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 .
Specification
The lengthy specification has not been checked to the extent necessary to determine the presence of all possible minor errors. Applicant’s cooperation is requested in correcting any errors of which applicant may become aware in the specification.
Claim Objections
Claims 1, 21, and 22 objected to because of the following informality: patient-specific metabolic models trained to predict effects of candidate treatment recommendations…and effects each previously adjusted intervention parameter on the metabolic state; should read “patient-specific metabolic models trained to predict effects of candidate treatment recommendations…and effects of each previously adjusted intervention parameter on the metabolic state;”.
Claim 2 objected to because of the following informality: wherein each candidate treatment recommendation…represents a distinct combination of intervention parameters of the population of intervention parameters should read “wherein each candidate treatment recommendation…represents a distinct combination of intervention parameters of the population of intervention parameters.” (claim is missing a period at the end of the last limitation).
Claim 4 objected to because of the following informality: wherein each of the plurality of candidate treatment recommendation should read “wherein each of the plurality of candidate treatment recommendations”.
Claim 6 objected to because of the following informality: and assigning the patient to either a first subset…or a second subset of patients insensitive to the intervention parameter based the historical changes. should read “and assigning the patient to either a first subset…or a second subset of patients insensitive to the intervention parameter based on the historical changes.”.
Claim 9 objected to because of the following informality: wherein the candidate treatment recommendation comprises instructions or adjusting a plurality of intervention parameters should read “wherein the candidate treatment recommendation comprises instructions for adjusting a plurality of intervention parameters”.
Claim 10 objected to because of the following informality: based on a comparison of the overall sensititvity of each category of patients; should read “based on a comparison of the overall sensitivity of each category of patients;”.
Claim 15 objected to because of the following informality: inputting the feature vector representation into each of the patient-specific metabolic models of the digital twin for each patient of the cohort should read “inputting the feature vector representation into each of the patient-specific metabolic models of the digital twin for each patient of the cohort.” (claim is missing a period at the end of the last limitation).
Claim 20 objected to because of the following informality: generating, for each of the one or more candidate treatment recommendation on the shortlist, should read “generating, for each of the one or more candidate treatment recommendations on the shortlist,”.
Claim Objection - Allowable Subject Matter
Claim 12 has no outstanding rejection over the cited prior art. The closest identified arts are Hendriks et al. (U.S. Patent Application Publication No. 2022/0102010), hereinafter Hendriks, and Barbiero et al. “Graph Representation Forecasting of Patient’s Medical Conditions: Toward a Digital Twin”, hereinafter Barbiero.
Hendriks discloses: (Hendriks Abstract) “Proposed are methods and systems for generating a prediction for a first organ of a subject using a Digital Twin model of a second, different organ of the subject. By obtaining an understanding of a relationship between the first and second organ, a Digital Twin prediction for the second organ is used to determine a prediction for the first organ”. This is similar to the limitation inputting the candidate treatment recommendation and the effect predicted by the primary model to each secondary model to predict effects of the candidate treatment recommendation on aspects of the metabolic state of the patient corresponding to the secondary model within claim 12, however Hendriks explicitly discloses (Hendriks [0045]) “Proposed embodiments may therefore leverage a DT model for a second organ when a DT model for a first organ is not available”, that is, rather than having a primary model and a secondary model, each modelling an aspect of a patient’s metabolic state, Hendriks only has a second organ and a first organ, with a model only for the first organ.
Barbiero discloses: (Barbiero Pg. 5) “GNNs natively allow the design of complex systems using a modular approach. First, the complexity of the human body is broken up by developing independent subsystems representing genomic alterations, biological pathways, and organ physiology. Each subsystem can be represented as a different node or a network of nodes in a GNN”. Barbiero therefore has multiple models representing aspects of a patient’s physiological systems, including metabolic systems, however Barbiero does not clearly use predicted effects of a treatment on a primary model to predict effects of the treatment on a secondary model.
Additionally, claim 18 has no outstanding rejection over the cited prior art. The closest identified prior art is Bulut et al. (U.S. Patent Application Publication No. 2021/0241116), hereinafter Bulut.
Bulut discloses: (Bulut [0045]) “Digital twin parameter calculate modules 532A and 532B may be configured to analyze various parameters and/or characteristics of digital twins-particularly from the pool of pain-free digital twins 5022-? to determine which pain estimation outputs should be weighted more or less heavily than others by weighting engine 530. For example, some of pain-free digital twins 5022-? may be deemed more accurate than others, e.g., because more accurate and/or extensive medical information is available about the underlying subjects”. This is similar to the first and third limitations recited in claim 18, however the second limitation, data sparsity or data collinearity between intervention parameters of the candidate treatment recommendation and the target improvement of the candidate treatment recommendation;, has no equivalent in Bulut.
