic testing”, but claim 1 does not recite an output indicating “the patient” being a candidate.
Claim 8 recites the phrase “the trained artificial intelligence model is trained to further produce a numeric output …” which renders the claim indefinite. It is unclear if this phrase requires an active step of training an already trained AI model as indicated by “is trained”, or if it is a product by process limitation that defines the training process previously performed to acquire the trained AI model. To overcome this rejection, it is suggested to clarify the interpretation of this phrase. For examination purposes, this phrase is being interpreted as a product by process limitation.
Claim 9, lines 5-6, recites the phrase “determining, based on the electronic health record, a patient corresponding to the electronic health record has undergone a genetic test” which renders the claim indefinite. It is unclear which electronic health record is being referenced out of “a plurality of electronic health records” as recited in line 3. It is also unclear whether how the determination being “based” on the health record is different from “corresponding” to the health record, or if there are two separate health records and the determination is based on one and corresponding to another. To overcome this rejection, it is suggested to amend the phrase to something similar to “determining, based on an electronic health record from the plurality of electronic health records, that a patient has undergone a genetic test”.
Claim 9, lines 9-10, recites the phrase “wherein the set of phecodes are obtained based on the electronic health record” which renders the claim indefinite. It is unclear which electronic health record is being referenced out of “a plurality of electronic health records” as recited in line 3.
Claim 9, line 10, recites the phrase “the phenotypic representation” which renders the claim indefinite because it is unclear which phenotypic representation is being reference out of each phenotypic representation for each electronic health record, as recited in lines 7-8.
Claim 9, line 11, recites the phrase “the plurality of ICD codes” which renders the claim indefinite because it is unclear which plurality of ICD codes is being referenced because lines 3-4 recite that each electronic health record has a plurality of ICD codes.
Claim 9, lines 13 and 15, recites the phrase “the phenotypic representation” which render the claim indefinite because it is unclear which phenotypic representation is being referenced from the plurality of phenotypic representations recited in lines 7-8.
Claim 9, line 18, recites the phrase “the candidate for genetic testing” which lacks antecedent basis. Lines 1-2 recite “identify candidates for genetic testing” and lines 16-17 recite “a candidate for the genetic test” but there is no recitation of “a candidate for genetic testing”.
Furthermore, claim 10 is also rejected because it depends on claim 9, which is rejected, and because it does not resolve the issue of indefiniteness.
Claim 10, lines 1-2, recites the phrase “wherein the indication of whether the patient has undergone the genetic test” which lacks antecedent basis in both claims 9-10.
Claim 12, lines 12-14, recites the phrase “wherein the trained artificial intelligence model is trained to produce an output indicating whether the patient is a candidate for genetic testing based on a training data phenotypic representation” which renders the claim indefinite. It is unclear if this phrase intends to require an active step of training an already trained AI model as indicated by “is trained”, or if it intends to be a product by process limitation that defines the training process previously performed to acquire the trained AI model. To overcome this rejection, it is suggested to clarify the interpretation of this phrase. For examination purposes, this phrase is being interpreted as a product by process limitation.
Claim 12, last 3 lines, recites the phrase “wherein the trained artificial intelligence model is trained using a training data set including a plurality of stored electronic health records, wherein for each stored electronic health record, only phecodes corresponding to ICD codes added to the electronic health record before a recorded date of a genetic test in the electronic health record” which renders the claim indefinite. It is unclear if this phrase intends to require an active step of training an already trained AI model as indicated by “is trained”, or if it intends to be a product by process limitation that defines the training process previously performed to acquire the trained AI model. To overcome this rejection, it is suggested to clarify the interpretation of this phrase. For examination purposes, this phrase is being interpreted as a product by process limitation.
Furthermore, claims 14-19 are also rejected because they depend on claim 12, which is rejected, and because they do not resolve the issue of indefiniteness.
Claim 14, line 4, recites the phrase “the electronic health record” which lacks antecedent basis. Claim 12, to which claim 14 depends, recites “electronic health record data” in line 1 and “electronic health record data for a patient” in lines 3-4. However, claim 12 does not recite an “electronic health record”.
Claim 16 recites the phrase “wherein the trained artificial intelligence model is trained to produce the output … by producing a numeric output …” which renders the claim indefinite. It is unclear if this phrase requires an active step of training an already trained AI model as indicated by “is trained”, or if it is a product by process limitation that defines the training process previously performed to acquire the trained AI model. To overcome this rejection, it is suggested to clarify the interpretation of this phrase. For examination purposes, this phrase is being interpreted as a product by process limitation.
Claim 17 recites the phrase “wherein the trained artificial intelligence model is trained to produce the output … by producing a first … and a second …” which renders the claim indefinite. It is unclear if this phrase requires an active step of training an already trained AI model as indicated by “is trained”, or if it is a product by process limitation that defines the training process previously performed to acquire the trained AI model. To overcome this rejection, it is suggested to clarify the interpretation of this phrase. For examination purposes, this phrase is being interpreted as a product by process limitation.
Claim 18 recites the phrase “the trained artificial intelligence model is trained to further produce a numeric output” which renders the claim indefinite. It is unclear if this phrase requires an active step of training an already trained AI model as indicated by “is trained”, or if it is a product by process limitation that defines the training process previously performed to acquire the trained AI model. To overcome this rejection, it is suggested to clarify the interpretation of this phrase. For examination purposes, this phrase is being interpreted as a product by process limitation.
Claim 18, lines 3-4, recites the phrase “the electronic controller” which lacks antecedent basis. For examination purposes, this phrase will be interpreted as “the electronic processor” because claim 12, line 1, amended the phrase “the electronic controller” to “the electronic processor”.
Claim 19, lines 8-9, recites the phrase “the set of phecodes being based at least in part on the plurality of ICD codes” which renders the claim indefinite because it is unclear which set of phecodes is being referenced out of each set of phecodes, as recited in claim 19 line 7, and which plurality of ICD codes is being referenced out of the plurality of ICD codes, as recited in claim 19 lines 5-6. To overcome this rejection it is suggested to amend the phrase to “wherein each set of phecodes is based at least in part on the corresponding plurality of ICD codes in the corresponding stored electronic health record”.
Claim 19, line 10, recites the phrase “the electronic health record” which renders the claim indefinite. It is unclear which electronic health record is being referenced out of the plurality of stored electronic health records, as recited in claim 19, line 4. To overcome this rejection, clarify which stored electronic health record is being referenced.
