DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Oath/Declaration
The declarations under 37 CFR 1.132 filed 07/28/2026 are sufficient to overcome the rejections of claims 1-20 based upon 35 USC 101 and 35 USC 103.
Response to Arguments
Rejection Under 101:
Applicant's arguments filed 07/28/2026 have been fully considered.
Applicant argues that the claims are not directed to an abstract idea because they do not fall within the grouping of abstract ideas. The claims are directed to a method performed by a unique, nested machine learning model architecture. The specific hardware implemented computational arrangement is a nested, multi-INN configuration rather than a generalized concept. Therefore, the claim is not directed to managing personal behavior or interactions (following rules or instructions).
In response to Applicant’s argument, the claims recite an abstract idea for being directed to following rules or instructions to analyze input data and make determinations about a person’s mental disorder and providing a recommendation. This would fall under organizing human activity. While the claims recite an abstract idea, after reviewing the SMEDs the arguments were persuasive regarding the practical application of the additional elements recited in the amended claim. Therefore, the rejection is withdrawn and this argument is moot.
Applicant argues that the claims are analogous to Example 39 where the claims were found to improve the accuracy of the technological process. Similarly, the claims use proportional sampling to improve the accuracy of an AI system. The training process applies representative densities to real-world occurrence frequency that represents a concrete computational constraint used to train a classifier and distinguish uncommon and common disorders. (SMED – Irmak, para. 12-13). Proportional real-world cases prevents the model from relying on artificial, equal-distribution datasets that fail to reflect clinical reality. See spec. [0092]. Since the claims improve the function of a computer/machine learning system, it does not recite an abstract idea.
In response to Applicant’s argument, as discussed above, the claims do recite an abstract idea, unlike Example 39. However, due to the claims being integrated into a practical application, the 101 rejection is withdrawn.
Applicant argues that the claims are integrated into a practical application. See SMED – Irmak, para 12-16. The claims provide a technical solution to the technical problem of class imbalance in training. See also spec. [0092]. This specific training constraint improves the technical accuracy of the model, moving it beyond a generic “black box.”
In response to Applicant’s argument, after considering the SMEDs, the declarations were found to be persuasive. The rejection of the 101 rejection is withdrawn.
Applicant claims a computational improvement over models known in the field through a multi-label output recited in the claims. See spec. [0029]. The claimed model overcomes the inefficiencies of training many models by training a single deep learning model that can learn to predict all labels at the same time. See SMED, para. 11, 15-16. This provides efficiency improvements in training due to its architecture and makes the solution scalable. Id. It makes deployment to production less error prone compared to hundred or deployed models. Id.
In response to Applicant’s argument, after considering the SMEDs, the declarations were found to be persuasive. The rejection of the 101 rejection is withdrawn.
Applicant argues that the similar to Desjardins, the machine learning model is trained on multiple tasks (i.e., multiple mental disorders). See SMED Utku Irmak, para 11. The present model computes probabilities concurrently which results in significant efficiency improvements for real-time interference compared to running hundreds of models in parallel. See SMED Irmak, para. 5-10.
In response to Applicant’s argument, after considering the SMEDs, the declarations were found to be persuasive. The rejection of the 101 rejection is withdrawn.
Applicant argues that the claimed invention improves the neural network (NN) by addressing certain specific NN structural constraints. The two input modules are associated in different formats of data and outputs are nested together by interacting or connected to each other in a hierarchical, ensemble, or stack form. These specifics amount to something more that is required under the Alice test.
In response to Applicant’s argument, after considering the SMEDs, the declarations were found to be persuasive. The rejection of the 101 rejection is withdrawn.
Applicant argues that the claims amount to significantly more than a routine or conventional activity. The claims recite an unconventional data aggregation that incorporates collateral information from a circle of individuals. See spec. para. 26-27. The flexible data collection improves data access needed for high-dimensional multi-label risk prediction. The combination of multi-source input with parallel INN processing is not well understood or routine practice in the industry. See SMED Irmak, para. 13-16.
In response to Applicant’s argument, the data aggregation is something that the courts have held to be mere data gathering and thus well understood and routine activity. However, the SMEDs were persuasive regarding the arguments directed to the practical application and thus the 101 rejection is withdrawn. Therefore, this argument is moot.
Rejection Under 103:
Applicant's arguments filed 07/28/2026 have been fully considered.
Applicant argues that the cited portion of the prior art fails to disclose or render obvious a model with the structure recited in the claims. LaBorde for example, does not teach a respective format associated with the two input modules. Brunner does not teach data transformation based on formats of the inputs modules, let alone transforming a single piece of data to different formats associated with the two input modules.
In response to Applicant’s arguments, the arguments appear to be directed to the amendment and is therefore moot. However, see below in light of the withdrawn rejection. See “Subject Matter Free of Prior Art.”
Applicant argues that the cited portion of the prior art fails to disclose or render obvious the concurrent computation of respective probabilities for multiple mental disorders, as recited in the claims.
In response to Applicant’s arguments, the arguments appear to be directed to the amendment and is therefore moot. However, see below in light of the withdrawn rejection. See “Subject Matter Free of Prior Art.”
Claim Rejections - 35 USC § 112
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 1-20 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. The dependent claims are also rejected for inheriting the issues of the independent claims.
Regarding claim 1 (and similarly claims 8, 15), the claims recite “pre-processing the pieces of input data to obtain input for a machine learning model, wherein the pre-processing includes… inputting the pre-processed pieces of input to a machine learning model, wherein the machine learning model comprises a plurality of input modules.” The recitation of “the machine learning model” is unclear if the model refers to the same “a machine learning model,” if Applicant is attempting to refer to different models, or if applicant is attempting to recite obtaining input for a model… inputting the input into one of the model. Appropriate clarification or correction is required.
Subject Matter Free of Prior Art
Claims 1-20 are free of prior art over LaBorde (US 2023/0138557) in view of Brunner (US2017/0249434). The prior art references, or reasonable combination thereof, could not be found to disclose, or suggest all the limitations found in the independent claims. The closest prior art is LaBorde (US 2023/0138557), which teaches using trained machine learning models to predict a suicide risk of patients based upon attributes of events and making recommendations to reduce the risk of suicides. Brunner (US2017/0249434 teaches detecting early signs of disease and recovery using a personalized health monitoring solution using smart devices. The references taken solely, or in combination, fail to provide the required limitations, and modification of any complementary combination of the references of record would be impermissible hindsight and not provide any advantages over their present application. The dependent claims are also free of prior art due to their corresponding dependency of the independent claims.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, 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 nonprovisional extension fee (37 CFR 1.17(a)) 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 mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to AMANDA R COVINGTON whose telephone number is (303)297-4604. The examiner can normally be reached Monday - Friday, 10 - 5 MT.
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, Jason B. Dunham can be reached at (571) 272-8109. 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.
/AMANDA R. COVINGTON/Examiner, Art Unit 3686
/RACHELLE L REICHERT/Primary Examiner, Art Unit 3686
Prosecution Insights