A search for the prior arts with PE2E Search, ip.com InnovationQ+ Search, and Google Patents Search has been conducted. Besides patent databases, searching over non-patent literature databases, such as Google Scholar, has also been performed. The prior arts searched and investigated in patent and non-patent domains do not fairly teach or suggest the above limitations recited in claims 12 and 18. However, it is noted that no single limitation renders the claim as a whole patentable. The limitations indicated as allowable render the claim patentable only when considered in combination with the other claim limitations.
Therefore, claim 12, the dependents of claim 12, claims 13 and 14, and claim 18 are objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
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.
Claim 2 is 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.
Regarding claim 2,
Claim 2 recites the limitation wherein each candidate treatment recommendation of the population of candidate treatments represents a distinct combination of intervention parameters of the population of intervention parameters. It is unclear what “the population of candidate treatments” is referring to, as no “population of candidate treatments” is referred to previously within claim 2 or its parent claim, claim 1, and thus the term “the population of candidate treatments” lacks antecedent basis. For examination purposes, the limitation will be interpreted as reading “wherein each candidate treatment recommendation of a population of candidate treatments represents a distinct combination of intervention parameters of the population of intervention parameters.”.
Prior Art
The following references are used for prior art claim rejections:
Sinisi et al. “Optimal Personalised Treatment Computation through In Silico Clinical Trials on Patient Digital Twins”
Bostic et al. (U.S. Patent Application Publication No. 2021/0202103)
Shamanna et al. “Reducing HbA1c in Type 2 Diabetes Using Digital Twin Technology-Enabled Precision Nutrition: A Retrospective Analysis”
Bulut et al. (U.S. Patent Application Publication No. 2021/0241116)
Ceballos Lentini et al. (U.S. Patent Application Publication No. 2025/0166794)
Allen et al. (U.S. Patent Application Publication No. 2018/0075194)
Abu El Ata et al. (U.S. Patent Application Publication No. 2022/0076841)
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-10, 15, and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Sinisi et al. “Optimal Personalised Treatment Computation through In Silico Clinical Trials on Patient Digital Twins”, hereinafter Sinisi, in view of Bostic et al. (U.S. Patent Application Publication No. 2021/0202103), hereinafter Bostic, further in view of Shamanna et al. “Reducing HbA1c in Type 2 Diabetes Using Digital Twin Technology-Enabled Precision Nutrition: A Retrospective Analysis”, hereinafter Shamanna.
Regarding claim 1,
Sinisi teaches A method comprising:
generating a plurality of candidate treatment recommendations ((Sinisi Pg. 13) “Given a patient clinical record C, its digital twin P(C), and possible alternative doses for each treatment drug at hand (i.e., sets A1, . . . , Anu , each one including dose zero), our backtracking-based algorithm performs a search in the space of possible treatments”) for causing a target improvement [in metabolic state,] ((Sinisi Pg. 11) “We define conditions that must be always and eventually satisfied by a successful treatment by means of invariants and goals, respectively…In our case study, the properties of interest are the conditions of successful downregulation treatments (Section 3). In particular, our invariant requires that the day of the first drug administration is between day 21 and day 25 of the menstrual cycle. Moreover, the value of all the biological quantities under observation must go and stay below their thresholds from the 9-th day after the first drug administration. Our goal condition instead requires that values for those quantities stay below their thresholds for 21 consecutive days”, a goal for a successful treatment of keeping target biological properties below thresholds corresponds to a target improvement, Sinisi does not teach metabolic state)
wherein each of the candidate treatment recommendations comprises an intervention parameter and instructions for adjusting the intervention parameter to cause the target improvement; ((Sinisi Pgs. 4-5) “A key topic in precision medicine is to develop pharmacological treatments optimised for any given individual, namely personalised treatments (see, e.g., [25, 26]). Several optimisation criteria can be defined. A typical criterion is the minimisation of the overall amount of drug used”, a selected amount of drug to use corresponds to an intervention parameter with instructions for adjusting the intervention parameter)
for each of the candidate treatment recommendations, generating, from a population of patients, a cohort of patients sensitive to adjustments to the intervention parameter, […] ((Sinisi Pgs. 7-8) “VPH models like GynCycle typically take into account inter-subject variability (i.e., the physiological differences among different individuals) by including suitable parameters in their equations…Different value assignments to model parameters yield different model time evolutions and/or different reactions to drug administrations, thus defining different Virtual Phenotypes (VPs). Intuitively, each VP represents a class of indistinguishable (as long as the VPH model is concerned) patients. Computing a complete set of VPs for the model at hand is the starting point to obtain a representative population of virtual patients (hence, ideally showing all possible phenotypes)”, (Sinisi Pg. 9) “To compute a personalised treatment for a human patient, we first need to compute a digital representation for her. This will be done by using clinical data (in the form of a clinical record C, Definition 2.2) available from that patient in order to select, from our representative population of VPs, the subset of VPs that are compatible with (i.e., fit) such data”, different virtual phenotypes that have different reactions to drug administrations, particularly greater reactivity to a drug, correspond to cohorts of patients sensitive to adjustments to an intervention parameter)
