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 12/31/2025. Claims 1, 9, 17, and 26 have been amended. Claims 1, 3, 5, 6, 8, 9, 11, 13, 14, 16, 17, 19, 21-26, 29, and 30 remain pending and have been examined.
Response to Arguments
A. Applicant’s arguments with respect to the claim for benefit of prior-filed Application No. 63/247,510 has been fully considered.
Applicant’s claim of priority is accepted on the basis that the claims as currently amended no longer positively recite a step of determining whether or not to recommend the patient for palliative care based upon the output of the execution of the patient analysis model for the patient. While claims 1, 9, and 17 recite “provide the output of the execution of the patient analysis model to one or more health care providers to determine whether or not to recommend the patient for palliative care based upon the output of the execution of the patient analysis model for the patient,” the portion reciting “to determine whether or not to recommend the patient for palliative care based upon the output of the execution of the patient analysis model for the patient” only constitutes non-functional descriptive material in the form of an intended reason to provide the output of the patient analysis model, and is not required to be performed by the claimed system or devices.
B. Applicant's arguments with respect to the rejection under 35 USC 112(a) have been fully considered.
The rejection is withdrawn on the basis that the claims as currently amended no longer positively recite a step of determining whether or not to recommend the patient for palliative care based upon the output of the execution of the patient analysis model for the patient. While claims 1, 9, and 17 recite “provide the output of the execution of the patient analysis model to one or more health care providers to determine whether or not to recommend the patient for palliative care based upon the output of the execution of the patient analysis model for the patient,” the portion reciting “to determine whether or not to recommend the patient for palliative care based upon the output of the execution of the patient analysis model for the patient” only constitutes non-functional descriptive material in the form of an intended reason to provide the output of the patient analysis model, and is not required to be performed.
C. Applicant's arguments with respect to the rejection of the claims under 35 USC 101 have been fully considered but they are not persuasive.
Applicant argues starting on page 12 of the response that the claims are not directed to a method of organizing human activity, and “instead to a specific, computer-implemented technological process that continuously trains, evaluates, updates, and retrains various machine-learning models using analysis of patient attributes and outcomes in a closed feedback loop,” and that the newly added amendments “constitute a specific implementation of machine learning with a feedback-based adjustment mechanism, not a method of organizing human activity.” Examiner respectfully disagrees.
Applicant asserts that “[m]ore specifically, at least the amendment of claims 1, 9, and 17 describe updating the trained machine-learning model with one or more adjustments to weight values such that the one or more adjustments to modify subsequent prediction of patient outcomes.” However, Examiner notes that claims 1, 9, and 17 do not explicitly recite the above subject matter. Claims 1, 9, and 17 recite “updating the patient analysis model based upon reinforcement learning to improve subsequent predictions” which, while it may be construed as encompassing some form of weight value adjustment, does not provide any manner of adjusting weight values beyond generic reinforcement learning.
With respect to Applicant’s assertion regarding Example 39, the use of a deep-learning model is not construed as falling within the scope of the abstract idea, and is an additional element addressed under Step 2A Prong 2 and Step 2B. Example 39 does not support an assertion that the use and training of a machine learning model within a claim means that the claim cannot therefore be directed to an abstract idea. A broad recitation of retraining a model using a general category of algorithm such as reinforcement learning is not analogous to the examples in MPEP 2106.04(a)(1), or analogous to the claim in Example 39 which was directed specifically to a particular manner of training a machine learning model.
Examiner further respectfully disagrees with the assertion that the claims recite specific improvements in machine learning. As addressed below, Applicant’s assertion that the use of multiple models results in an improvement in machine learning based on improvements in the ability to handle “scale issues with machine learning” is not supported by either the claims or the contents of the disclosure. Merely using machine learning to analyze data only amounts to an instruction to implement the data analysis using computing elements as tools, and the high-level recitation of multiple models does not support an assertion that the machine learning model itself is improved.
