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 .
Notice to Applicant
This communication is in response to the amendment filed 04/20/2026. Claims 1-20 are presented for examination.
Subject Matter Free of Prior Art
Claim(s) 1-20 are allowable over prior art because the prior art of record fail to expressly teach or suggest, either alone or in combination, the features found within the independent claims, in particular: “for each respective training example of the set of training examples... in response to determining that the quality level of the respective training example does not satisfy the predetermined quality threshold, generating an updated version of the training SOAP note of the respective training example using a first machine learning model, wherein a score indicative of at least one of a readability or grammar level of the updated version of the training SOAP is higher than a score indicative of at least one of a readability or grammar level of the respective training example, determining that the updated version of the training SOAP note of the respective training example corresponds to the training transcript of the respective training example by using a second machine-learning model that is different from the first machine- learning model to determine that facts in the updated version of the training SOAP note are also included in the training transcript, in response to determining that the updated version of the training SOAP note of the respective training example corresponds to the training transcript of the respective training example, replacing the training SOAP note in the respective training example with the updated version of the training SOAP note” Because the prior art does not teach or disclose the above features in the specific manner and combinations recited in independent claims 1, 8, 15, claims 1, 8, 15are hereby deemed to be allowable over prior art. Originally numbered dependent claims 2-7, 9-14, 16-20 incorporate the allowable features of originally numbered independent claims 1, 8, 15, through dependency, respectively.
However, the claims are still rejected under 101.
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 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception (i.e., a law of nature, a natural phenomenon, or an abstract idea) without significantly more. Based upon consideration of all of the relevant factors with respect to the claims as a whole, the claims are directed to non-statutory subject matter which do not include additional elements that are sufficient to amount to significantly more than the judicial exception because of the following analysis:
Claim 1 is drawn to a method which is within the four statutory categories (i.e., method). Claim 8 is drawn to a system which is within the four statutory categories (i.e., machine). Claim 15 is drawn to one or more non-transitory computer-readable media storing instructions which is within the four statutory categories (i.e., manufacture).
Independent claim 8 (which is representative of independent claims 1, 15) recites… accessing training data comprising a set of training examples, each training example of the set of training examples comprising a training transcript and a training Subjective, Objective, Assessment and Plan (SOAP) note corresponding to the training transcript; performing an evaluation process on the training data to result in evaluated training data, wherein the evaluated training data comprises evaluated training examples representing training examples in the set of training examples that have been modified by the evaluation process, and wherein performing the evaluation process on the training data comprises: for each respective training example of the set of training examples: determining that a quality level of the respective training example does not satisfy a predetermined quality threshold, in response to determining that the quality level of the respective training example does not satisfy the predetermined quality threshold, generating an updated version of the training SOAP note of the respective training example using a first…model, wherein a score indicative of at least one of a readability or grammar level of the updated version of the training SOAP is higher than a score indicative of at least one of a readability or grammar level of the respective training example, determining that the updated version of the training SOAP note of the respective training example corresponds to the training transcript of the respective training example by using a second…model that is different from the first…model to determine that facts in the updated version of the training SOAP note are also included in the training transcript, in response to determining that the updated version of the training SOAP note of the respective training example corresponds to the training transcript of the respective training example, replacing the training SOAP note in the respective training example with the updated version of the training SOAP note; and generating a fine-tuned…model using the evaluated training data by fine-tuning a machine-learning model, wherein the fine-tuned…model is configured to perform a sub-task associated with a task for generating a SOAP note.
Under its broadest reasonable interpretation, the limitations noted above, as drafted, covers certain methods of organizing human activity (i.e., managing personal behavior or relationships or interactions between people…following rules or instructions), but for the recitation of generic computer components. That is, other than reciting a “computer” to implement the method (claim 1), computing “system” (claims 8, 15), the claim encompasses rules or instructions to collect data, analyze the collected data, and output data based on the analysis. If a claim limitation, under its broadest reasonable interpretation, covers managing personal behavior or relationships or interactions between people, but for the recitation of generic computer components, then it falls within the “Certain Methods of Organizing Human Activity” grouping of abstract ideas. Accordingly, the claims recite an abstract idea.
