DETAILED ACTION
Notices to Applicant
This communication is a Final Office Action on the merits. Claims 1-2, 8, 14-15, 61-64, 68-76, and 90-91 as filed 06/11/2025, are currently pending and have been considered below.
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Priority
This application claims the benefit of and priority to U.S. Provisional Patent Application No. 63/034,368 filed on June 3, 2020, U.S. Provisional Patent Application No. 63/064,054 filed on August 11, 2020, and U.S. Provisional Patent Application No. 63/180,880 filed on April 28, 2021.
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 1-2, 8, 14-15, 61-64, 68-76, and 90-91 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception (i.e., an abstract idea) without significantly more.
Claims 1-2, 8, 14-15, 61-64, 68-76, and 90-91 are drawn to a method for subphenotyping acute respiratory distress syndrome patients, which is within the four statutory categories (i.e. method).
Independent Claim 1 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. Claim 1 recites:
1. A method, comprising:
obtaining or having obtained electronic health record (EHR) data for a subject exhibiting acute respiratory distress syndrome (ARDS); and
determining, through a patient subphenotype classifier trained on training data comprising HER data correlated to one or more subphenotypes, a classification of the subject selected from two or more subphenotypes from the EHR data for the subject without analyzing biomarker levels of the subject, wherein the patient subphenotype classifier receives at least a mean arterial pressure.
The above limitations, as drafted, is a method that, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components. That is, other than reciting the above bolded language, for example, “through a patient subphenotype classifier trained on training data,” nothing in the claim precludes the steps from practically being performed in the mind. For example, but for the “through a patient subphenotype classifier trained on training data,,” language, obtaining health data for a subject exhibiting acute respiratory distress syndrome and determining a classification of the subject from EHR data for the subject without analyzing biomarker levels of the subject while receiving at least a mean arterial pressure in the context of this claim encompasses the observation, evaluation, judgment, and/or opinion of electronic health record data for a subject. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
This judicial exception is not integrated into a practical application. In particular, the claim only recites the above bolded additional elements of “through a patient subphenotype classifier trained on training data,,” to perform the obtaining and analyzing limitations. The additional element is recited at a high-level of generality (i.e., a patient subphenotype classifier such as a machine learned model trained using a machine learning implemented method, such as any one of a linear regression algorithm, logistic regression algorithm, decision tree algorithm, support vector machine classification, Naive Bayes classification, K-Nearest Neighbor classification, random forest algorithm, deep learning algorithm, gradient boosting algorithm, and dimensionality reduction techniques such as manifold learning, principal component analysis, factor analysis, autoencoder regularization, and independent component analysis, or combinations thereof. In various embodiments, the predictive model is trained using supervised learning algorithms, unsupervised learning algorithms, semi-supervised learning algorithms (e.g., partial supervision), weak supervision, transfer, multi-task learning, or any combination thereof performed by a processor as it relates to a general purpose computer component (Application Specification [0024], [00130]-[00131]). As such, the limitations amount to no more than mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea. See MPEP 2106.05(f). Accordingly, these additional elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea.
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the above bolded additional elements, for example, “through a patient subphenotype classifier trained on training data,” to perform the claimed limitations amounts to no more than mere instructions to apply the exception using a generic computer component (i.e., a patient subphenotype classifier such as a machine learned model trained using a machine learning implemented method, such as any one of a linear regression algorithm, logistic regression algorithm, decision tree algorithm, support vector machine classification, Naive Bayes classification, K-Nearest Neighbor classification, random forest algorithm, deep learning algorithm, gradient boosting algorithm, and dimensionality reduction techniques such as manifold learning, principal component analysis, factor analysis, autoencoder regularization, and independent component analysis, or combinations thereof. In various embodiments, the predictive model is trained using supervised learning algorithms, unsupervised learning algorithms, semi-supervised learning algorithms (e.g., partial supervision), weak supervision, transfer, multi-task learning, or any combination thereof performed by a processor as it relates to a general purpose computer component (Application Specification [0024], [00130]-[00131]). Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. See MPEP 2106.05(f). The claim is not patent eligible.