for each patient of the cohort of patients, predicting effects of the candidate treatment recommendation on a [metabolic] state of the patient by inputting the candidate treatment recommendation to a digital twin of the patient, […] ((Sinisi Pg. 5) “starting from clinical data collected from a real patient, we first compute her digital twin, which defines a digital representation of that patient physiology. Then, we compute in silico, by means of intelligent search on such a digital twin, the lightest (in terms of overall amount of administered drug) treatment still effective for that patient”, performing a search of treatments to find ones that are effective corresponds to predicting effects of candidate treatments, Sinisi does not teach metabolic state of patients)
identifying, from the plurality of candidate treatment recommendations, a shortlist comprising one or more effective treatments identified based on their effects predicted by the plurality of patient-specific [metabolic models] of the digital twin; ((Sinisi Pg. 14) “Since we seek the optimal personalised treatment for the input digital twin, our algorithm does not stop at the first found successful treatment. Indeed, along the lines of [47], it keeps track of the lightest successful treatment found so far, i.e., the one envisioning the administration of the minimum overall drug amount”, Sinisi does not teach metabolic models for digital twins)
Bostic teaches the following further limitations that Sinisi does not teach:
[…] the digital twin comprising a plurality of patient-specific [metabolic] models ((Bostic [0123]) “In embodiments, the machine learning simulation uses a plurality of digital twins of the patient”, (Bostic [0288]) “For example, in some embodiments, the digital twin of the patient may be a digital representation of an entire body of the patient, of a biological system of the patient such as the cardiovascular system or the respiratory system, and/or of an organ of the patient such as a lung, a liver, or a heart of the patient”, Bostic does not teach metabolic models) trained to predict effects of candidate treatment recommendations on [metabolic] states ((Bostic [0295]) “In some embodiments, the digital twin module 1302 may be configured to simulate one or more potential future health states of the patient using one or more of the digital twins of the patient, the digital twin of the population of patients, and the one or more machine learning modules. The one or more machine learning modules may intake the digital twin of the patient and, using machine learning and/or deep learning and training related thereto, simulate a plurality of future health states of the patient. The future health states of the patient may be simulated according to variables, such as a time frame, a treatment schedule, a prescription drug schedule, a lifestyle, potential developments in one or more health issues experienced by the patients, any other suitable variable for use in simulation, and/or a combination thereof”, Bostic does not teach metabolic states) based on a training dataset of previously adjusted intervention parameters and effects each previously adjusted intervention parameter on the [metabolic] state; ((Bostic [0085]) “In embodiments, the machine learned model corresponds to the medication. In embodiments, the machine learned model is trained on a plurality of training data samples that respectively correspond to a plurality of previous patients that were prescribed the medication, wherein each training data sample includes respective patient attributes of the respective previous patient and an outcome related to the medication for the previous patient”, patient outcomes when prescribed medication corresponds to effects of a previously adjusted intervention parameter, Bostic does not teach metabolic states)
and displaying the shortlist of candidate treatment recommendations to a patient or medical provider via an application on a computing device. ((Bostic [0310]) “The platform 100 may present the one or more personalized treatment plans to the user of the platform 100, thereby allowing the user, such as a healthcare professional, to enact the personalized treatment plan to provide personalized healthcare to the patient”)
At the time of filing, one of ordinary skill in the art would have motivation to combine Sinisi and Bostic by taking the method for generating potential treatments with intervention parameters for causing a target improvement, and testing the potential treatments on digital twins of a cohort of patients to identify effective treatments, taught by Sinisi, and adding a plurality of patient-specific models that predict treatment effects based on training data, and displaying successful treatments via an application, taught by Bostic, as using a plurality of models allows for a more complex and realistic simulation of human physiological systems, training the models with historical data allows the models to learn from and identify patterns in the data, both of which increase the accuracy of the models, and displaying the treatments using an application allows for effective and convenient communication of the relevant results. Such a combination would be obvious.