Claim 19, lines 12 and 15, recites the phrase “the training data set,” which renders the claim indefinite because it is unclear if the phrase refers to “a training data set” in claim 19, line 3, or if it refers to “a training data set” in claim 12, line 18. To overcome this rejection, it is suggested to clarify which training data set is being referenced.
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.
Claims 1, 3-10, 12 and 14-19 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Any newly recited portions herein are necessitated by claim amendment.
Step 2A, Prong 1:
In accordance with MPEP § 2106, claims found to recite statutory subject matter (Step 1: YES) are then analyzed to determine if the claims recite any concepts that equate to an abstract idea, law of nature or natural phenomena (Step 2A, Prong 1). In the instant application, claims 1 and 3-8 recite a method, claims 9-10 recite a method, and claims 12 and 14-19 recite a system. The instant claims recite the following limitations that equate to one or more categories of judicial exception:
Claim 1 recites “generating a phenotypic representation based on rarity of a set of phecodes across the group of patients and diversity of phecodes of the patient, wherein the set of phecodes are obtained based on the electronic health record data, and the phenotypic representation indicates prevalence of a subset of the set of phecodes; applying a trained artificial intelligence model comprising a classifier to the phenotypic representation by processing the phenotypic representation using the classifier to obtain a pattern of a group of phenotypes, wherein the trained artificial intelligence model is trained to produce an output indicating whether a training data patient is a candidate for genetic testing based on a training data phenotypic representation; wherein the trained artificial intelligence model is trained using a training data set including a plurality of stored electronic health records, wherein for each stored electronic health record, only phecodes corresponding to ICD codes added to the electronic health record before a recorded date of a genetic test in the electronic health record.”
Claim 3 recites “wherein the phenotypic representation includes at least one selected from a group consisting of: a binary matrix indicating presence or absence of each of a plurality of phecodes in the electronic health record data converted from International Classification of Diseases (ICD) codes in the electronic health record data, a matrix of phecode counts indicating a number of occurrences of each phecode in the electronic health record data converted from the ICD codes in the electronic health record data, and a phenotypic risk score.”
Claim 4 recites “further comprising automatically scheduling the patient for a genetic testing procedure in response to the transmitted output signal.”
Claim 6 recites “wherein the trained artificial intelligence model is trained to produce the output indicating whether the patient is the candidate for the genetic testing by producing a numeric output indicative of a probability that the patient has a genetic disorder.”
Claim 7 recites “wherein the trained artificial intelligence model is trained to produce the output indicating whether the patient is the candidate for the genetic testing by producing a first output indicating whether the patient is the candidate for the genetic testing and a second output identifying a specific genetic disorder.”
Claim 8 recites “wherein the trained artificial intelligence model is trained to further produce a numeric output indicative of a relative probability that the patient has a genetic disorder based on the phenotypic representation,”
Claim 9 recites “determining, based on the electronic health record, a patient corresponding to the electronic health record has undergone a genetic test; obtaining training data by generating a phenotypic representation for each electronic health record based on rarity of a set of phecodes across a group of patients and diversity of phecodes of the patient, wherein the set of phecodes are obtained based on the electronic health record, and the phenotypic representation indicates prevalence of a subset of the set of phecodes, the set of phecodes being based at least in part on the plurality of ICD codes; training the machine-learning model with the training data by processing the phenotypic representation using a classifier to obtain a pattern of a group of phenotypes, the pattern indicating a constellation of the group of phenotypes, wherein the machine-learning model is trained to receive as input the phenotypic representation and to produce as output an indication of whether the patient corresponding to the set of phecodes is a candidate for the genetic test; wherein the trained artificial intelligence model is trained using a training data set including a plurality of stored electronic health records, wherein for each stored electronic health record, only phecodes corresponding to ICD codes added to the electronic health record before a recorded date of a genetic test in the electronic health record.”
Claim 10 recites “wherein the indication of whether the patient has undergone the genetic test includes an indication of a specific genetic test of a plurality of genetic tests, and wherein the machine-learning model is trained to produce as output an identification of the specific genetic test.”
Claim 12 recites “generate an input data set based on the electronic health record data for the patient by generating a phenotypic representation based on rarity of a set of phecodes across a group of patients and diversity of phecodes of the patient, wherein the set of phecodes are obtained based on the electronic health record data, and the phenotypic representation indicates prevalence of a subset of the set of phecodes indicated by the electronic health record data; apply a trained artificial intelligence model to the input data set by processing the phenotypic representation using a classifier to obtain a pattern of a group of phenotypes, the pattern indicating a constellation of the group of phenotypes, wherein the trained artificial intelligence model is trained to produce an output indicating whether the patient is a candidate for genetic testing based on a training data phenotypic representation; wherein the trained artificial intelligence model is trained using a training data set including a plurality of stored electronic health records, wherein for each stored electronic health record, only phecodes corresponding to ICD codes added to the electronic health record before a recorded date of a genetic test in the electronic health record.”
Claim 14 recites “wherein the phenotypic representation includes at least one selected from a group consisting of: a binary matrix indicating presence or absence of each of a plurality of phecodes in the electronic health record converted from International Classification of Diseases (ICD) codes in the electronic health record data, a matrix of phecode counts indicating a number of occurrences of each phecode in the converted electronic health record converted from the ICD codes in the electronic health record data, and a phenotypic risk score.”
Claim 15 recites “automatically schedule the patient for a genetic testing procedure in response to the transmitted output signal.”
Claim 16 recites “wherein the trained artificial intelligence model is trained to produce the output indicating whether the patient is the candidate for the genetic testing by producing a numeric output indicative of a probability that the patient has a genetic disorder.”
Claim 17 recites “wherein the trained artificial intelligence model is trained to produce the output indicating whether the patient is the candidate for the genetic testing by producing a first output indicating whether the patient is the candidate for the genetic testing and a second output identifying a specific genetic disorder.”