Dependent claims 2, 8, 14-15, 61-64, 68-76, and 90-91 include limitations of the independent claim and are directed to the same abstract idea as discussed above and incorporated herein. The dependent claims are rejected under 35 U.S.C. § 101 because they are directed to non-statutory subject matter. These additional claims recite what the patient data is and how it is analyzed. These information characteristics do not integrate the judicial exception into a practical application, and, when viewed individually or as a whole, they do not add anything substantial beyond the observation, evaluation, judgment, and/or opinion patient data. Furthermore, the combination of elements does not indicate a significant improvement to the functioning of a computer or any other technology. Therefore the dependent claims are rejected under 35 U.S.C. § 101.
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-2, 8, 14-15, 61-62, 68-76 are rejected under 35 U.S.C. 103 as being unpatentable over Wilson, Jennifer G., and Carolyn S. Calfee. "ARDS subphenotypes: understanding a heterogeneous syndrome." Annual Update in Intensive Care and Emergency Medicine 2020 (2020): 67-79 (hereinafter “Wilson et al.”) in view of Sinha, Pratik, Matthew M. Churpek, and Carolyn S. Calfee. "Machine learning classifier models can identify acute respiratory distress syndrome phenotypes using readily available clinical data." American journal of respiratory and critical care medicine 202.7 (2020): 996-1004 (hereinafter “Sinha et al.”) and U.S. Patent Application Pub. No. 2020/0178903 A1 (hereinafter “Chaudhuri et al.”).
RE: Claim 1 (Currently Amended) Wilson et al. teaches the claimed:
1. (Currently Amended) A method, comprising: obtaining or having obtained electronic health record (EHR) data for a subject exhibiting acute respiratory distress syndrome (ARDS) ((Wilson et al., Pg. 72, Lines 11-14) (using an approach to identify subgroups within a heterogeneous population called latent class analysis, two distinct subphenotypes of ARDS were identified based on combined clinical and biologic data from patients enrolled in two large clinical trial cohorts)); and
determining, through a patient subphenotype […], a classification of the subject selected from two or more subphenotypes ((Wilson et al., Pg. 72, Lines 14-19; Pg. 76, Lines 20-22) (The “hyperinflammatory” subphenotype was characterized by enhanced inflammation, fewer ventilator-free days, and increased mortality compared to the “hypoinflammatory” subphenotype (Fig. 5.2); These two subphenotypes have been found in subsequent independent analyses of multiple other ARDS trial cohorts, and the poor prognosis associated with the hyperinflammatory phenotype persists; machine learning applied to electronic health record data and identified four subphenotypes with different generic and inflammatory markers and markedly different mortality rates)).
Wilson et al. fails to explicitly teach, but Sinha et al. teaches the claimed:
((Sinha et al., Pg. 996, Abstract, Pg. 998) (the objective of this study was to develop models to classify ARDS phenotypes using readily available clinical data only; Only data that were used in the original LCA-modeling studies and that were deemed to be readily available in routine clinical workflow at the point of trial enrollment were considered for predictor variables)).
One of ordinary skill in the art at the time of the effective filing date would have found it obvious to combine the method of developing models to classify ARDS phenotypes using readily available clinical data only as taught by Sinha et al. within the method of subphenotyping of ARDS patients as taught by Wilson et al. with the motivation of providing point-of-care assays that plasma biomarkers currently lack (Sinha et al., Pg. 996, Abstract).
a patient subphenotype classifier trained on training data comprising EHR data correlated to one or more subphenotypes Chaudhuri et al., [0032], [0034]) (the computer-implemented method and system of monitoring a patient with respect to a particular medical condition includes receiving a long time interval of time synchronized parameter data for each of at least two physiological parameters, and processing the long time interval of parameter data using a trained machine learning model to identify patterns learned by the model and calculate a risk severity index of the particular medical condition based on a set of feature values identified in the parameter data; the slope of each parameter dataset in each segment 33 may be analyzed. For example, the slope may be classified into one of the above described three stages of ARDS exemplified at FIG. 1-stage one where SpO2 is declining));
wherein the patient subphenotype classifier receives at least a mean arterial pressure ((Chaudhuri et al., [0025]) (the processing unit may be configured to process the physiological data recorded by the sensors in a blood pressure cuff to calculate systolic, diastolic, and mean blood pressure values for the patient)) .