Claim 1 recites additional elements (i.e., computer to implement the method; a first machine learning model; a second machine-learning model; a fine-tuned machine-learning model). Claim 8 recites additional elements (i.e., A system comprising: one or more processing systems; and one or more computer-readable media storing instructions; a first machine learning model; a second machine-learning model; a fine-tuned machine-learning model). Claim 15 recites additional elements (i.e., One or more non-transitory computer-readable media storing instructions; one or more processors; a system; a first machine learning model; a second machine-learning model; a fine-tuned machine-learning model). Looking to the specifications, a computer having one or more processing systems, one or more computer-readable media storing instructions is described at a high level of generality (¶ 0038; ¶ 0089; ¶ 0104; ¶ 0129; ¶ 0168-0171; ¶ 0178-0182; ¶ 0188), such that it amounts to no more than mere instructions to apply the exception using generic computer components. Also, “a first machine learning model,” “a second machine-learning model,” and “a fine-tuned machine-learning model” is described at a high level of generality (i.e., no description of the mechanism for accomplishing the result), such that using machine learning models amounts to no more than a recitation of the words "apply it" (or an equivalent), such as mere instructions to implement an abstract idea on a computer, and only generally links the use of a judicial exception to a particular technological environment or field of use (i.e., computer technology), which does not impose meaningful limits on the scope of the claim. Looking at the limitations as an ordered combination adds nothing that is not already present when looking at the elements individually. The additional elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. Accordingly, the claims are directed to an abstract idea.
Reevaluated under step 2B, the additional elements noted above do not provide “significantly more” when taken either individually or as an ordered combination. The use of a general purpose computer or computers (i.e., a computer having one or more processing systems, one or more computer-readable media storing instructions) amounts to no more than mere instructions to apply the exception using generic computer components and does not impose any meaningful limitation on the computer implementation of the abstract idea, so it does not amount to significantly more than the abstract idea. Also, “a first machine learning model,” “a second machine-learning model,” and “a fine-tuned machine-learning model” is described at a high level of generality (i.e., no description of the mechanism for accomplishing the result), such that using machine learning models amounts to no more than a recitation of the words "apply it" (or an equivalent), such as mere instructions to implement an abstract idea on a computer, and only generally links the use of a judicial exception to a particular technological environment or field of use (i.e., computer technology), which does not impose meaningful limits on the scope of the claim. Looking at the limitations as an ordered combination adds nothing that is not already present when looking at the elements individually. The combination of elements does not indicate a significant improvement to the functioning of a computer or any other technology and their collective functions merely provide a conventional computer implementation of the abstract idea. Furthermore, the additional elements or combination of elements in the claims, other than the abstract idea per se, amount to no more than a recitation of generally linking the abstract idea to a particular technological environment or field of use, as the courts have found in Parker v. Flook; similarly, the current invention merely limits the claimed calculations to the healthcare industry which does not impose meaningful limits on the scope of the claim. Therefore, there are no limitations in the claims that transform the judicial exception into a patent eligible application such that the claims amount to significantly more than the judicial exception.
Dependent claims 2-7, 9-14, 16-20 include all the limitations of the parent claims and further elaborate on the abstract idea discussed above and incorporated herein.
Claims 2-7, 9-14, 16-20 further define the analysis and organization of data for the performance of the abstract idea and do not recite any additional elements. Thus, the claims do not integrate the abstract idea into a practical application and do not provide “significantly more.”
Although the dependent claims add additional limitations, they only serve to further limit the abstract idea by reciting limitations on what the information is and how it is received and used. These information characteristics do not change the fundamental analogy to the abstract idea grouping of “Certain Methods of Organizing Human Activity,” and, when viewed individually or as a whole, they do not add anything substantial beyond the abstract idea. Furthermore, the combination of elements does not indicate a significant improvement to the functioning of a computer or any other technology. Therefore, the claims when taken as a whole are ineligible for the same reasons as the independent claims.
Response to Arguments
Applicant's arguments filed 04/20/2026 have been fully considered but they are not persuasive. Applicant’s arguments will be addressed hereinbelow in the order in which they appear in the response filed 04/20/2026.