Claim 18 recites “wherein the trained artificial intelligence model is trained to further produce a numeric output indicative of a relative probability that the patient has a genetic disorder based on the phenotypic representation,”
Claim 19 recites “generate a training data set by generating a set of phecodes for each stored electronic health record of the plurality of health records, the set of phecodes being based at least in part on the plurality of ICD codes, determining, based on the electronic health record, a patient corresponding to the electronic health record has undergone a genetic test, and including in the training data set, for each stored electronic health record, the generated set of phecodes and an indication of whether the patient has undergone the genetic test; and training an artificial intelligence model based on the training data set, wherein the artificial intelligence model is trained to receive as input a set of phecodes for a patient and to produce as output an indication of whether the patient is the candidate for the genetic test.”
Limitations reciting a mental process.
The following limitations that are recited directly above in claims 1, 3, 6-10, 12, and 16-29 equate to a mental process because they are similar to the concepts of collecting information, analyzing it, and displaying certain results of the collection and analysis in Electric Power Group, LLC, v. Alstom (830 F.3d 1350, 119 USPQ2d 1739 (Fed. Cir. 2016)), which the courts have identified as concepts that can be practically performed in the human mind. The above cited limitations are: “(i) generating a phenotypic representation, (ii) generating an input data set, (iii) applying a trained artificial intelligence model that is trained to provide an output by applying a classifier to the phenotypic representation data, (iv) wherein the trained artificial intelligence model produces a numeric output indicative of a probability, (v) wherein the trained artificial intelligence model produces a first and second output, (vi) wherein the trained artificial intelligence model produces a numeric output indicative of a relative probability, (viii) determining a patient has undergone a genetic test, (ix) training the machine-learning model with a training set to produce an output of an indication and an identification, (x) generating the training set, (xi) generate an input data set, (xii) apply a trained artificial intelligence model to the input data set to produce an output, (xiii) generate a training data set by generating a set of phecodes, (xiv) training an artificial intelligence model based on the training data asset to produce an indication, and (xv) include in the training data set only phecode corresponding to ICD codes.”
Furthermore, the above cited limitations in claims 1, 3, 66-10, 12, and 16-29 that recite a mental process are recited at such a high level of generality that a human could practically perform them with their mind or by using pen and paper. For example, a human could practically collect data to generate a phenotypic representation, an input dataset, and a training set. A human could also use pen and paper to perform the mathematical operations of an unspecified trained artificial intelligence or machine-learning model that uses a classifier. The broadest reasonable interpretation (BRI) of a trained artificial intelligence or machine-learning model that uses a classifier includes it being a logistic regression, which has had its parameters updated. A human could practically input data into the updated logistic regression and perform its calculations. A human could also use pen and paper to train the logistic regression as it merely requires performing calculations and updating the model’s parameters. Therefore, these limitations equate to reciting a mental process.
Limitations reciting a mathematical concept.
The following limitations that are recited directly above in claims 1, 3, 6-10, 12, 14 and 16-19 equate a mathematical concept because these limitations are similar to the concepts of organizing and manipulating information through mathematical correlations in Digitech Image Techs., LLC v Electronics for Imaging, Inc. (758 F.3d 1344, 111 U.S.P.Q.2d 1717 (Fed. Cir. 2014)), which the courts have identified as mathematical concepts. The limitations are: (i) applying a trained artificial intelligence model that is trained to produce an output, (ii) a binary matrix, (iii) a matrix of phecode counts, (iv) a phenotypic risk score, (v) wherein the trained artificial intelligence model produces a numeric output indicative of a probability, (vi) wherein the trained artificial intelligence model produces a first and second output, (vii) wherein the trained artificial intelligence model produces a numeric output indicative of a relative probability, (viii) training the machine-learning model with a training set to produce an indication and an identification, (ix) apply a trained artificial intelligence model to the input data set to produce an output, (x) training an artificial intelligence model based on the training data asset to produce an indication, and (xi) processing the phenotypic representation using a classifier to obtain a pattern of a group of phenotypes, the pattern indicating a constellation of the group of phenotypes.
Furthermore, the BRI of a trained artificial intelligence model and a machine-learning model that uses an unspecified classifier includes them both being a logistic regression with updated parameters. Using an updated logistic regression to produce a numerical value equates to using a mathematical function to perform calculations that result in a numerical value. Moreover, the BRI of training an artificial intelligence model and machine-learning model, which both may be a logistic regression under their BRI, includes performing gradient descent, which is a mathematical function that performs calculations. Therefore, these limitations equate to reciting a mathematical concept.
Limitations reciting organization of human activity.
The following limitations that are recited directly above in claims 4 and 15 equate to organizing human activity because they are similar to managing personal behavior or relationships or interactions between people (See MPEP 210604(a)(2).II.C). The limitations are: “schedule/scheduling the patient for a genetic testing procedure”. The BRI of this limitation includes that a healthcare provider schedules an appointment between a patient and doctor.
Limitations includes in the recited judicial exception.
Regarding the above cited limitations in claims 1, 9 and 12 of “wherein the trained artificial intelligence model is trained using a training data set including a plurality of stored electronic health records, wherein for each stored electronic health record, only phecodes corresponding to ICD codes added to the electronic health record before a recorded date of a genetic test in the electronic health record”. This limitation is included in the mathematical concept and mental process of training a ML model and applying a trained AI model.
As such, claims 1, 3-10, 12 and 14-19 recite an abstract idea (Step 2A, Prong 1: Yes).
Step 2A, Prong 2:
Claims found to recite a judicial exception under Step 2A, Prong 1 are then further analyzed to determine if the claims as a whole integrate the recited judicial exception into a practical application or not (Step 2A, Prong 2). The judicial exception is not integrated into a practical application because the claims do not recite additional elements that reflect an improvement to a computer, technology, or technical field (MPEP § 2106.04(d)(1) and 2106.5(a)), require a particular treatment or prophylaxis for a disease or medical condition (MPEP § 2106.04(d)(2)), implement the recited judicial exception with a particular machine that is integral to the claim (MPEP § 2106.05(b)), effect a transformation or reduction of a particular article to a different state or thing (MPEP § 2106.05(c)), nor provide some other meaningful limitation (MPEP § 2106.05(e)). Rather, the claims include limitations that equate to an equivalent of the words “apply it” and/or to instructions to implement an abstract idea on a computer (MPEP § 2106.05(f)), insignificant extra-solution activity (MPEP § 2106.05(g)), and field of use limitations (MPEP § 2106.05(h)). The instant claims recite the following additional elements:
Claim 1 recites “accessing, from a non-transitory computer-readable memory, the electronic health record data for a patient in a group of patients; and transmitting an output signal to notify that the patient is the candidate for generic testing in response to detecting the group of phenotypes in the electronic health record data of the patient.”