One of ordinary skill in the art at the time of the effective filing date would have found it obvious to combine the using a trained machine learning model utilizing mean blood pressure values of a patient to ARDS stage classification as taught by Chaudhuri et al. within the method of subphenotyping of ARDS patients as taught by Wilson et al. and the method of developing models to classify ARDS phenotypes using readily available clinical data only as taught by Sinha et al. with the motivation of tracking, visualizing, and predicting the severity of a particular medical condition based on parameter data for the measured physiological data parameters (Chaudhuri et al., [0001]).
RE: Claim 2 (Currently Amended) Wilson, Sinha et al., and Chaudhuri et al. teach the claimed:
2. (Currently Amended) The method of claim 1, wherein the patient subphenotype classifier receives one or more input variables comprising heart rate, ((Wilson et al., pg. 72, Fig. 5.2; pg. 74) (hypoinflammatory subphenotype High bicarbonate, hyperflammatory subphenotype Low bicarbonate; a lower PaO2:FiO2 ratio not only identifies patients at higher risk of death, but also reflects patients with greater lung weight who may be more likely to benefit from recruitment maneuvers, higher PEEP or prone positioning)).
RE: Claim 8 (Previously Presented) Wilson, Sinha et al., and Chaudhuri et al. teach the claimed:
8. (Currently Amended) The method of claim 2,wherein the patient subphenotype classifier further receives one or more input variables comprising partial pressure of carbon dioxide, PaO2/FiO2, platelet count, age, gender, positive end- expiratory pressure, tidal volume, body mass index, plateau pressure, minute ventilation, and vasopressor use in prior 24 hours ((Wilson et al., pg. 72, Fig. 5.2; pg. 74) (hypoinflammatory subphenotype with higher ventilator-free days, hyperflammatory subphenotype with lower ventilator-free days; a lower PaO2:FiO2 ratio not only identifies patients at higher risk of death, but also reflects patients with greater lung weight who may be more likely to benefit from recruitment maneuvers, higher PEEP or prone positioning; Similarly, one could hypothesize that among patients with severe ARDS, those who have a plateau pressure >30 cmH2O or an unfavorable driving pressure despite adherence to a lung-protective ventilation strategy may be more likely to benefit from “lung rest” with ultra-low tidal volumes on ECMO)).
RE: Claim 14 (Original) Wilson, Sinha et al., and Chaudhuri et al. teach the claimed:
14. (Original) The method of claim 1, wherein the patient subphenotype classifier comprises:(A) a subphenotyping submodel that outputs a prediction for an ARDS subphenotype ((Sinha et al., pg. 999) (for the primary analysis, the final tuned clinical classifier model to assign the hyperinflammatory phenotype)); and (B) a mortality submodel that outputs a prediction of an ARDS mortality rate ((Sinha et al., pg. 999) (to evaluate differential treatment responses, logistic regression models were constructed by introducing interaction terms of phenotype assignment and treatment, with mortality at Day 90 as the outcome variable)).
One of ordinary skill in the art at the time of the effective filing date would have found it obvious to combine the method of a clinical classifier for an ARDS phenotype assignment and with mortality as the outcome variable as taught by Sinha et al. within the method of subphenotyping of ARDS patients as taught by Wilson et al. and the using a trained machine learning model utilizing mean blood pressure values of a patient to ARDS stage classification as taught by Chaudhuri et al. with the motivation of providing point-of-care assays that plasma biomarkers currently lack (Sinha et al., Pg. 996, Abstract).
RE: Claim 15 (Original) Wilson, Sinha et al., and Chaudhuri et al. teach the claimed:
15. (Original) The method of claim 14, wherein the prediction for the ARDS subphenotype outputted by the subphenotyping submodel serves as an input to the mortality submodel ((Sinha et al., pg. 999) (for the primary analysis, the final tuned clinical classifier model to assign the hyperinflammatory phenotype)); and (B) a mortality submodel that outputs a prediction of an ARDS mortality rate ((Sinha et al., pg. 999-1000) (to evaluate differential treatment responses, logistic regression models were constructed by introducing interaction terms of phenotype assignment and treatment, with mortality at Day 90 as the outcome variable; When phenotype was assigned by the clinical classifier model, mortality at Day 90 was significantly higher in the hyperinflammatory phenotype compared with the hypoinflammatory phenotype)).