Likewise, Applicant’s disclosure does not support the assertion that “[t]he ML models are separated to improve the speed and reduce the computing resources required to execute.” No description is provided of such improvements or how the recited combination of models would yield those improvements in speed and computing resources. Examiner notes that Applicant does not cite or reference any portions of the specification or drawings which would support the above assertions.
Applicant further argues starting on page 15 that the claims do not recite any enumerated category of abstract idea. Examiner respectfully disagrees.
As addressed in Step 2A Prong 1 below, the claims recite limitations falling within the scope of a method of organizing human activity including modeling a variety of medical data such as diagnoses, medications, lab results, and demographic information, using the results to analyze medical and claim data associated with a patient to determine risk and mortality, and the results to a healthcare provider. A person or clinician could perform these tasks as part of evaluating a patient’s information and conferring with a healthcare provider. Applicant does not provide any arguments specifically addressing why the limitations listed under Step 2A Prong 1 do not fall within the scope of a certain method of organizing human activity.
Examiner additionally notes that the claims recite elements which would fall within the scope of a mental process, including those reciting, on the second day of hospital admission, determining a patient risk score and predict a mortality outcome for the patient based on a plurality of patient-specific claim information, a plurality of patient-specific health records, and a plurality of patient-specific health records for the patient for at least the last 24 hours. Applicant’s remarks regarding the use of computer components to store the patient analysis model and individual component models would only amount to instructions to implement using machines as tools, and the receipt of the data analyzed would amount to insignificant extra-solution activity.
With respect to Applicant’s argument that the claims “are not drawn to, and do not supersede, any "basic tool of scientific and technological work," whether a claim recites an abstract idea under Step 2A Prong 1 is determined based on whether the claim recites limitations falling within the scope of an abstract idea, and such limitations are not required to be directed to a basic tool of scientific and technological work.
Applicant further argues starting on page 18 that the claims recite additional elements which integrate the exception into a practical application, and that the claims recite “a specific improvement over the prior art of medical analysis technology.” Examiner respectfully disagrees. Examiner initially disagrees with Applicant’s characterization of “medical analysis” as a technological field in the context of Step 2A Prong 2. A bare assertion that one of ordinary skill in the art would recognize medical analysis as a technological field is not sufficient to support such a characterization. Applicant asserts that “the system described herein improves the computer processing of such analysis by intelligently reducing the amount of computer resources required for processing and displaying.” However, Applicant does not provide further support for this assertion, and Examiner is unable to find such support within Applicant’s disclosure. Examiner also notes that the asserted improvement lacks nexus with the claim limitations.
Applicant further argues that the claims “recite specific improvements to machine learning,” asserting that “the combination of multiple models are used to handle the scale issues with machine learning” and that “by training four separate machine learning models and then combining those models, the present claims describe using deep learning to handle the potentially large volumes of data.” Examiner respectfully disagrees. Examiner initially notes that the only recitation of machine learning in claim 1 is the characterization of the patient analysis model as a “deep-learning model,” while claims 9 and 17 do not recite any of the models being machine learning models. While claims 1, 9, and 17 recite the models being trained using information, a broad recitation of training a model does not require the use of computers to perform machine learning, for example in the case of linear regression or similar models. The use of multiple models, or the combination of models recited at a high level of generality, does not support a conclusion that the claims recite specific improvements to machine learning. While claim 29 recites the models as including LSTM and neural network models, these model types are only recited at a high level of generality. Likewise, the use of a binary cross-entropy loss function and sigmoid function as recited in claim 30 are each generically applicable and common elements of different machine learning functions and are only recited at a high level of generality as used within the claimed model. Furthermore, Applicant’s assertion about the combination of models being used to “handle the scale issues with machine learning” lacks support in the disclosure and nexus with the claims.