Claim 5 recites “further comprising performing a genetic testing procedure in response to the transmitted output signal.”
Claim 8 recites “and the method further comprising transmitting a second output signal to a health care provider device identifying the genetic disorder in response to determining that the relative probability indicated by the numeric output exceeds a threshold.”
Claim 9 recites “accessing a plurality of electronic health records, each electronic health record including a plurality of International Classification of Diseases (ICD) codes; and transmitting an output signal to notify that the patient is the candidate for genetic testing in response to detecting the constellation of the group of phenotypes in the electronic health record data of the patient.”
Claim 12 recites “the system comprising an electronic processor configured to: access, from a non-transitory computer-readable memory, electronic health record data for a patient; and transmit an output signal to notify that the patient is the candidate for generic testing in response to detecting the group of phenotypes in the electronic health record data of the patient.”
Claim 15 recites “wherein the electronic processor is further configured to”
Claim 18 recites “and wherein the electronic controller is further configured to transmit a second output signal to a health care provider device identifying the genetic disorder in response to determining that the relative probability indicated by the numeric output exceeds a threshold.”
Claim 19 recites “wherein the electronic processor is further configured to: accessing a plurality of stored electronic health records, each stored electronic health record including a plurality of International Classification of Diseases (ICD) codes,”
Regarding the above cited limitations in claims 1, 12, 15 and 18-29 of the non-transitory computer-readable memory, the system, and the electronic processor, there are no limitations that these components require anything other than a generic computing system. Therefore, these limitations equate to mere instructions to implement an abstract idea on a generic computer, which the courts have established does not render an abstract idea eligible in Alice Corp. 573 U.S. at 223, 110 USPQ2d at 1983.
Regarding the above cited limitations in claims 1, 8-9, 12, and 18-19 of accessing electronic health records, transmitting an output signal, and receiving input data, these limitations equate to insignificant, extra-solution activity of mere data gathering/outputting that is used to collect data for the machine learning/artificial intelligence model to then output the results.
Regarding the above cited limitation in claim 5, this limitation equates to the equivalent of the words “apply it” because it recites only the idea of an outcome but fails to recite details for how the outcome solves a problem (See MPEP 2106.05(f)(1)). Specifically, the judicial exception of indicating that a patient is a candidate for genetic testing based on one or more phenotypes, as recited in claim 1, does not lead to an improvement in the genetic testing procedure, which is recited at a high level of generality. Claim 5 also equates to generally linking the use of a judicial exception to a particular technological environment because it performs a generically recited genetic test after the judicial exception of indicating that the patient is a candidate for a genetic test. This merely limits the use of the judicial exception to a particular technological use. Generally linking the use of a judicial exception to a particular technological environment cannot integrate a judicial exception into a practical application (MPEPM 2106.05(h)).
As such, claims 1, 3-10, 12 and 14-19 are directed to an abstract idea (Step 2A, Prong 2: No).
Step 2B:
Claims found to be directed to a judicial exception are then further evaluated to determine if the claims recite an inventive concept that provides significantly more than the judicial exception itself (Step 2B). These claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception because these claims recite additional elements that equate to instructions to apply the recited exception in a generic way and/or in a generic computing environment (MPEP § 2106.05(f)) and to well-understood, routine and conventional (WURC) limitations (MPEP § 2106.05(d)). The instant claims recite the following additional elements:
Claim 1 recites “accessing, from a non-transitory computer-readable memory, the electronic health record data for a patient in a group of patients; and transmitting an output signal to notify that the patient is the candidate for generic testing in response to detecting the group of phenotypes in the electronic health record data of the patient.”
Claim 5 recites “further comprising performing a genetic testing procedure in response to the transmitted output signal.”
Claim 8 recites “and the method further comprising transmitting a second output signal to a health care provider device identifying the genetic disorder in response to determining that the relative probability indicated by the numeric output exceeds a threshold.”
Claim 9 recites “accessing a plurality of electronic health records, each electronic health record including a plurality of International Classification of Diseases (ICD) codes; and transmitting an output signal to notify that the patient is the candidate for genetic testing in response to detecting the group of phenotypes in the electronic health record data of the patient.”
Claim 12 recites “the system comprising an electronic controller configured to: access, from a non-transitory computer-readable memory, electronic health record data for a patient; and transmit an output signal to notify that the patient is the candidate for generic testing in response to detecting the group of phenotypes in the electronic health record data of the patient.”
Claim 15 recites “wherein the electronic controller is further configured to”
Claim 18 recites “and wherein the electronic processor is further configured to transmit a second output signal to a health care provider device identifying the genetic disorder in response to determining that the relative probability indicated by the numeric output exceeds a threshold.”
Claim 19 recites “wherein the electronic controller is further configured to: accessing a plurality of stored electronic health records, each stored electronic health record including a plurality of International Classification of Diseases (ICD) codes,”
Regarding the above cited limitations in claims 1, 12, 15 and 18-19 of the non-transitory computer-readable memory, the system, and the electronic processor, these limitations equate to be instructions to implement an abstract idea on a generic computing environment, which the courts have established does not provide an inventive concept in Intellectual Ventures I LLC v. Capital One Bank (USA), 792 F.3d 1363, 1367, 115 USPQ2d 1636, 1639 (Fed. Cir. 2015).
Regarding the above cited limitations in claims 1, 8-9, 12 and 18-19 of accessing, transmitting, and receiving, these limitations equate to receiving/transmitting data over a network, which the courts have established as WURC limitation of a generic computer in buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014).
Regarding the above cited limitation in claim 5, this limitation equates to be WURC because Crowley et al. (“Crowley”; Nature reviews Clinical oncology 10, no. 8 (2013): 472-484; previously cited on PTO892 mailed 07/30/2024) discloses a review on liquid biopsies for monitoring cancer-genetics in the blood (title). The review explores how tumor-associated mutations detectable in blood can be used in the clinic after diagnosis, including for the assessment of prognosis (abstract). Crowley shows in Table 1 tumor-associated genetic aberrations in circulating free DNA, wherein the column for technique discloses various genetic testing methods that were performed on various patients with different tumor types. The genetic tests include PCR, TAm-Seq Digital PCR and BEAMing (Table 1). In Table 1 of Crowley, there are referenced specific articles that uses these genetic testing techniques, wherein at least three of them discuss using computer software in combination with performing genetic testing on cancer patients, indicating that they are WURC limitations when in combination with a computer.