One of ordinary skill in the art at the time of the effective filing date would have found it obvious to combine the method of a clinical classifier for an ARDS phenotype assignment and with mortality as the outcome variable as taught by Sinha et al. within the method of subphenotyping of ARDS patients as taught by Wilson et al. and the using a trained machine learning model utilizing mean blood pressure values of a patient to ARDS stage classification as taught by Chaudhuri et al. with the motivation of providing point-of-care assays that plasma biomarkers currently lack (Sinha et al., Pg. 996, Abstract).
RE: Claim 61 (Previously Presented) Wilson, Sinha et al., and Chaudhuri et al. teach the claimed:
61. (Previously Presented) The method of claim 1 wherein the patient subphenotype classifier is trained using a training dataset comprising patient data from one or more clinical trial datasets ((Sinha et al., pg. 996, Abstract) (Three randomized controlled trial cohorts served as the training data set (ARMA [High vs. Low VT], ALVEOLI [Assessment of Low VT and Elevated End-Expiratory Pressure to Obviate Lung Injury], and FACTT [Fluids and Catheter Treatment Trial]; n = 2,022), and a fourth served as the validation data set (SAILS [Statins for Acutely Injured Lungs from Sepsis]; n = 745). A gradient boosted machine algorithm was used to develop classifier models using 24 variables (demographics, vital signs, laboratory, and respiratory variables) at enrollment)).
One of ordinary skill in the art at the time of the effective filing date would have found it obvious to combine training of a classifier model utilizing controlled trial cohorts served as the training data set as taught by Sinha et al. within the method of subphenotyping of ARDS patients as taught by Wilson et al. and the using a trained machine learning model utilizing mean blood pressure values of a patient to ARDS stage classification as taught by Chaudhuri et al. with the motivation of providing point-of-care assays that plasma biomarkers currently lack (Sinha et al., Pg. 996, Abstract).
RE: Claim 62 (Original) Wilson, Sinha et al., and Chaudhuri et al. teach the claimed:
62. (Original) The method of claim 61, wherein the one or more clinical trial datasets are any of ARMA dataset, KARMA dataset, LARMA dataset, ALVEOLI dataset, EDEN dataset, FACTT dataset, SAILS dataset, ROSE dataset, eICU-CRD dataset, and the Brazilian ART dataset ((Sinha et al., pg. 996, Abstract) (Three randomized controlled trial cohorts served as the training data set (ARMA [High vs. Low VT], ALVEOLI [Assessment of Low VT and Elevated End-Expiratory Pressure to Obviate Lung Injury], and FACTT [Fluids and Catheter Treatment Trial]; n = 2,022), and a fourth served as the validation data set (SAILS [Statins for Acutely Injured Lungs from Sepsis]; n = 745). A gradient boosted machine algorithm was used to develop classifier models using 24 variables (demographics, vital signs, laboratory, and respiratory variables) at enrollment)).
One of ordinary skill in the art at the time of the effective filing date would have found it obvious to combine training of a classifier model utilizing controlled trial cohorts served as the training data set such as ARMA, ALVEOLI, FACTT, SAILS, etc. as taught by Sinha et al. within the method of subphenotyping of ARDS patients as taught by Wilson et al. and the using a trained machine learning model utilizing mean blood pressure values of a patient to ARDS stage classification as taught by Chaudhuri et al. with the motivation of providing point-of-care assays that plasma biomarkers currently lack (Sinha et al., Pg. 996, Abstract).