Shamanna teaches the following further limitations that neither Sinisi, nor Bostic teach:
[…] the sensitivity of a patient representing a likelihood that adjustments to the intervention parameter will affect the metabolic state of the patient; ((Shamanna Pg. 4) “Machine learning algorithms analyzed the macronutrients, micronutrients, and biota nutrients from the database to determine the drivers of glucose response to specific foods for each participant. Factors found to be associated with glycemic response were analyzed for each participant. Participants were then provided with a set of specific food recommendations each day with the aim to avoid glucose spikes”, glycemic response and glucose spikes correspond to a metabolic state)
At the time of filing, one of ordinary skill in the art would have motivation to combine Sinisi, Bostic, and Shamanna by taking the method for generating potential treatments with intervention parameters for causing a target improvement testing the potential treatments on digital twins, comprising a plurality of models, of a cohort of patients to identify effective treatments, and displaying successful treatments via an application, taught by Sinisi and Bostic, and directing the method towards simulating a metabolic state, with sensitivity of each patient being the likelihood that changes to an intervention parameter change each patients’ metabolic state, taught by Shamanna, as Shamanna teaches: (Shammana Pg. 3) “Despite the importance of good glycemic control and the association of postprandial glycemic response (PPGR) with diabetes complications, predicting the impact of specific foods on PPGR has been challenging due to the high variability in different people’s response to the same food”, and thus applying the method of personalized treatment recommendation to the area results in substantial improvement to treatment. Such a combination would be obvious.
Regarding claim 2,
Sinisi, Bostic, and Shamanna jointly teach The method of claim 1,
Shamanna further teaches:
wherein each candidate treatment recommendation of the population of candidate treatments represents a distinct combination of intervention parameters of the population of intervention parameters ((Shamanna Pg. 4) “Machine learning algorithms analyzed the macronutrients, micronutrients, and biota nutrients from the database to determine the drivers of glucose response to specific foods for each participant. Factors found to be associated with glycemic response were analyzed for each participant. Participants were then provided with a set of specific food recommendations each day with the aim to avoid glucose spikes”, specific food recommendations based on macronutrients, micronutrients, and biota nutrients correspond to candidate treatment recommendations based on intervention parameters)
At the time of filing, one of ordinary skill in the art would have motivation to combine the method jointly taught by Sinisi, Bostic, and Shamanna for the parent claim of claim 2, claim 1. No new embodiments are introduced, so the reason to combine is the same as for the parent claim.
Regarding claim 3,
Sinisi, Bostic, and Shamanna jointly teach The method of claim 1, further comprising:
Sinisi further teaches:
responsive to receiving a domain for the plurality of candidate treatment recommendations, identifying patient data recorded for the population of patients within the domain, ((Sinisi Pg. 15) “In particular, such our overall dataset consists of 21 patient clinical records for a total of around 800 measurements. Each clinical record comprises measurements for the hormones that play the most important role during the human menstrual cycle: LH, FSH, E2, and P4”)
wherein the domain represents types of patient data to be measured for evaluating each candidate treatment; ((Sinisi Pgs. 11-12) “The treatment might have different duration in order to address different patient reactions. In particular, a downregulation treatment is considered successful (effective) for a patient in case the blood concentrations of a given set of hormones and other physiological quantities go below certain thresholds within 9 days from the first drug administration, and stay always below such thresholds for the following 21 days”)
and generating the plurality of candidate treatment recommendations based on patient data identified within the domain ((Sinisi Pg. 11) “When a UNDET model trajectory satisfies the goal conditions, the monitor output turns to SUCCESS, informing the caller that the input u(t) defines a successful treatment (i.e., an effective treatment which always satisfies invariants)…In our case study, the properties of interest are the conditions of successful downregulation treatments (Section 3). In particular, our invariant requires that the day of the first drug administration is between day 21 and day 25 of the menstrual cycle. Moreover, the value of all the biological quantities under observation must go and stay below their thresholds from the 9-th day after the first drug administration. Our goal condition instead requires that values for those quantities stay below their thresholds for 21 consecutive days”)
At the time of filing, one of ordinary skill in the art would have motivation to combine the method jointly taught by Sinisi, Bostic, and Shamanna for the parent claim of claim 3, claim 1. No new embodiments are introduced, so the reason to combine is the same as for the parent claim.
Regarding claim 4,
Sinisi, Bostic, and Shamanna jointly teach The method of claim 1, further comprising:
Shamanna further teaches:
defining a population of intervention parameters, wherein each intervention parameter of the population represents a feature of the candidate treatment to be input to the patient-specific metabolic model; ((Shamanna Pg. 4) “Machine learning algorithms analyzed the macronutrients, micronutrients, and biota nutrients from the database to determine the drivers of glucose response to specific foods for each participant”, macronutrients, micronutrients, and biota nutrients are intervention parameters that are also features of treatments)
and generating the plurality of candidate treatment recommendations based on the population of intervention parameters, wherein each of the plurality of candidate treatment recommendation represents a distinct combination of intervention parameters of the population of intervention parameters. ((Shamanna Pg. 4) “Machine learning algorithms analyzed the macronutrients, micronutrients, and biota nutrients from the database to determine the drivers of glucose response to specific foods for each participant. Factors found to be associated with glycemic response were analyzed for each participant. Participants were then provided with a set of specific food recommendations each day with the aim to avoid glucose spikes”, specific food recommendations based on macronutrients, micronutrients, and biota nutrients correspond to candidate treatment recommendations based on intervention parameters)
At the time of filing, one of ordinary skill in the art would have motivation to combine the method jointly taught by Sinisi, Bostic, and Shamanna for the parent claim of claim 4, claim 1. No new embodiments are introduced, so the reason to combine is the same as for the parent claim.