When these additional elements are considered individually and in combination, they do not provide an inventive concept because they equate to be WURC functions/components of a generic computing system and because they have been shown to be WURC limitations for performing genetic testing in combination with a generic computer, as shown by Morley. Therefore, these additional elements do not transform the claimed judicial exception into a patent-eligible application of the judicial exception and do not amount to significantly more than the judicial exception itself (Step 2B: No).
As such, claims 1, 3-10, 12 and 14-19 are not patent eligible.
Response to Arguments under 35 USC 101
Applicant's arguments filed 11/07/2025 have been fully considered but they are not persuasive.
Applicant argues that the independent claims do not recite a specific type of artificial intelligence (AI) or machine learning (ML) model nor do they recite a specific training algorithm. Thus, the claims at most involve math but do not recite a mathematical concept (pg. 10, para. 3 of Applicant’s remarks). Applicant’s argument is not persuasive for the following reasons:
MPEP 2106.04(a)(2).I.C recites:
“A claim that recites a mathematical calculation, when the claim is given its broadest reasonable interpretation in light of the specification, will be considered as falling within the "mathematical concepts" grouping. A mathematical calculation is a mathematical operation (such as multiplication) or an act of calculating using mathematical methods to determine a variable or number, e.g., performing an arithmetic operation such as exponentiation … a claim does not have to recite the word "calculating" in order to be considered a mathematical calculation. For example, a step of "determining" a variable or number using mathematical methods or "performing" a mathematical operation may also be considered mathematical calculations when the broadest reasonable interpretation of the claim in light of the specification encompasses a mathematical calculation.”
The BRI of claims 1 and 12 do not require an active step of training the AI model. Rather, they require applying a previously trained AI model. The BRI of “apply/applying a trained AI model comprising a classifier to the phenotypic representation by processing the phenotypic representation using the classifier to obtain a pattern of a group of phenotypes” includes inputting phenotypic representations into a logistic regression, performing the calculations of the logistic regression, and outputting a classification (i.e., probability). This interpretation is reinforced by the specification in para. [33] [54] as well as Figure 4A.
The BRI of claim 9 requires training a ML model. Although claim 9 does not name a specific ML model or disclose a specific training algorithm, the BRI of training a ML model, in light of the specification, includes training a logistic regression using gradient descent.
Applicant’s reference to the August 2024 Memo on Patent Eligibility is acknowledged. However, it is noted that the claims are more generically directed to unspecific AI and ML models whereas Subject Matter Eligibility Examples 39 and 47 are directed toward neural networks.
Applicant argues that several limitations in the claims do not recite a mental process because they are of a scale and use temporal censoring constraints that cannot be practically performed in the human mind (pg. 10, para. 4 – pg. 11, para. 5 of Applicant’s remarks). Applicant’s arguments are not persuasive for the following reasons:
In claims 1, 9 and 12, the limitations of “accessing/access” have been identified as an additional element. In claims 3 and 14, the BRI of a phenotypic representation includes a phenotypic risk score, which is a numerical value calculated using a formula. A human is capable of calculating a score given a formula on pen and paper. In claims 9 and 19, the BRI of generating phecodes from ICD codes includes a mental process of linking ICD codes to clinical phenotypes using a Phecode map. In claims 1, 9 and 12, the BRI of “generating a phenotypic representation based on a rarity … indicates prevalence of a subset of the set of phecodes” includes selecting phecodes based on rarity and diversity to generate a phenotypic representation.
As discussed in Claim Rejections 112(b), the limitation in claims 1 and 12 of “wherein the trained AI model is trained using a training data set … only phecodes corresponding to ICD codes added to the HER record before a recorded date of a genetic test in the EHR” has been identified as a product by process and does not recite an active step of training. However, the same limitation in claim 9 is not being interpreted as a product by process limitation. The same limitation in claim 9 merely further limits the type of data being used to train the ML model. As discussed in the rejection above, the BRI of training an unspecified ML model includes mathematical concepts and a mental process.
Regarding Applicant’s comments of the limitations being at such a scale and using temporal constraint such that a human is not practically capable of performing them, it is noted that the temporal constraints are part of a product by process in claims 1 and 12. In claim 9, the temporal constraints merely further limit training data, wherein the training data still recites a mathematical concept and mental process. The “scale” has not been defined by the claims such that it precludes performance by a human using pen and paper. It is also noted that a human could filter information by data (i.e., remove data before a given date).
Applicant argues a practical application of a domain-specific training and inference process implemented on a processor/memory structure (pg. 11, para. 6 – pg. 12, para. 1 of Applicants’ remarks). Applicant’s arguments are not persuasive:
It is noted that claims 1 and 12 do not require an active step of training. However, claim 9 does require an active step of training. Nonetheless, the active training step in claim 9 when viewed in combination with the other limitations in claim 9 do not integrate the abstract idea into a practical application. The limitations in claims 1, 9 and 12 of “generating a phenotypic representation”, “applying a trained AI model”, and “training the ML model” recite an abstract idea. MPEP 2106.05(a) recites “It is important to note, the judicial exception alone cannot provide the improvement”.
Even when these limitations are viewed in combination with the additional element of “transmitting an output signal to notify that the patient is the candidate for genetic testing” there still appears to be no integration into a practical application. This is because “transmitting an output signal” equates to insignificant extra-solution activity of necessary data outputting, which cannot integrate a judicial exception into a practical application (MPEP 2106.05(g)(3)).
Applicant argues that the limitation in claims 1, 9 and 12 of “wherein the trained AI model is trained using … only phecodes corresponding to ICD codes added to the EHR before a recorded date of a genetic test” is not WURC because it was found to be free from the prior art (pg. 12, para. 2 of Applicant’s remarks). Applicant’s argument is not persuasive for the following reasons:
This limitation is part of the judicial exception and thus does not recite an additional element. Furthermore, MPEP 2106.04(I) recites “The Supreme Court’s decisions make it clear that judicial exceptions need not be old or long-prevalent, and that even newly discovered or novel judicial exceptions are still exceptions.”