RE: Claim 68 (Previously Presented) W Wilson, Sinha et al., and Chaudhuri et al. teach the claimed:
68. (Previously Presented) A method for identifying a mortality prognosis for a subject, the method comprising: obtaining a classification of the subject exhibiting acute respiratory distress syndrome (ARDS), the classification of the subject selected from two or more subphenotypes and determined using the method of claim 1 ((Wilson et al., Pg. 72, Lines 11-19; Pg. 76, Lines 20-22) (using an approach to identify subgroups within a heterogeneous population called latent class analysis, two distinct subphenotypes of ARDS were identified based on combined clinical and biologic data from patients enrolled in two large clinical trial cohorts; “hyperinflammatory” subphenotype was characterized by enhanced inflammation, fewer ventilator-free days, and increased mortality compared to the “hypoinflammatory” subphenotype (Fig. 5.2); These two subphenotypes have been found in subsequent independent analyses of multiple other ARDS trial cohorts, and the poor prognosis associated with the hyperinflammatory phenotype persists; machine learning applied to electronic health record data and identified four subphenotypes with different generic and inflammatory markers and markedly different mortality rates)); and
identifying a mortality prognosis for the subject based at least in part on the classification, wherein responsive to the classification of the subject comprising subphenotype B from the two or more subphenotypes, the mortality prognosis identified for the subject comprises high mortality risk, and wherein responsive to the classification of the subject comprising subphenotype A from the two or more subphenotypes, the mortality prognosis identified for the subject comprises low mortality risk ((Wilson et al., Pg. 72, Lines 11-19) (using an approach to identify subgroups within a heterogeneous population called latent class analysis, two distinct subphenotypes of ARDS were identified based on combined clinical and biologic data from patients enrolled in two large clinical trial cohorts; “hyperinflammatory” subphenotype was characterized by enhanced inflammation, fewer ventilator-free days, and increased mortality compared to the “hypoinflammatory” subphenotype (Fig. 5.2); These two subphenotypes have been found in subsequent independent analyses of multiple other ARDS trial cohorts, and the poor prognosis associated with the hyperinflammatory phenotype persists)).
RE: Claim 69 (Original) Wilson, Sinha et al., and Chaudhuri et al. teach the claimed:
69. (Original) The method of claim 68, wherein low mortality risk comprises at least one of reduced risk of hospital mortality, reduced risk of ICU mortality, reduced risk of 28-day mortality, reduced risk of 90-day mortality, reduced risk of 180-day mortality, and reduced risk of 6-month mortality relative to high mortality risk ((Wilson et al., pg. 70, lines 10-11; pg. 75, Table 5.3) (Several studies have shown that ARDS onset >48 h after ICU admission is associated with higher mortality; 90-mortality of Interventional/trial cohorts)).
RE: Claim 70 (Previously Presented) Wilson, Sinha et al., and Chaudhuri et al. teach the claimed:
70. (Previously Presented) The method of claim 68, wherein low mortality risk further comprises positive patient outcome, wherein high mortality risk further comprises negative patient outcome, and wherein positive patient outcome comprises at least one of shorter hospital length of stay, shorter ICU length of stay and more ventilator-free days relative to negative patient outcome ((Wilson et al., Pg. 72, Lines 14-19) (The “hyperinflammatory” subphenotype was characterized by enhanced inflammation, fewer ventilator-free days, and increased mortality compared to the “hypoinflammatory” subphenotype (Fig. 5.2))).
RE: Claim 71 (Previously Presented) Wilson, Sinha et al., and Chaudhuri et al. teach the claimed:
71. (Previously Presented) The method of claim 1, further identifying a therapy recommendation for the subject based at least in part on the classification, wherein responsive to the classification of the subject comprising subphenotype B from the two or more subphenotypes, the therapy recommendation identified for the subject comprises one or more of neuromuscular blockade (NMB) therapy or no NMB therapy, high PEEP or low PEEP, no treatment or methylprednisolone, dexamethasone, no lisofylline, ketoconazole, catheter and fluid treatment, recruitment maneuver, statins, or full or trophic enteral feeding and wherein responsive to the classification of the subject comprising subphenotype A from the two or more subphenotypes, the therapy recommendation identified for the subject comprises one or more of NMB therapy, low PEEP therapy, no methylprednisolone, no treatment or dexamethasone, no treatment or lisofylline, no treatment or ketoconazole, no combination of catheter and fluid treatment, no recruitment maneuver, statins as a preemptive therapy, or full enteral feeding ((Wilson et al., pg. 74, lines 1-7; pg. 75, Table 5.3) (a lower PaO2:FiO2 ratio not only identifies patients at higher risk of death, but also reflects patients with greater lung weight who may be more likely to benefit from recruitment maneuvers, higher PEEP or prone positioning; Similarly, one could hypothesize that among patients with severe ARDS, those who have a plateau pressure >30 cmH2O or an unfavorable driving pressure despite adherence to a lung-protective ventilation strategy may be more likely to benefit from “lung rest” with ultra-low tidal volumes on ECMO; subphenotype hypoinflammatory response and hyperinflammatory response to high vs low PEEP and 90-day mortality)).