Regarding claim 5,
Sinisi, Bostic, and Shamanna jointly teach The method of claim 1, wherein generating the cohort of patients comprises:
Bostic further teaches:
accessing patient data for the population of patients, ((Bostic [0451]) “receiving, at a healthcare data system computing device including one or more processors, health information from one or more healthcare communication sources, wherein the health information includes one or more of genetic, environmental, and lifestyle data related to an individual patient and one or more of genetic, environmental, and lifestyle data related to a population of patients;”) the patient data comprising labels describing the sensitivity of each patient of the population of patients to the intervention parameter; ((Bostic [0308]) “The platform 100 may calculate a desirability score of the patient, the population of patients, and/or one or more patients included in the population of patients, the desirability score being based at least partially on how closely the genetic, environmental, health state, and/or lifestyle properties of the patient, the population of patients, and/or one or more patients included in the population of patients fit criteria of suitable patients for the healthcare research program”, a desirability score based on health state or lifestyle properties for one or more patients in the population corresponds to a sensitivity of patients in the population to the intervention parameter)
and generating the cohort of patients based on patients sensitive to the intervention parameter in the candidate treatment recommendation based on the accessed patient data ((Bostic [0460]) “determining, at the healthcare data system computing device, whether one or more of said patient, said population of patients, and a subset of said population of patients are desired by said healthcare researcher based on the comparison of the desirability data to the health information”, a subset of a population of patients that have desirable health information data for research corresponds to a cohort of patients sensitive to an intervention parameter)
At the time of filing, one of ordinary skill in the art would have motivation to combine the method jointly taught by Sinisi, Bostic, and Shamanna for the parent claim of claim 5, claim 1. No new embodiments are introduced, so the reason to combine is the same as for the parent claim.
Regarding claim 6,
Sinisi, Bostic, and Shamanna jointly teach The method of claim 5, wherein assigning the label describing the sensitivity of a patient to the intervention parameter to the patient comprises:
Sinisi further teaches:
determining historical changes in a [metabolic] state of the patient caused by previous adjustments to the intervention parameter; ((Sinisi Pg. 7) “The model defines the time evolution of overall 33 biological quantities (mostly blood concentration of hormones) and the Pharmacokinetics/Pharmacodynamics (PKPD) of several pharmaceutical drugs used in assisted reproduction (in particular, we will focus on GnRH analogues like Triptorelin) by means of highly non-linear differential equations”, (Sinisi Pg. 8) “VPH models like GynCycle typically take into account inter-subject variability (i.e., the physiological differences among different individuals) by including suitable parameters in their equations…Different value assignments to model parameters yield different model time evolutions and/or different reactions to drug administrations, thus defining different Virtual Phenotypes (VPs). Intuitively, each VP represents a class of indistinguishable (as long as the VPH model is concerned) patients”, a model of a patient indistinguishable from a current patient that simulates time evolution of a hormonal state caused by pharmaceutical drugs corresponds to determining historical changes in a state of a patient caused by previous adjustments, Shamanna teaches metabolic states)
Bostic further teaches:
and assigning the patient to either a first subset of patients (Bostic [0460]) “determining, at the healthcare data system computing device, whether one or more of said patient, said population of patients, and a subset of said population of patients are desired by said healthcare researcher based on the comparison of the desirability data to the health information”) sensitive to the intervention parameter or a second subset of patients insensitive to the intervention parameter based the historical changes. (((Bostic [0308]) “The platform 100 may calculate a desirability score of…the population of patients, and/or one or more patients included in the population of patients, the desirability score being based at least partially on how closely the genetic, environmental, health state, and/or lifestyle properties of…the population of patients, and/or one or more patients included in the population of patients fit criteria of suitable patients for the healthcare research program”, a desirability score based on health state or lifestyle properties for one or more patients in the population corresponds to a sensitivity of patients in the population to the intervention parameter)
At the time of filing, one of ordinary skill in the art would have motivation to combine the method jointly taught by Sinisi, Bostic, and Shamanna for the parent claim of claim 6, claim 5. No new embodiments are introduced, so the reason to combine is the same as for the parent claim.