Applicant further asserts that claims 1, 9, and 17 “provide specific information about how the model is trained” and “provides a very specific machine learning model and a specific manner of training that machine learning model, which is significantly more than just applying an abstract idea.” Examiner respectfully disagrees that these elements integrate the abstract idea into a practical application. While the claims recite the models as trained using information such as diagnosis codes, procedure codes, medication codes, lab records, vital signs records, and demographics, these limitations only describe types of information used in the model and fall within the scope of the abstract idea. The events being weighted differently likewise falls within the scope of the abstract idea. Merely reciting the use of particular data to train models, without more, is not sufficient to conclude that the claims integrate the abstract idea into a practical application.
Applicant’s assertion on page 21 that depending claim 23 “provides a specific method of training the model for improved efficiency” and “provides a more efficient and accurate manner for training a model to predict mortality outcomes” is likewise unpersuasive and lacks nexus with the claims. No description is provided in the disclosure of the training data influencing the computing resources required to train the model or improving the accuracy. Examiner also notes that selecting training data falls within the scope of the abstract idea, and simply not selecting undesired training data, on its own, is not sufficient to integrate the abstract idea into a practical application or amount to significantly more.
Applicant further states on page 21 that “the claims recite a number of additional elements, which are used in performing the steps recited in the independent claims, and these recitations extend well beyond the scope of any alleged abstract idea.” However, the limitations subsequently listed by Applicant fall within the scope of the abstract idea with the exception of the electronic healthcare records, which only constitutes an instruction to implement the recited health records using computing elements.
Applicant further argues on page 23 that the claims recite a specific improvement over the prior art in the field of decision-making analysis technology, referencing paragraphs 3-8 of the specification. However, these cited portions only provide a description of the general field of palliative and end-of-life care followed by a high-level description of the claim elements. A broad assertion that a claimed invention is relevant to a field or use is not sufficient to establish that the claims recite a specific improvement in a technology or technological field. Examiner also notes that any improvements must stem from additional elements within the claims rather than solely the abstract idea.
Applicant lastly argues starting on page 22 that the claims are similar to those in Diamond v. Diehr, stating that the present claims “provide an improvement to the technological field of medical analysis by dynamically updating predicted mortality probabilities based on newly added features of patients.” Examiner respectfully disagrees. The Court in Diehr concluded that the claims at issue amounted to significantly more than the recited mathematical formula based on the use of that formula to control the recited rubber press, i.e. a particular machine, and consequently providing an improvement in rubber manufacturing technology. However, Applicant’s asserted improvement of “dynamically updating predicted mortality probabilities based on newly added features of patients” lies entirely within the scope of the abstract idea itself.
The rejection under 35 USC 101 is maintained.
D. Applicant’s arguments with respect to claims 1, 3, 5, 6, 8, 9, 11, 13, 14, 16, 17, 19, 21-26, 29, and 30 under 35 USC 103 have been considered and are persuasive. The rejection is withdrawn based on the previously relied upon Chi reference no longer constituting valid prior art in light of the Inventor Declaration under 37 CFR 1.130 filed 12/31/2025.
Claim Objections
The previous objection to claim 26 is withdrawn based on the amendment filed 12/31/2025.
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, 5, 6, 8, 9, 11, 13, 14, 16, 17, 19, 21-26, 29, and 30 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. Claims 1, 3, 5, 6, 8, 21-23, 26, 29, and 30 are drawn to a system, claims 9, 11, 13, 14, 16, and 24 are drawn to a method, and claims 17, 19, and 25 are drawn to a computer device, each of which is within the four statutory categories.