In the remarks, Applicant argues in substance that:
Regarding the 101 rejections,
“Claims 1, 8, and 15 are directed to a specific computational architecture that solves a technological problem inherent in the fine-tuning of machine-learning models. The Specification explicitly identifies the technical problem: traditional machine-learning techniques utilized for automated SOAP note generation are "prone to errors and instability." This instability is caused by the "quality of the information being collected during the patient encounter and lack of readily available, low-cost, and high-quality training data"…Claim 1, as amended, provides a specific technical solution to this problem by implementing an automated multi-model veracity loop. Unlike a generic "apply it" instruction, Claim 1 recites a precise, recursive logical structure. For example, Claim 1 generates an updated version of a training example based on an objective "score indicative of at least one of a readability or grammar level." This ensures data enhancement is governed by measurable computational metrics rather than subjective human judgment. In another example, Claim 1 uses a "second machine-learning model that is different from the first" to "determine that facts in the updated version ... are also included in the training transcript." This creates a technological filter that verifies the semantic veracity of the first model's output against the raw source data…Claim 1, as amended, reflects the disclosed improvement of ensuring "training data consistency and quality" by reciting the specific automated mechanisms (the second-model fact filter) that reduce error- prone fine-tuning…A person of ordinary skill in the field of machine learning would readily identify that using a secondary model as a veracity checker during a fine-tuning protocol is a non-abstract improvement to the functioning of the model training environment. By confining the alleged abstract idea to this specific, unconventional verification loop, the claims clearly provide a "technologically- rooted solution to a technical problem"…The claims, as amended herein, likewise recite and reflect a system-level, technological improvement: a multi-model veracity filter that optimizes the fine-tuning protocol to achieve superior model stability…a human cannot practically extract and cross- reference objective fact lists across thousands of high-dimensional training pairs in real-time while simultaneously evaluating grammar and readability scores at the scale required for model stability”; and
“the ordered combination would still amount to "significantly more," as the Office Action itself acknowledges that the prior art fails to teach or suggest the features of the evaluation process. This corroborates that the specific recursive inter-model arrangement is not well-understood, routine, and conventional in the relevant context…an improved machine- learning training system that reduces error propagation…amounts to significantly more than the abstract idea itself.”
It is respectfully submitted that Examiner has considered Applicant’s arguments and does not find them persuasive. Examiner has attempted to address all of the arguments presented by Applicant; however, any arguments inadvertently not addressed are not persuasive for at least the following reasons:
In response to Applicant’s argument that (a) regarding the 101 rejections,
“Claims 1, 8, and 15 are directed to a specific computational architecture that solves a technological problem inherent in the fine-tuning of machine-learning models. The Specification explicitly identifies the technical problem: traditional machine-learning techniques utilized for automated SOAP note generation are "prone to errors and instability." This instability is caused by the "quality of the information being collected during the patient encounter and lack of readily available, low-cost, and high-quality training data"…Claim 1, as amended, provides a specific technical solution to this problem by implementing an automated multi-model veracity loop. Unlike a generic "apply it" instruction, Claim 1 recites a precise, recursive logical structure. For example, Claim 1 generates an updated version of a training example based on an objective "score indicative of at least one of a readability or grammar level." This ensures data enhancement is governed by measurable computational metrics rather than subjective human judgment. In another example, Claim 1 uses a "second machine-learning model that is different from the first" to "determine that facts in the updated version ... are also included in the training transcript." This creates a technological filter that verifies the semantic veracity of the first model's output against the raw source data…Claim 1, as amended, reflects the disclosed improvement of ensuring "training data consistency and quality" by reciting the specific automated mechanisms (the second-model fact filter) that reduce error- prone fine-tuning…A person of ordinary skill in the field of machine learning would readily identify that using a secondary model as a veracity checker during a fine-tuning protocol is a non-abstract improvement to the functioning of the model training environment. By confining the alleged abstract idea to this specific, unconventional verification loop, the claims clearly provide a "technologically- rooted solution to a technical problem"…The claims, as amended herein, likewise recite and reflect a system-level, technological improvement: a multi-model veracity filter that optimizes the fine-tuning protocol to achieve superior model stability…a human cannot practically extract and cross- reference objective fact lists across thousands of high-dimensional training pairs in real-time while simultaneously evaluating grammar and readability scores at the scale required for model stability”:
It is respectfully submitted that Applicant argues “Claims 1, 8, and 15 are directed to a specific computational architecture that solves a technological problem inherent in the fine-tuning of machine-learning models. The Specification explicitly identifies the technical problem: traditional machine-learning techniques utilized for automated SOAP note generation are "prone to errors and instability." This instability is caused by the "quality of the information being collected during the patient encounter and lack of readily available, low-cost, and high-quality training data."” However, “SOAP note generation” and the "quality of the information being collected during the patient encounter and lack of readily available, low-cost, and high-quality training data" addresses administrative problems, and not a technical problem to any specific devices, technology, or computers for that matter, and thus, the claims do not provide a technical solution. The claims of the present invention do not improve machine-learning models, but use machine learning based architecture to allegedly improve generating data (i.e., SOAP notes). For example, the computing system did not cause the argued problem and thus it is not a technical problem caused by the technological environment to which the claims are confined. Even if the claims provide the alleged improvements (i.e., “improving consistency and quality of the generated SOAP note”), any alleged benefits of the invention are at best, an improvement to the abstract idea of rules or instructions to collect data, analyze the collected data, and output data based on the analysis. However, an improved abstract idea is still an abstract idea and the claims do not provide a technical improvement.