Regarding claim 7,
Sinisi, Bostic, and Shamanna jointly teach The method of claim 1, wherein generating the cohort of patients comprises:
Shammana further teaches:
identifying, from the population of patients, a subset of patients whose metabolic state is below a threshold metabolic state; ((Shamanna Pg. 3) “Program enrollees were required to have adequate hepatic and renal function to be included in the study, with the former defined as an aspartate transaminase or alanine transaminase ratio ≤ 3-fold the upper limit of normal”)
Bostic further teaches:
and generating the cohort of patients from the subset of patients ((Bostic [0308]) “The platform 100 may calculate a desirability score of the patient, the population of patients, and/or one or more patients included in the population of patients, the desirability score being based at least partially on how closely the genetic, environmental, health state, and/or lifestyle properties of the patient, the population of patients, and/or one or more patients included in the population of patients fit criteria of suitable patients for the healthcare research program”)
At the time of filing, one of ordinary skill in the art would have motivation to combine the method jointly taught by Sinisi, Bostic, and Shamanna for the parent claim of claim 7, claim 1. No new embodiments are introduced, so the reason to combine is the same as for the parent claim.
Regarding claim 8,
Sinisi, Bostic, and Shamanna jointly teach The method of claim 1, wherein generating the cohort of patients comprises:
Bostic further teaches:
determining a long-term effect of adjustments to the intervention parameter on each patient of the population of patients ((Bostic [0296]) “In some embodiments, the digital twin module 1302 may be configured to simulate one or more potential future health states of the population of patients using one or both of the digital twin of the population of patients, and the one or more machine learning modules…The future health states of the population of patients may be simulated according to variables, such as a time frame, a treatment schedule, a prescription drug schedule, a lifestyle, potential developments in one or more health issues experienced by the population of patients, any other suitable variable for use in simulation, and/or a combination thereof. Simulating based on time frame may include simulating a potential health state of the population of patients in one or more, seconds, minutes, hours, days, months, years, or any other suitable time frame…For example, the digital twin module 1302 may simulate a health state of a population of heart disease patients according to potentially prescribing heart disease medication to the population of patients and advising that the population of patients take a small dose of aspirin regularly”) based on historical changes in the metabolic state of the patient; ((Bostic [0085]) “In embodiments, the machine learned model corresponds to the medication. In embodiments, the machine learned model is trained on a plurality of training data samples that respectively correspond to a plurality of previous patients that were prescribed the medication, wherein each training data sample includes respective patient attributes of the respective previous patient and an outcome related to the medication for the previous patient”, patient outcomes when prescribed medication corresponds to historical changes in a patient state, Shamanna teaches metabolic states)
and generating the cohort of patients (Bostic [0460]) “determining, at the healthcare data system computing device, whether one or more of said patient, said population of patients, and a subset of said population of patients are desired by said healthcare researcher based on the comparison of the desirability data to the health information”) based on the long-term effect of adjustments to the intervention parameter determined for each patient of the population of patients. (((Bostic [0308]) “The platform 100 may calculate a desirability score of…the population of patients, and/or one or more patients included in the population of patients, the desirability score being based at least partially on how closely the genetic, environmental, health state, and/or lifestyle properties of…the population of patients, and/or one or more patients included in the population of patients fit criteria of suitable patients for the healthcare research program”, a desirability score based on health state or lifestyle properties for one or more patients in the population corresponds to the long-term effect on patients in the population to the intervention parameter)
At the time of filing, one of ordinary skill in the art would have motivation to combine the method jointly taught by Sinisi, Bostic, and Shamanna for the parent claim of claim 8, claim 1. No new embodiments are introduced, so the reason to combine is the same as for the parent claim.