Step 2A(1)
Claim 1 recites, in part, performing the steps of:
storing a patient analysis model comprises a combination of a plurality of models, wherein a first model of the plurality of models is trained on diagnosis codes, procedure codes, and medication codes, wherein a second model of the plurality of models, is trained on lab records, wherein a third model of the plurality of models is trained on vital signs records, wherein a fourth model of the plurality of models is trained on demographic information, wherein events are weighted differently based on the amount of time since the event’s occurrence;
receiving a patent identifier associated with a patient;
retrieving a plurality of patient-specific claim information associated with the patient identifier;
retrieving a plurality of patient-specific health records associated with the patient identifier;
retrieving the plurality of patient-specific health records for the patient for at least the last 24 hours;
on the second day of hospital admission, executing the patient analysis model to determine a patient risk score and predict a mortality outcome for the patient based on the plurality of patient-specific claim information and the plurality of patient-specific health records, and the plurality of patient-specific health records for the patient for at least the last 24 hours; and
providing the output of the execution of the patient analysis model to one or more health care providers to determine whether or not to recommend the patient for palliative care based upon the output of the execution of the patient analysis model; and
presenting the recommendation about palliative care for the patient to a healthcare provider;
updating the patient analysis model to improve subsequent predictions.
These limitations amount to a form of managing personal behavior or relationships or interactions between people, and therefore fall within the scope of a method of organizing human activity. Fundamentally the process is that of modeling a variety of medical data such as diagnoses, medications, lab results, and demographic information, using the results to analyze medical and claim data associated with a patient to determine risk and mortality, and providing the results to a healthcare provider. A person or clinician could perform these tasks as part of evaluating a patient’s risk and mortality information and then conferring with a healthcare provider.
Independent claims 9 and 17 recite similar limitations and also recite an abstract idea under the same analysis.
Step 2A(2)
This judicial exception is not integrated into a practical application because the additional elements within the claims only amount to:
A. Instructions to Implement the Judicial Exception. MPEP 2106.05(f)
Claim 1 recites additional elements of a) a patient analysis computer device comprising at least one processor and at least one memory device, and recited as implementing or performing the subsequent data processing functions such as retrieving information and determining the patient risk score, b) the patient analysis model being a “deep-learning” model, and c) health records being electronic health records, and d) using reinforcement learning to update the patient analysis model.
Claims 9 and 17 recite additional elements of a) a patient analysis computer device comprising at least one processor and at least one memory device, and recited as implementing or performing the subsequent data processing functions such as retrieving information and determining the patient risk score, b) the health records being electronic health records, and c) using reinforcement learning to update the patient analysis model.
Paragraph 35 of the specification as originally filed describes a processor as including “any programable system including systems using micro-controllers; reduced instruction set circuits (RISC), application-specific integrated circuits (ASICs), logic circuits, and any other circuit or processor capable of executing the functions described herein.” Paragraph 40 describes memory as including RAM, ROM, EPROM, and other types of memory devices. The patient analysis computer device comprising at least one processor and at least one memory device is therefore given its broadest reasonable construction as encompassing generic computing elements.
Paragraphs 25-27 describe a deep-learning model being used to predict mortality, and lists several techniques including LSTMs, deep neural networks, and random forest models. The recitation of a “deep learning” model is therefore construed as encompassing generic forms of corresponding algorithms.
Paragraphs 24, 30, and 69 describe the use of patient electronic health records, but do not provide any description beyond their use as a collection of patient data. Paragraph 83 describes servers as storing EHRs for patients. The electronic health record is therefore given its broadest reasonable construction as encompassing generic computing elements.
Paragraph 100 states that “[i]n various embodiments, the implemented ML methods and algorithms are directed toward at least one of a plurality of categorizations of machine learning, such as supervised learning, unsupervised learning, and reinforcement learning.” Paragraph 103 further provides that “ML methods and algorithms directed toward reinforcement learning may receive a user-defined reward signal definition, receive a data input, utilize a decision-making model to generate a ML output based on the data input, receive a reward signal based on the reward signal definition and the ML output, and alter the decision-making model so as to receive a stronger reward signal for subsequently generated ML outputs.” The disclosure therefore discloses reinforcement learning as one possible option alongside supervised learning and unsupervised learning, and Examiner notes that the only further disclosure amounts to a general description of how reinforcement learning algorithms function. Examiner additionally notes that “reinforcement learning” is not a singular algorithm, but rather describes an entire category of distinct algorithms. No disclosure is provided of any specific reinforcement learning algorithms beyond this high-level disclosure of altering the model based on the reward signal. The use of reinforcement learning to update the patient analysis model is therefore construed as encompassing any generic reinforcement learning algorithm.