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claims 1, 3-5, 7, 12, 14-15 and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Kowsari et al. (“Kowsari”; WO 2020/154324 A1, pg. 1-17; published 07/30/2020; previously cited on PTO892 mailed 07/30/2024) in view of Bastarache et al. (“Bastarache”; Science 359, no. 6381 (2018): 1233-1239; previously cited on PTO892 mailed 08/12/2025).
The bold and italicized text below are the limitations of the instant claims, and the italicized text serves to map the prior art onto the instant claims.
The rejection of claims 1, 3-5 and 7 is updated from the previous office action to address the new limitations.
The rejection of claims 12, 14-15 and 17 is newly applied and is necessitated by claim amendment.
Claims 1 and 12:
A method of evaluating electronic health record data to identify genetic disorders, the method comprising:
A system for evaluating electronic health record data to identify genetic disorders, the system comprising an electronic processor configured to:
Kowsari discloses systems/methods for designing a personalized action plan for a patient using genomic and phenotype data derived from electronic health record (EHR) data, and suggesting a personalized genetic testing recommendation [130] (Figure 11).
accessing, from a non-transitory computer-readable memory, the electronic health record data for a patient in a group of patients;
access, from a non-transitory computer-readable memory, electronic health record data for a patient;
Kowsari teaches that the system includes a non-transitory computer-readable medium that implements a cloud-based genomic/phenotypic data access among a plurality of digital computers, wherein the phenotypic data may be EHR from patients [11-12].
generating a phenotypic representation based on rarity of a set of phecodes across the group of patients and diversity of phecodes of the patient, wherein the set of phecodes are obtained based on the electronic health record data, and the phenotypic representation indicates prevalence of a subset of the set of phecodes;
generate an input data set based on the electronic health record data for the patient by generating a phenotypic representation based on rarity of a set of phecodes across a group of patients and diversity of phecodes of the patient, wherein the set of phecodes are obtained based on the electronic health record data, and the phenotypic representation indicates prevalence of a subset of the set of phecodes indicated by the electronic health record data;
Kowsari shows in Figure 10 genomic/phenotypic data (phenotypic representation) being inputted to and labeled by machine learning and artificial intelligence models, which are then used to improve personalized information delivered to a patient to make better decisions about their health [41].
Although Kowsari teaches generating a phenotypic representation (i.e., phenotype data), Kowsari does not teach that the phenotypic data is based on a rarity of a set of phecodes across a group of patients and diversity of phecodes of the patient, wherein the set of phecodes are obtained from the EHR, or the phenotypic data indicates a prevalence of a subset of the set of phecodes.
Bastarache mapped clinical features of 1204 Mendelian disease into phenotypes acquired from EHR and summarized the phenotypes into phenotype risk scores (PheRSs) (abstract). Bastarache discloses “We created a map from HPO terms to consolidated billing codes from the EHR called phecodes. Phecodes enable high-throughput ascertainment of EHR phenotypes and have been widely used to replicate known genetic associations and discover new ones (21–23). By mapping HPO terms to phecodes, we can express “phenotype syndromes” patterned after Mendelian diseases in OMIM in terms of clinical phenotypes that can be rapidly derived from the HER” (the set of phecodes are obtained from EHR) (pg. 1, col. 3, para. 2).
Bastarache also teaches “The PheRS for a given Mendelian disease is defined as the sum of clinical features observed in a given subject weighted by the log inverse prevalence of the feature” (based on rarity of a set of phecodes across the group of patients and diversity of phecodes of the patient … indicates prevalence of a subset of the set of phecodes) (pg. 1, col. 3, para. 2).
applying a trained artificial intelligence model comprising a classifier to the phenotypic representation by processing the phenotypic representation using the classifier to obtain a pattern of a group of phenotypes,
apply a trained artificial intelligence model to the input data set by processing the phenotypic representation using a classifier to obtain a pattern of a group of phenotypes,
Kowsari shows in Figure 10 using labeled datasets to train machine learning and artificial intelligence models that personalize information a patient receives based on their phenotype data [40]. Figure 10 shows that the machine learning and artificial intelligence models contain a classification and clustering engine (classifier) [134]. The models are trained on patient phenotype data and are updated and applied to additional data over time (applying a trained artificial intelligence model) [40] [116] [121] [127] [137]. The models identify and group data (e.g., dynamic data which is obtained from EHR) [112-113]. Figure 7 shows that the grouped data may be clustered using a health data graph (pattern of a group of phenotypes) [115-117].
wherein the trained artificial intelligence model is trained to produce an output indicating whether a training data patient is a candidate for genetic testing based on a training data phenotypic representation;
wherein the trained artificial intelligence model is trained to produce an output indicating whether the patient is a candidate for genetic testing based on a training data phenotypic representation;
Kowsari teaches that the trained models become better over time based on receiving more data [116] [137]. The data acquired from EHR records is inputted into an artificial intelligence and machine learning algorithm to match individuals to personalized genetic testing recommendations [128-130].
transmitting an output signal to notify that the patient is the candidate for genetic testing in response to detecting the group of phenotypes in the electronic health record data of the patient,
transmit an output signal to notify that the patient is the candidate for genetic testing in response to detecting the group of phenotypes in the electronic health record data of the patient,
Kowsari discloses a patient portal 102 that digitally displays information to the user [96]. Figure 11 shows a personalized action plan tailored to each individual based on a user’s health data graph [41]. One of the outputs may be a personalized genetic testing recommendation [130].
wherein the trained artificial intelligence model is trained using a training data set including a plurality of stored electronic health records, wherein for each stored electronic health record, only phecodes corresponding to ICD codes added to the electronic health record before a recorded date of a genetic test in the electronic health record.
As discussed in Claim Rejections 112(b), the broadest reasonable interpretation of this limitation is a product-by-process. MPEP 2113.I recites “The patentability of a product does not depend on its method of production. If the product in the product-by-process claim is the same as or obvious from a product of the prior art, the claim is unpatentable even though the prior product was made by a different process." In the instant case, the product is “the trained artificial intelligence model” and the training process previously performed to derive the product is “using a training data set including a plurality of stored …” Thus, any trained AI model capable of “using a classifier to obtain a pattern of a group of phecodes”, even if trained using a difference process than the process recited in claims 1 and 12, will read on the trained AI model of claims 1 and 12.
As described above, Kowsari discloses a trained classifier capable of detecting patterns of groups of phenotypes (Figure 7 and 10) [115-117] [121] [134].