RE: Claim 72 (Previously Presented) Wilson, Sinha et al., and Chaudhuri et al. teach the claimed:
72. (Previously Presented) A method for identifying candidate subjects to be provided a therapy, the method comprising: for one or more subjects, obtaining a classification of the subject exhibiting acute respiratory distress syndrome (ARDS), the classification of the subject selected from two or more subphenotypes and determined using the method of claim 1 ((Wilson et al., Pg. 72, Lines 11-19; Pg. 76, Lines 20-22) (using an approach to identify subgroups within a heterogeneous population called latent class analysis, two distinct subphenotypes of ARDS were identified based on combined clinical and biologic data from patients enrolled in two large clinical trial cohorts; “hyperinflammatory” subphenotype was characterized by enhanced inflammation, fewer ventilator-free days, and increased mortality compared to the “hypoinflammatory” subphenotype (Fig. 5.2); These two subphenotypes have been found in subsequent independent analyses of multiple other ARDS trial cohorts, and the poor prognosis associated with the hyperinflammatory phenotype persists; machine learning applied to electronic health record data and identified four subphenotypes with different generic and inflammatory markers and markedly different mortality rates)); and
determining whether the subject is a candidate subject based at least in part on the classification ((Wilson et al., pg. 74, lines 1-7) (a lower PaO2:FiO2 ratio not only identifies patients at higher risk of death, but also reflects patients with greater lung weight who may be more likely to benefit from recruitment maneuvers, higher PEEP or prone positioning; Similarly, one could hypothesize that among patients with severe ARDS, those who have a plateau pressure >30 cmH2O or an unfavorable driving pressure despite adherence to a lung-protective ventilation strategy may be more likely to benefit from “lung rest” with ultra-low tidal volumes on ECMO)).
RE: Claim 73 (Original) Wilson, Sinha et al., and Chaudhuri et al. teach the claimed:
73. (Original) The method of claim 72, wherein the therapy is a neuromuscular blockade (NMB) therapy, and wherein determining whether the subject is a candidate subject comprises determining that the subject is a likely responder responsive to the classification of the subject comprising subphenotype A from the two or more subphenotypes ((Wilson et al., pg. 69, lines 4-10) (Multiple large clinical ARDS trials have used the PaO2:FiO2 ratio for prognostic enrichment. For example, the ACURASYS trial of early continuous neuromuscular blockade, the PROSEVA trial of prone positioning, and the ROSE trial reevaluating early continuous neuromuscular blockade all targeted patients with moderate-to-severe ARDS (PaO2:FiO2 ratio < 150 mmHg). All three of these trials had mortality endpoints, and all three had mortality rates in the control arms that exceeded 40%)).
RE: Claim 74 (Original) Wilson, Sinha et al., and Chaudhuri et al. teach the claimed:
74. (Original) The method of claim 72, wherein the therapy is a neuromuscular blockade (NMB) therapy, and wherein determining whether the subject is a candidate subject comprises determining that the subject is unlikely to be a responder responsive to the classification of the subject comprising subphenotype B from the two or more subphenotypes ((Wilson et al., pg. 69, lines 4-10) (Multiple large clinical ARDS trials have used the PaO2:FiO2 ratio for prognostic enrichment. For example, the ACURASYS trial of early continuous neuromuscular blockade, the PROSEVA trial of prone positioning, and the ROSE trial reevaluating early continuous neuromuscular blockade all targeted patients with moderate-to-severe ARDS (PaO2:FiO2 ratio < 150 mmHg). All three of these trials had mortality endpoints, and all three had mortality rates in the control arms that exceeded 40%)).
RE: Claim 75 (Original) Wilson, Sinha et al., and Chaudhuri et al. teach the claimed:
75. (Original) The method of claim 72, wherein the therapy is a low positive end-expiratory pressure (PEEP) treatment, and wherein determining whether the subject is a candidate subject comprises determining that the subject is likely to be a responder responsive to the classification of the subject comprising subphenotype A from the two or more subphenotypes ((Wilson et al., pg. 74, lines 1-7; pg. 75, Table 5.3) (a lower PaO2:FiO2 ratio not only identifies patients at higher risk of death, but also reflects patients with greater lung weight who may be more likely to benefit from recruitment maneuvers, higher PEEP or prone positioning; Similarly, one could hypothesize that among patients with severe ARDS, those who have a plateau pressure >30 cmH2O or an unfavorable driving pressure despite adherence to a lung-protective ventilation strategy may be more likely to benefit from “lung rest” with ultra-low tidal volumes on ECMO; subphenotype hypoinflammatory response and hyperinflammatory response to high vs low PEEP and 90-day mortality)).