The above elements therefore only amount to the use of computing devices within their ordinary capacity as tools to implement functions within the abstract idea. For example, the processor is only recited at a high level of generality as used to implement the various data receiving and processing functions, and the EHR is only recited at a high level of generality as electronic health records. Likewise, the model being a “deep learning” model is only recited at a high level of generality as a type of model used in the analysis, and the use of reinforcement learning is only recited at a high level of generality and disclosed broadly as one possible type of machine learning paradigm. These elements are therefore not sufficient to integrate the abstract idea into a practical application.
The above claims, as a whole, are therefore directed to an abstract idea.
Step 2B
The present claims do not include additional elements that are sufficient to amount to more than the abstract idea because the additional elements or combination of elements amount to no more than a recitation of:
A. Instructions to Implement the Judicial Exception. MPEP 2106.05(f)
As explained above, claims 1, 9, and 17 only recite the patient analysis computer device comprising at least one processor and at least one memory device, deep-learning model, electronic health records, and use of reinforcement learning as tools for performing the steps of the abstract idea, and mere instructions to perform the abstract idea using a computer is not sufficient to amount to significantly more than the abstract idea. MPEP 2106.05(f).
Thus, taken alone, the additional elements do not amount to significantly more than the above-identified judicial exception. Looking at the limitations as an ordered combination adds nothing that is not already present when looking at the elements taken individually.
Depending Claims
Claims 3 and 11 recite wherein the patient has been admitted to a hospital. These limitations fall within the scope of the abstract idea as set out above.
Claims 5 and 13 recite iteratively reanalyzing the patient’s conditions and predictions. These limitations fall within the scope of the abstract idea as set out above.
Claim 5 recites the additional element of the at least one processor performing the function of iteratively reanalyzing the patient’s conditions and predictions.
As cited above, paragraph 35 of the specification describes a processor as including “any programable system including systems using micro-controllers; reduced instruction set circuits (RISC), application-specific integrated circuits (ASICs), logic circuits, and any other circuit or processor capable of executing the functions described herein.” The at least one processor is therefore given its broadest reasonable construction as encompassing generic computing elements.
The use of a processor therefore only amounts to the use of a computing device within its ordinary capacity as a tool to implement functions within the abstract idea. In this case, the processor is only recited at a high level of generality as used to implement the data processing function of iteratively reanalyzing the patient’s conditions and predictions. This element is therefore not sufficient to integrate the abstract idea into a practical application or to amount to significantly more than the abstract idea.
Claims 6 and 14 recite wherein a course of action includes at least one of palliative care and hospice care. These limitations fall within the scope of the abstract idea as set out above.
Claims 8 and 16 recite submitting patient information from a current hospital stay to the patient analysis model. These limitations fall within the scope of the abstract idea as set out above.
Claim 8 recites the additional element of the at least one processor performing the function of submitting the patient information.
As cited above, paragraph 35 of the specification describes a processor as including “any programable system including systems using micro-controllers; reduced instruction set circuits (RISC), application-specific integrated circuits (ASICs), logic circuits, and any other circuit or processor capable of executing the functions described herein.” The at least one processor is therefore given its broadest reasonable construction as encompassing generic computing elements.
The use of a processor therefore only amounts to the use of a computing device within its ordinary capacity as a tool to implement functions within the abstract idea. In this case, the processor is only recited at a high level of generality as used to submit the patient information to the model. This element is therefore not sufficient to integrate the abstract idea into a practical application or to amount to significantly more than the abstract idea.
Claim 19 recites iteratively determining a course of action for the patient, wherein the course of action includes at least one of palliative care and hospice care. These limitations fall within the scope of the abstract idea as set out above.