Applicant argues “a specific technical solution to this problem by implementing an automated multi-model veracity loop. Unlike a generic "apply it" instruction, Claim 1 recites a precise, recursive logical structure. For example, Claim 1 generates an updated version of a training example based on an objective "score indicative of at least one of a readability or grammar level." This ensures data enhancement is governed by measurable computational metrics rather than subjective human judgment. In another example, Claim 1 uses a "second machine-learning model that is different from the first" to "determine that facts in the updated version ... are also included in the training transcript." This creates a technological filter that verifies the semantic veracity of the first model's output against the raw source data…Claim 1, as amended, reflects the disclosed improvement of ensuring "training data consistency and quality" by reciting the specific automated mechanisms (the second-model fact filter) that reduce error- prone fine-tuning…A person of ordinary skill in the field of machine learning would readily identify that using a secondary model as a veracity checker during a fine-tuning protocol is a non-abstract improvement to the functioning of the model training environment. By confining the alleged abstract idea to this specific, unconventional verification loop, the claims clearly provide a "technologically- rooted solution to a technical problem"…The claims, as amended herein, likewise recite and reflect a system-level, technological improvement: a multi-model veracity filter that optimizes the fine-tuning protocol to achieve superior model stability.” However, the claims to which Applicant seem to refer as the “multi-model veracity loop,” “precise, recursive logical structure,” “generates an updated version of a training example based on an objective "score indicative of at least one of a readability or grammar level,"” "determine that facts in the updated version ... are also included in the training transcript" are interpreted as part of the abstract idea of rules or instructions followed to collect data, analyze the collected data, and output relevant data based on the analysis accordingly, and not additional elements to be interpreted in Step 2A, Prong Two. The machine-learning models are described at a high level of generality (i.e., no description of the mechanism for accomplishing the result), such that using them amounts to no more than a recitation of the words "apply it" (or an equivalent), such as mere instructions to implement an abstract idea on a computer, and only generally links the use of a judicial exception to a particular technological environment or field of use (i.e., computer technology), which does not impose meaningful limits on the scope of the claim. Looking at the limitations as an ordered combination adds nothing that is not already present when looking at the elements individually. Furthermore, with regards to “data enhancement is governed by measurable computational metrics rather than subjective human judgment,” the courts have indicated that “Mere automation of manual processes” may not be sufficient to show an improvement in computer functionality. See: MPEP § 2106.05(a)(I). The claims of the present invention do not improve any specific devices, technology (i.e., machine learning), or computers for that matter, and thus, the claims do not provide a technical solution, but use machine learning based architecture to allegedly improve generating data (i.e., SOAP notes). For example, the computing system did not cause the argued problem and thus it is not a technical problem caused by the technological environment to which the claims are confined. Even if the claims provide the alleged improvements (i.e., “ensuring "training data consistency and quality"”), any alleged benefits of the invention are at best, an improvement to the abstract idea of rules or instructions to collect data, analyze the collected data, and output data based on the analysis. However, an improved abstract idea is still an abstract idea and the claims do not provide a technical improvement.