RE: Claim 76 (Original) Wilson, Sinha et al., and Chaudhuri et al. teach the claimed:
76. (Original) The method of claim 72, wherein the therapy is a high positive end-expiratory pressure (PEEP) treatment, and wherein determining whether the subject is a candidate subject comprises determining that the subject is likely to be a responder responsive to the classification of the subject comprising subphenotype B from the two or more subphenotypes ((Wilson et al., pg. 74, lines 1-7; pg. 75, Table 5.3) (a lower PaO2:FiO2 ratio not only identifies patients at higher risk of death, but also reflects patients with greater lung weight who may be more likely to benefit from recruitment maneuvers, higher PEEP or prone positioning; Similarly, one could hypothesize that among patients with severe ARDS, those who have a plateau pressure >30 cmH2O or an unfavorable driving pressure despite adherence to a lung-protective ventilation strategy may be more likely to benefit from “lung rest” with ultra-low tidal volumes on ECMO; subphenotype hypoinflammatory response and hyperinflammatory response to high vs low PEEP and 90-day mortality)).
Claims 63-64 are rejected under 35 U.S.C. 103 as being unpatentable over Wilson, Jennifer G., and Carolyn S. Calfee. "ARDS subphenotypes: understanding a heterogeneous syndrome." Annual Update in Intensive Care and Emergency Medicine 2020 (2020): 67-79 (hereinafter “Wilson et al.”) in view of Sinha, Pratik, Matthew M. Churpek, and Carolyn S. Calfee. "Machine learning classifier models can identify acute respiratory distress syndrome phenotypes using readily available clinical data." American journal of respiratory and critical care medicine 202.7 (2020): 996-1004 (hereinafter “Sinha et al.”), and U.S. Patent Application Pub. No. 2020/0178903 A1 (hereinafter “Chaudhuri et al.”), and further in view of Rice, Todd W., et al. "Comparison of the SpO2/FIO2 ratio and the PaO2/FIO2 ratio in patients with acute lung injury or ARDS." Chest 132.2 (2007): 410-417 (hereinafter “Rice et al.”).
RE: Claim 63 (Previously Presented) Wilson, Sinha et al., and Chaudhuri et al. teach the claimed:
63. (Previously Presented) The method of claim 61, wherein the patient data is derived from a sub-cohort of patients of the one or more clinical trial datasets ((Wilson et al., Pg. 72, Lines 11-14) (using an approach to identify subgroups within a heterogeneous population called latent class analysis, two distinct subphenotypes of ARDS were identified based on combined clinical and biologic data from patients enrolled in two large clinical trial cohorts)).
Wilson et al., Sinha et al., and Chaudhuri fail to explicitly teach, but Rice et al. teaches the claimed:
wherein the sub-cohort of patients are characterized by having a ratio of arterial oxygen concentration to the fraction of inspired oxygen (P/F ratio) of less than or equal to 200 ((Rice et al., pg. 410) (Acute hypoxic respiratory failure, as defined by the Pao2/fraction of inspired oxygen (Fio2) ratio (or P/F ratio) is one of the criteria for ALI/ARDS that was developed by an American European Consensus Conference (AECC) in 1994; A P/F ratio ≤ 300 and ≤ 200, respectively, are utilized to define ALI and ARDS)).
One of ordinary skill in the art at the time of the effective filing date would have found it obvious to combine the criteria for ALI/ARDS of a P/F ratio ≤ 300 and ≤ 200, respectively, as taught by Rice et al. within the method of subphenotyping of ARDS patients as taught by Wilson et al., the method of developing models to classify ARDS phenotypes using readily available clinical data only as taught by Sinha et al., and the using a trained machine learning model utilizing mean blood pressure values of a patient to ARDS stage classification as taught by Chaudhuri et al. with the motivation of providing minimally invasive approaches for defining ALI and ARDS (Rice et al., pg. 410).