Claim 19 recites the additional element of the at least one processor performing the function of iteratively determining the course of action for the patient.
As cited above, paragraph 35 of the specification describes a processor as including “any programable system including systems using micro-controllers; reduced instruction set circuits (RISC), application-specific integrated circuits (ASICs), logic circuits, and any other circuit or processor capable of executing the functions described herein.” The at least one processor is therefore given its broadest reasonable construction as encompassing generic computing elements.
The use of a processor therefore only amounts to the use of a computing device within its ordinary capacity as a tool to implement functions within the abstract idea. In this case, the processor is only recited at a high level of generality as used to implement the data processing function of iteratively determining the course of action for the patient. This element is therefore not sufficient to integrate the abstract idea into a practical application or to amount to significantly more than the abstract idea.
Claim 21 recites wherein the plurality of patient-specific health records include both inpatient and outpatient data. These limitations fall within the scope of the abstract idea as set out above.
Claim 21 recites the additional element of the patient-specific health records being electronic health records.
Paragraphs 24, 30, and 69 describe the use of patient electronic health records, but do not provide any description beyond their use as a collection of patient data. Paragraph 83 describes servers as storing EHRs for patients. The electronic health record is therefore given its broadest reasonable construction as encompassing generic computing elements.
The use of “electronic” health records only amounts to the use of computing elements as tools to implement functions within the abstract idea. In this case the EHR is only recited at a high level of generality as an “electronic” version of health records. This element is therefore not sufficient to integrate the abstract idea into a practical application or to amount to significantly more than the abstract idea.
Claim 22 recites receiving additional patient-specific health records for the patient, re-executing the patient analysis model to determine an updated patient risk score, and presenting updated information about the patient to a healthcare provider. These limitations fall within the scope of the abstract idea as set out above.
Claim 22 recites the additional elements of the at least one processor performing the subsequent data analysis functions and the patient-specific health records being electronic health records.
As cited above, paragraph 35 of the specification describes a processor as including “any programable system including systems using micro-controllers; reduced instruction set circuits (RISC), application-specific integrated circuits (ASICs), logic circuits, and any other circuit or processor capable of executing the functions described herein.” The at least one processor is therefore given its broadest reasonable construction as encompassing generic computing elements.
Paragraphs 24, 30, and 69 describe the use of patient electronic health records, but do not provide any description beyond their use as a collection of patient data. Paragraph 83 describes servers as storing EHRs for patients. The electronic health record is therefore given its broadest reasonable construction as encompassing generic computing elements.
The above elements therefore only amount to the use of computing elements as tools to implement functions within the abstract idea. In this case, the processor is only recited at a high level of generality as used to implement the data processing functions of receiving the additional patient-specific health records, updating the risk score, and presenting updated information, and the EHR is only recited at a high level of generality as “electronic” health records. These elements are therefore not sufficient to integrate the abstract idea into a practical application or to amount to significantly more than the abstract idea.
Claims 23, 24, and 25 recite receiving a plurality of health records for a plurality of patients, filtering the plurality of health records to exclude admissions related to psychiatry, labor/delivery, and bone marrow transplants, and training the patient analysis model using the plurality of filtered health records. These limitations fall within the scope of the abstract idea as set out above.
Claims 23 and 25 recite the additional elements of a) the at least one processor performing the subsequent data analysis functions, b) the health records being electronic health records, and c) a long short-term memory (LSTM) technique as used to train the patient analysis model.
Claim 24 recites the additional elements of a) the health records being electronic health records, and b) a long short-term memory (LSTM) technique as used to train the patient analysis model.
As cited above, paragraph 35 of the specification describes a processor as including “any programable system including systems using micro-controllers; reduced instruction set circuits (RISC), application-specific integrated circuits (ASICs), logic circuits, and any other circuit or processor capable of executing the functions described herein.” The at least one processor is therefore given its broadest reasonable construction as encompassing generic computing elements.