Applicant argues “a human cannot practically extract and cross- reference objective fact lists across thousands of high-dimensional training pairs in real-time while simultaneously evaluating grammar and readability scores at the scale required for model stability.” However, the claims of the present invention covers the sub-grouping of managing personal behavior or relationships or interactions between people in the “Certain Methods of Organizing Human Activity” grouping of abstract ideas, and not a concept performed in the human mind in the “Mental Processes” grouping, as Applicant now seems to argue. Furthermore, the courts have indicated that “Mere automation of manual processes” may not be sufficient to show an improvement in computer functionality. See: MPEP § 2106.05(a)(I).
Thus, the claims are directed to an abstract idea and the claim as a whole does not integrate the recited judicial exception into a practical application.
“the ordered combination would still amount to "significantly more," as the Office Action itself acknowledges that the prior art fails to teach or suggest the features of the evaluation process. This corroborates that the specific recursive inter-model arrangement is not well-understood, routine, and conventional in the relevant context…an improved machine- learning training system that reduces error propagation…amounts to significantly more than the abstract idea itself”:
Applicant argues “the ordered combination would still amount to "significantly more," as the Office Action itself acknowledges that the prior art fails to teach or suggest the features of the evaluation process.” However, Applicant fails to specify “the ordered combination” to which Applicant refers as amounting to “significantly more.” Regardless, looking at the limitations as an ordered combination adds nothing that is not already present when looking at the elements individually, as noted in Office Action dated 02/12/2026 and previously above.
Applicant argues “the prior art fails to teach or suggest the features of the evaluation process.” However, per MPEP § 2106.05(I): “the search for an inventive concept should not be confused with a novelty or non-obviousness determination…As made clear by the courts, the "‘novelty’ of any element or steps in a process, or even of the process itself, is of no relevance in determining whether the subject matter of a claim falls within the § 101 categories of possibly patentable subject matter…a claim for a new abstract idea is still an abstract idea. The search for a § 101 inventive concept is thus distinct from demonstrating § 102 novelty…Because [novelty and obviousness] are separate and distinct requirements from eligibility, patentability of the claimed invention under 35 U.S.C. 102 and 103 with respect to the prior art is neither required for, nor a guarantee of, patent eligibility under 35 U.S.C. 101.”
Applicant argues “the specific recursive inter-model arrangement is not well-understood, routine, and conventional in the relevant context.” However, whether the elements define only well-understood, routine, conventional activity is not a standalone test for determining eligibility, but an exemplary consideration in a non-limiting list of considerations.
Applicant argues “an improved machine- learning training system that reduces error propagation…amounts to significantly more than the abstract idea itself.” However, the claims of the present invention do not improve machine-learning models, but use machine learning based architecture to allegedly “[reduce] error propagation” by generating data (i.e., SOAP notes). For example, the computing system did not cause the argued problem and thus it is not a technical problem caused by the technological environment to which the claims are confined. The machine-learning models are described at a high level of generality (i.e., no description of the mechanism for accomplishing the result), such that using them amounts to no more than a recitation of the words "apply it" (or an equivalent), such as mere instructions to implement an abstract idea on a computer, and only generally links the use of a judicial exception to a particular technological environment or field of use (i.e., computer technology), which does not impose meaningful limits on the scope of the claim. Looking at the limitations as an ordered combination adds nothing that is not already present when looking at the elements individually. Even if the claims provide the alleged improvements (i.e., “reduces error propagation”), any alleged benefits of the invention are at best, an improvement to the abstract idea of rules or instructions to collect data, analyze the collected data, and output data based on the analysis. However, an improved abstract idea is still an abstract idea and the claims do not provide a technical improvement.
Thus, the claim as a whole does not amount to significantly more than the judicial exception.
Thus, Examiner maintains the 101 rejections of claims 1-20, which have been updated to address Applicant’s amendments and remarks and to comply with the 2019 Revised Patent Subject Matter Eligibility Guidance in the above Office Action and the 2024 Guidance Update on Patent Subject Matter Eligibility, Including on Artificial Intelligence in the above Office Action.
Conclusion
THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any extension fee pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Emily Huynh whose telephone number is (571)272-8317. The examiner can normally be reached on M-Th 8-5 PM.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Robert Morgan can be reached on (571) 272-6773.The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/EMILY HUYNH/Primary Examiner, Art Unit 3683
Prosecution Insights