Claims 3 and 14:
Kowsari discloses using phenotype data derived from EHR but does not disclose that the phenotype data includes a phenotypic risk score. Bastarache discloses a phenotype risk score called PheRS derived from phenotypes captured from EHR (abstract).
Claims 7 and 17:
Kowsari shows in Figure 10 shows a classifier being trained on patient genomic/phenotypic data. The data acquired from EHR records is inputted into an artificial intelligence and machine learning algorithm to match individuals to personalized genetic testing recommendations [128-130]. However, Kowsari does not disclose that the model identifies a specific genetic disorder. Bastarache shows in Figures 1 and 2 that the PheRSs were capable of identifying specific disease from phecodes such as cystic fibrosis and Marfan syndrome.
Prima facie case for obviousness for claims 1, 3, 7, 12, 14 and 17:
An invention would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the instant invention if some teaching or motivation in the prior art would have led that person to combine the prior art teachings to arrive at the claimed invention.
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the instant invention to have modified the phenotype data derived from EHR, as taught by Kowsari, to include a PheRS derived from phecodes taken from EHR, as taught by Bastarache. The motivation for doing so is taught by Bastarache who states “Genetic association studies often examine features independently, potentially missing subpopulations with multiple phenotypes that share a single cause”, wherein the PheRS was able to “augment rare-variant interpretation and may identify subsets of patients with distinct genetic causes for common diseases” (abstract). The motivation from Bastarache aligns with the intention of Kowsari for personalizing action plans for patients that have genetic variants detected by their method [126]. The combination would allow Kowsari to detect rare-variants and identify subsets of patients with distinct genetic causes for common disease, as taught by Bastarache.
Furthermore, one of ordinary skill in the art would have had a reasonable expectation of success for using the PheRSs of Bastarache in Kowsari because the PheRS is a phenotype score derived from EHR, wherein Kowsari already discloses using phenotype data derived from EHR [6]. There would have also been a reasonable expectation of success because Bastarache discloses that EHRs linked to genetic data helps drive genomic discovery and define clinical phenotypes associated with rare variants (pg. 1, col. 2, para. 2), and Bastarache discloses validating the efficacy of PheRS (pg. 1, col. 3, para. 3). Bastarache also states “Incorporation of richer EHR data, such as laboratory results and clinical notes (31), could increase the resolving power of PheRS. Furthermore, this method may be used with other combinations of phenotypes that do not follow established Mendelian patterns, perhaps based on undiagnosed patients with unusual presentations” (pg. 5, col. 3, para. 2).
Claims 4 and 15:
Kowsari discloses “A health provider or clinical staff orders the test for a patient based on the outcome of the test recommendation. This may also be initiated by the patient. The sample collection process can be performed at home or in a clinical setting. As a result of the testing, genotype data or biomarker data (e.g., from a genetic or health test) and a lab result for the test are obtained” [131]. Kowsari does not disclose automatically ordering a genetic test. However, this limitation was obvious over the teachings of teachings of Kowsari. MPEP 2144.04.III states that broadly providing an automation step in a claim is not sufficient to distinguish the claim over the prior art. Therefore, it would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the instant invention to have automated scheduling a genetic test for a patient because the health care provider was already manually ordering the genetic test in Kowsari.
Claim 5:
Kowsari discloses an analysis of a user’s health data [127] that includes a third step of providing a personalized genetic testing recommendation to a patient [130] and a fourth step of testing the patient based on the recommendation resulting in genotype data [131].
Claims 6, 8, 16 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Kowsari et al. (“Kowsari”; WO 2020/154324 A1, pg. 1-17; published 07/30/2020; previously cited on PTO892 mailed 07/30/2024) in view of Bastarache et al. (“Bastarache”; Science 359, no. 6381 (2018): 1233-1239; previously cited on PTO892 mailed 08/12/2025), as applied above to claims 1 and 12, and in further view of Banda et al. (“Banda”; NPJ Digit Med. 2019; 2: 23; previously cited on PTO892 mailed 07/30/2024).
The rejection of claims 6 and 8 is updated from the previous office action to address the new limitations.
The rejection of claims 16 and 18 is newly applied and is necessitated by claim amendment.
The limitations of claims 1 and 12 have been taught in the rejection above by Kowsari and Bastarache.
Claims 6 and 16:
Kowsari states that the trained models become better over time based on receiving more data with more data [137], [116]. The data acquired from EHR records is inputted into an artificial intelligence and machine learning algorithm to match individuals to personalized genetic testing recommendations [128-130]. However, Kowsari does not teach that the output contains a probability that a patient has a genetic disorder. Banda discloses a classifier to identify potential familial hypercholesterolemia (FH) using EHRs (abstract). Banda states that the classifier produces a probability for each patient as to whether they have FH (pg. 2, col. 2, para. 2; Figure 1b; pg. 3, col. 1, para. 2).
Claims 8 and 18:
Kowsari states that patient genomic data, phenotype data, interpretations, and reports can be transferred to health care providers through a custom web or mobile application [83] (Figure 1E). This may also be done through an API of a health care provider to transmit data over as secure virtual private network [93]. However, Kowsari does not teach a relative probability. Banda discloses a classifier to identify potential familial hypercholesterolemia (FH), which is a genetic condition, using electronic health records (EHR) (abstract). Banda states that the classifier produces a probability for each patient as to whether they have FH (pg. 2, col. 2, para. 2; Figure 1b; pg; 3, col. 1, para. 2). Banda states that the classification produced by the classifier, which are probabilities (pg. 2, col. 2, para. 2), correctly flagged 84% of patients at the highest probability threshold (abstract).
Prima facie case for obviousness:
An invention would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the instant invention if some teaching or motivation in the prior art would have led that person to combine the prior art teachings to arrive at the claimed invention. Kowsari discloses methods/systems for analyzing a patient’s health data to recommend a genetic test that is then performed on the patient and added to a clinical report [128-133]. Kowsari also discusses clustering and classifying patients with machine learning/artificial intelligence models using genotypic and phenotypic data [134] (Figure 10), wherein the phenotype data comes from electronic health records (EHR) [6]. Banda discloses using a machine learning classifier to classify patients that have a genetic condition called familial hypercholesterolemia using electronic health records (title and abstract), wherein the classifier improved diagnostic classification.