Regarding claim 9,
Sinisi, Bostic, and Shamanna jointly teach The method of claim 1,
Bostic further teaches:
wherein the candidate treatment recommendation comprises instructions or adjusting a plurality of intervention parameters […] ((Bostic [0130]) “In embodiments, the method includes receiving simulation instructions, the simulation instructions including one or more drug treatment regimens; simulating the one or more drug treatment regimens on one or both of said individual patient and a population of patients using at least one of the digital twin of said individual patient and the digital twin of said population of patients”)
and generating the cohort of patients (Bostic [0460]) “determining, at the healthcare data system computing device, whether one or more of said patient, said population of patients, and a subset of said population of patients are desired by said healthcare researcher based on the comparison of the desirability data to the health information”) based on the overall sensitivity of each patient of the population of patients (((Bostic [0308]) “The platform 100 may calculate a desirability score of…the population of patients, and/or one or more patients included in the population of patients, the desirability score being based at least partially on how closely the genetic, environmental, health state, and/or lifestyle properties of…the population of patients, and/or one or more patients included in the population of patients fit criteria of suitable patients for the healthcare research program”, a desirability score based on health state for one or more patients in the population corresponds sensitivity of patients in the population)
Sinisi further teaches:
[…] and generating the cohort of patients comprises: for each patient of the population of patients, determining a sensitivity of each patient ((Sinisi Pgs. 7-8) “Different value assignments to model parameters yield different model time evolutions and/or different reactions to drug administrations, thus defining different Virtual Phenotypes (VPs). Intuitively, each VP represents a class of indistinguishable (as long as the VPH model is concerned) patients. Computing a complete set of VPs for the model at hand is the starting point to obtain a representative population of virtual patients (hence, ideally showing all possible phenotypes)”) to each intervention parameter of the plurality; ((Sinisi Pg. 7) “The model defines the time evolution of overall 33 biological quantities (mostly blood concentration of hormones) and the Pharmacokinetics/Pharmacodynamics (PKPD) of several pharmaceutical drugs used in assisted reproduction (in particular, we will focus on GnRH analogues like Triptorelin) by means of highly non-linear differential equations”, hormones and administrations of drugs are intervention parameters)
and determining an overall sensitivity of the patient to the candidate treatment recommendation based on the sensitivity of the patient to each intervention parameter of the plurality; ((Sinisi Pg. 14) “Since we seek the optimal personalised treatment for the input digital twin, our algorithm does not stop at the first found successful treatment. Indeed, along the lines of [47], it keeps track of the lightest successful treatment found so far, i.e., the one envisioning the administration of the minimum overall drug amount”, a minimal amount of a drug necessary to suitably adjust intervention parameters of a patient for a treatment recommendation corresponds to a sensitivity of a patient to a treatment recommendation, the amount of drug administered is based on the sensitivity of the measured hormones to adjustment based on the administration of the drug)
At the time of filing, one of ordinary skill in the art would have motivation to combine the method jointly taught by Sinisi, Bostic, and Shamanna for the parent claim of claim 9, claim 1. No new embodiments are introduced, so the reason to combine is the same as for the parent claim.
Regarding claim 10,
Sinisi, Bostic, and Shamanna jointly teach The method of claim 1, wherein generating the cohort of patients comprises:
Bostic further teaches:
categorizing the population of patients into categories of patients with a shared [metabolic] state; ((Bostic [0012]) “and classifying the patient within a population of patients, wherein said population of patients is determined based on one or more of lifestyle, diagnosis and/or prognosis, and present or previous healthcare treatments as categorized using the machine learning module”, metabolic state taught by Shamanna)
and generating the cohort of patients ((Bostic [0460]) “determining, at the healthcare data system computing device, whether one or more of said patient, said population of patients, and a subset of said population of patients are desired by said healthcare researcher based on the comparison of the desirability data to the health information”) based on the category of patients most sensitive to the candidate treatment recommendation. ((Bostic [0308]) “The platform 100 may calculate a desirability score of…the population of patients, and/or one or more patients included in the population of patients, the desirability score being based at least partially on how closely the genetic, environmental, health state, and/or lifestyle properties of…the population of patients, and/or one or more patients included in the population of patients fit criteria of suitable patients for the healthcare research program”, a desirability score based on health state for one or more patients in the population corresponds sensitivity of patients in the population)
Sinisi further teaches:
for each category of patients, ((Sinisi Pg. 7) “Different value assignments to model parameters yield different model time evolutions and/or different reactions to drug administrations, thus defining different Virtual Phenotypes (VPs). Intuitively, each VP represents a class of indistinguishable (as long as the VPH model is concerned) patients”) predicting an effect of the candidate treatment recommendation on each patient of the category ((Sinisi Pg. 12) “Since the input of our algorithm is a digital twin P(C) computed from a patient clinical record C and consisting of a set of np ∈ N+ Virtual Phenotypes (VPs) (see Section 2.5), during the search we drive the simulation of np independent copies of our VPH model (each one connected to a copy of the monitor checking for treatment invariants and goals, as explained in Section 4.1)”) by inputting the candidate treatment recommendation to a patient-specific [metabolic] model of the patient; ((Sinisi Pg. 11) “In particular, given a model trajectory x(λ, t) under a given input u, our monitor output is UNDET as long as x(λ, t) satisfies the invariants (in other words, the monitor decision is ‘undetermined’ as long as there is hope to extend the current treatment into a successful treatment) and goes to and stays at value FAIL as soon as invariants are violated. When a UNDET model trajectory satisfies the goal conditions, the monitor output turns to SUCCESS, informing the caller that the input u(t) defines a successful treatment (i.e., an effective treatment which always satisfies invariants)”, a candidate treatment resulting in success or failure corresponds to an effect of the candidate treatment, Shamanna teaches metabolic models)
and determining an overall sensitivity of the category of patients ((Sinisi Pg. 9) “Intuitively, the digital twin of a human patient is a digital representation of the patient physiology in the form of all VPs entailing model behaviours that fit the patient clinical measurements”) to the candidate treatment recommendation based on the predicted effect of the candidate treatment recommendation on each patient of the cohort; ((Sinisi Pg. 14) “Since we seek the optimal personalised treatment for the input digital twin, our algorithm does not stop at the first found successful treatment. Indeed, along the lines of [47], it keeps track of the lightest successful treatment found so far, i.e., the one envisioning the administration of the minimum overall drug amount”, a minimal amount of a drug necessary to suitably adjust intervention parameters of a patient for a treatment recommendation corresponds to a sensitivity of a patient to a treatment recommendation)
and determining a category of patients most sensitive to the candidate treatment recommendation based on a comparison of the overall sensititvity of each category of patients; (Sinisi Pg. 23, Fig. 9 compares the dosage and timing of the drug Triptorelin for each Virtual Phenotype checked, the Virtual Phenotypes matching the patient with the least amount of the dosage, C16, are the most sensitive)
PNG
media_image1.png
587
772
media_image1.png
Greyscale
At the time of filing, one of ordinary skill in the art would have motivation to combine the method jointly taught by Sinisi, Bostic, and Shamanna for the parent claim of claim 10, claim 1. No new embodiments are introduced, so the reason to combine is the same as for the parent claim.
Regarding claim 15,
Sinisi, Bostic, and Shamanna jointly teach The method of claim 1, wherein predicting the effect of the candidate treatment recommendation on the metabolic state of the patient further comprises:
Sinisi further teaches:
encoding a feature vector representation of the candidate treatment recommendation; ((Sinisi Pg. 6) “u(t) ∈ Rnu is called input vector, and models exogenous inputs (e.g., sequences of drug administrations);”)
and inputting the feature vector representation ((Sinisi Pg. 6) “In our setting, administrations of nu different pharmaceutical drugs (clinical actions) define inputs to the VPH model. Indeed, for each i ∈ [nu], the i-th component ui of the input vector u defines a function that associates to each time instant t ∈ R0+ the administered dose of the i-th drug, to be chosen within a set Ai of possible doses (which always includes dose zero)”) into each of the patient-specific [metabolic] models of the digital twin for each patient of the cohort ((Sinisi Pg. 9) “To compute a personalised treatment for a human patient, we first need to compute a digital representation for her. This will be done by using clinical data (in the form of a clinical record C, Definition 2.2) available from that patient in order to select, from our representative population of VPs, the subset of VPs that are compatible with (i.e., fit) such data”, (Sinisi Pg. 19) “Each VP in a compact digital twin has been encoded in a Modelica (http://www.modelica.org) model (encompassing the GynCycle VPH model taking clinical actions as input, a parameter vector assignment, and a monitor to check for treatment invariants and goals) and has been simulated using JModelica v2.1”)
At the time of filing, one of ordinary skill in the art would have motivation to combine the method jointly taught by Sinisi, Bostic, and Shamanna for the parent claim of claim 15, claim 1. No new embodiments are introduced, so the reason to combine is the same as for the parent claim.
Regarding claim 19,
Sinisi, Bostic, and Shamanna jointly teach The method of claim 1, further comprising:
Sinisi further teaches:
identifying candidate treatment recommendations that cause a threshold improvement in metabolic state by comparing the predicted changes in [metabolic] states of the cohort of patients to a threshold improvement, ((Sinisi Pgs. 11-12) “a downregulation treatment is considered successful (effective) for a patient in case the blood concentrations of a given set of hormones and other physiological quantities go below certain thresholds within 9 days from the first drug administration, and stay always below such thresholds for the following 21 days”)
wherein the predicted changes in [metabolic] states represent the effectiveness of a candidate treatment recommendation; ((Sinisi Pgs. 11-12) “a downregulation treatment is considered successful (effective) for a patient in case the blood concentrations of a given set of hormones and other physiological quantities go below certain thresholds within 9 days from the first drug administration, and stay always below such thresholds for the following 21 days”, Shamanna teaches metabolic states)
extracting intervention parameters adjusted in the one or more effective treatments of the candidate treatment recommendations; ((Sinisi Pgs. 11-12) “a downregulation treatment is considered successful (effective) for a patient in case the blood concentrations of a given set of hormones and other physiological quantities go below certain thresholds”, blood concentrations of hormones are intervention parameters that are adjusted)
and generating an aggregate treatment recommendation comprising adjustments to one or more of