Paragraphs 24, 30, and 69 describe the use of patient electronic health records, but do not provide any description beyond their use as a collection of patient data. Paragraph 83 describes servers as storing EHRs for patients. The electronic health record is therefore given its broadest reasonable construction as encompassing generic computing elements.
Paragraph 27 states that the machine learning techniques “can include, but are not limited to, long short-term memory (LSTM), deep neural networks (DNN), random forest (RF), and logistic regression (LR).”
The above elements only amount to the use of computing elements as tools to implement functions within the abstract idea. In this case, the processor is only recited at a high level of generality as used to implement the data processing functions of receiving the additional health records, updating the risk score, and presenting updated information, and the EHR is only recited at a high level of generality as “electronic” health records. The LSTM is similarly only recited at a high level of generality as a technique used to train the patient analysis model, and is described in the disclosure as just one possible mechanism for training the model. These elements are therefore not sufficient to integrate the abstract idea into a practical application or to amount to significantly more than the abstract idea.
Claim 26 recites wherein the plurality of events include events with outcomes of interest including if the patient died while in the hospital, was discharged to hospice, or had a recorded date of death within 30 days of discharge. These limitations fall within the scope of the abstract idea as set out above.
Claim 29 recites the additional elements of wherein the first model, the second model, and the third model are bidirectional long short-term memory models, wherein the fourth model is a neural network.
Paragraph 45 states that “[t]he deep-learning model is composed of a combination of four models representing the above groups…[t]hree models are bidirectional long short-term memory (LSTM) models and the final is a neural network model…”.
The above elements only amount to the use of computing elements as tools to implement functions within the abstract idea. Specifically, each of the models is only recited at a high level of generality within the claim as a respective modeling technique used in the patient analysis model, and are likewise only disclosed broadly in terms of their respective categories of algorithm. These elements are therefore not sufficient to integrate the abstract idea into a practical application or to amount to significantly more than the abstract idea.
Claim 30 recites the additional elements of an output layer of the deep-learning patient analysis model using a binary cross-entropy loss function for the output layer, and a Sigmoid function used as an activation function for a hidden layer of the deep-learning patient analysis model.
Paragraph 45 provides the only disclosure of the above elements, reflecting the language of the claims in stating that “[a] binary cross-entropy loss function can be employed as the output layer and a Sigmoid function was used as the activation function for the hidden layer.” No further disclosure of these elements or how they are employed in the model is provided.
The above elements only amount to the use of computing elements as tools to implement functions within the abstract idea. Specifically, the binary cross-entropy loss function is only recited at a high level of generality within the claim the output layer, and the sigmoid function is likewise only recited at a high level of generality as used as an activation function. Examiner notes that binary cross-entropy loss functions and sigmoid functions are standard mathematical functions which can be used as part of various machine learning models. These elements are therefore not sufficient to integrate the abstract idea into a practical application or to amount to significantly more than the abstract idea.
Claims 1, 3, 5, 6, 8, 9, 11, 13, 14, 16, 17, 19, 21-26, 29, and 30 are therefore rejected under 35 U.S.C. 101 as being directed to non-statutory subject matter.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Goldberg et al (US Patent Application Publication 2018/0293353) describes a system which uses a model to determine a patient risk score based on EMR data and advises on treatments such as palliative and hospice care.
Mettert et al, Identification of Hospital Patients in Need of Palliative Care – A Predictive Score.
Fiorentino et al, The Palliative Performance Scale Predicts Mortality in Hospitalized Patients with COVID-19.
Monteiro et al, Prediction of mortality in Intensive Care Units: a multivariate feature selection.
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 WILLIAM G LULTSCHIK whose telephone number is (571)272-3780. The examiner can normally be reached 9am - 5pm.
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, Fonya Long can be reached at (571) 270-5096. 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.
/Gregory Lultschik/Examiner, Art Unit 3682