Therefore, one of ordinary skill in the art would have been motivated to combine the classifier of Banda that has improved diagnostic power to the method/system of Kowsari and Bastarache because it would have improved classification of phenotype data of patients, thereby recommending genetic tests that may be more relevant to a particular genetic disease. Furthermore, one of ordinary skill in the art would have had a reasonable expectation of success by combining Banda to Kowsari and Bastarache because Kowsari already uses machine learning models to classify/cluster electronic health record data.
Claim 19 is rejected under 35 U.S.C. 103 as being unpatentable over Kowsari et al. (“Kowsari”; WO 2020/154324 A1, pg. 1-17; published 07/30/2020; previously cited on PTO892 mailed 07/30/2024) in view of Bastarache et al. (“Bastarache”; Science 359, no. 6381 (2018): 1233-1239; previously cited on PTO892 mailed 08/12/2025), as applied above to claim 12, and in further view of Wu et al. (“Wu”; JMIR medical informatics 7, no. 4 (2019): e14325; previously cited on PTO892 mailed 08/12/2025).
This rejection is newly recited and is necessitated by claim amendment.
The bold and italicized text below are the limitations of the instant claims, and the italicized text serves to map the prior art onto the instant claims.
The limitations of claim 12 have been taught in the rejection above by Kowsari, Bastarache, and AMCI.
Claim 19:
wherein the electronic processor is further configured to: generate a training data set by accessing a plurality of stored electronic health records, each stored electronic health record including a plurality of International Classification of Diseases (ICD) codes,
Kowsari discloses methods/systems for facilitating genetic data exchange among different users (abstract). The system includes a non-transitory computer-readable medium that implements a cloud-based genomic/phenotypic data access among a plurality of digital computers, wherein the phenotypic data may be Electronic Health Care Records (EHR) from patients [11-12].
Kowsari and Bastarache do not teach that EHR contain ICD codes. However, this limitation is taught below by Wu.
Wu discloses accessing ICD-10 and ICD-10-CM codes and mapping them to phecodes (abstract) (pg. 3, col. 1, para. 1) (Figure 1).
generating a set of phecodes for each stored electronic health record of the plurality of health records, the set of phecodes being based at least in part on the plurality of ICD codes,
Bastarache teaches generating phecodes from Human Phenotype Ontology from EHR data (pg. 1, col. 3, para. 2).
Neither Kowsari nor Bastarache, disclose generating phecodes from ICD codes.
Wu discloses using databases from Vanderbilt that contained EHRs with associated ICD-10 and ICD-10-CM codes that were then mapped to phecodes (pg. 3, col. 1 – col. 2) (Figure 1).
determining, based on the electronic health record, a patient corresponding to the electronic health record has undergone a genetic test, and
Kowsari shows in Figure 7 that genotypic data acquired from genetic tests can be obtained through EHRs [115] and inputted into a patient’s health data graph. Kowsari also shows in Figure 10 that genetic tests are part of an individual’s health data graph that is used to train the machine learning/artificial intelligence models [121].
including in the training data set, for each stored electronic health record, the generated set of phecodes and an indication of whether the patient has undergone the genetic test; and training an artificial intelligence model based on the training data set, wherein the artificial intelligence model is trained to receive as input a set of phecodes for a patient and to produce as output an indication of whether the patient is the candidate for the genetic test.
Kowsari shows in Figure 10 the genomic/phenotypic data being inputted to and labeled by machine learning and artificial intelligence models, which are then used to improve personalized information delivered to a patient to make better decisions about their health [41]. The trained models become better over time based on receiving more data [137], [116]. Figure 10 shows that genetic tests are part of an individual’s health data graph that is used to train the machine learning/artificial intelligence model [121]. Figure 11 shows a personalized action plan tailored to each individual based on a user’s health data graph [41]. One of the outputs may be a personalized genetic testing recommendation [130].
Bastarache in combination with Kowsari teaches that phecodes are input into the machine learning classifier of Figure 10.
Prima facie case for obviousness:
An invention would have been prima facie obvious to one of ordinary skill in the art before the effective date of the instant invention if there was a finding that the prior art contained a method/system that differed from the instant invention by the substitution of some components with other components, wherein the results of the substitution would have been predictable. Thus, it would have been prima facie obvious to one of ordinary skill in the art to have substituted Human Phenotype Ontology (HPO) of Bastarache with ICD-10 and ICD-10-CM codes of Wu because HPO, ICD-10, and ICD-10-CM codes can be converted to phecodes. The result of substituting these components would have yielded predictable results because Bastarache uses phecodes to calculate PheRSs, wherein the phecodes derived from ICD-10 and ICD-10-CM in Wu could be used to calculate the PheRSs.
Response to Arguments under 35 USC 103
Applicant's arguments filed 11/07/2025 have been fully considered but they are persuasive only in part.
Applicant argues that incorporation of claims 11/20, which were found to be free from the prior art, into independent claims 1, 9 and 12 overcomes the 103 rejections (pg. 12, para. 4 – pg. 13, para. 3 of Applicant’s remarks). This argument is persuasive for claim 9, which requires an active step of training. However, claims 1 and 12 do not require an active step of training, as discussed in Claim Interpretation. Thus, the trained classifier of Kowsari still reads on the trained classifier of instant claims 1 and 12.
Conclusion
No claims are allowed.
Claims 9-10 are free from the prior art because the prior art does not fairly teach or suggest the following limitations in claim 9: “wherein the training data includes a plurality of stored electronic health records, wherein for each stored electronic health record, only phecodes corresponding to ICD codes added to the electronic health record before a recorded date of a genetic test in the electronic health record.” The closest prior art is Kowsari et al. (“Kowsari”; WO 2020/154324 A1, pg. 1-17; published 07/30/2020; previously cited on PTO892 mailed 07/30/2024). Kowsari uses phenotype data collected from electronic health records to train a model and classify phenotype data to create personalized health recommendations. However, Kowsari does not teach using phecodes nor using phecodes collected before a date of a genetic test.
Inquiries
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Noah A. Auger whose telephone number is (703)756-4518. The examiner can normally be reached M-F 7:30-4:30 EST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Karlheinz Skowronek can be reached on (571) 272-9047. 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.
/N.A.A./Examiner, Art Unit 1687
/KAITLYN L MINCHELLA/Primary Examiner, Art Unit 1685