RE: Claim 64 (Previously Presented) Wilson, Sinha et al., and Chaudhuri et al. teach the claimed:
64. (Previously Presented) The method of claim 61, wherein the patient data is derived from a sub-cohort of patients of the one or more clinical trial datasets ((Wilson et al., Pg. 72, Lines 11-14) (using an approach to identify subgroups within a heterogeneous population called latent class analysis, two distinct subphenotypes of ARDS were identified based on combined clinical and biologic data from patients enrolled in two large clinical trial cohorts)).
Wilson et al., Sinha et al., and Chaudhuri fail to explicitly teach, but Rice et al. teaches the claimed:
wherein the sub-cohort of patients are characterized by having a ratio of arterial oxygen concentration to the fraction of inspired oxygen (P/F ratio) of less than or equal to 300 ((Rice et al., pg. 410) (Acute hypoxic respiratory failure, as defined by the Pao2/fraction of inspired oxygen (Fio2) ratio (or P/F ratio) is one of the criteria for ALI/ARDS that was developed by an American European Consensus Conference (AECC) in 1994; A P/F ratio ≤ 300 and ≤ 200, respectively, are utilized to define ALI and ARDS)).
One of ordinary skill in the art at the time of the effective filing date would have found it obvious to combine the criteria for ALI/ARDS of a P/F ratio ≤ 300 and ≤ 200, respectively, as taught by Rice et al. within the method of subphenotyping of ARDS patients as taught by Wilson et al., the method of developing models to classify ARDS phenotypes using readily available clinical data only as taught by Sinha et al., and the using a trained machine learning model utilizing mean blood pressure values of a patient to ARDS stage classification as taught by Chaudhuri et al. with the motivation of providing minimally invasive approaches for defining ALI and ARDS (Rice et al., pg. 410).
Claims 90-91 are rejected under 35 U.S.C. 103 as being unpatentable over Wilson, Jennifer G., and Carolyn S. Calfee. "ARDS subphenotypes: understanding a heterogeneous syndrome." Annual Update in Intensive Care and Emergency Medicine 2020 (2020): 67-79 (hereinafter “Wilson et al.”) in view of Sinha, Pratik, Matthew M. Churpek, and Carolyn S. Calfee. "Machine learning classifier models can identify acute respiratory distress syndrome phenotypes using readily available clinical data." American journal of respiratory and critical care medicine 202.7 (2020): 996-1004 (hereinafter “Sinha et al.”), and U.S. Patent Application Pub. No. 2020/0178903 A1 (hereinafter “Chaudhuri et al.”), and further in view of Zhang, Zhongheng. "Identification of three classes of acute respiratory distress syndrome using latent class analysis." PeerJ 6 (2018): e4592. (hereinafter “Zhang et al.”).
RE: Claim 90 (Previously Presented) Wilson, Sinha et al., and Chaudhuri et al. teach the claimed:
90. (Previously Presented) The method of claim 72,
Wilson et al., Sinha et al., and Chaudhuri fail to explicitly teach, but Zhang et al. teaches the claimed:
wherein the therapy is full enteral feeding, and wherein determining whether the subject is a candidate subject comprises determining that the subject is likely to be a responder responsive to the classification of the subject comprising subphenotype A from the two or more subphenotypes ((Zhang et al., pg. 2, lines 32-39) (The study was a secondary analysis of the early versus delayed enteral feeding to treat people with acute lung injury or acute respiratory distress syndrome (EDEN) study; The study randomized 1,000 patients within 48 h of developing ARDS requiring mechanical ventilation in approximately equal numbers to receive either trophic or full enteral feeding for the first six days; The effectiveness of both interventions was comparable with regard to clinical outcomes such as 60-day mortality, ventilator-free days and infectious complications)).
One of ordinary skill in the art at the time of the effective filing date would have found it obvious to combine the analysis for early versus delated enteral feeding to treat people with ARDS as taught by Zhang et al. within the method of subphenotyping of ARDS patients as taught by Wilson et al., the method of developing models to classify ARDS phenotypes using readily available clinical data only as taught by Sinha et al., and the using a trained machine learning model utilizing mean blood pressure values of a patient to ARDS stage classification as taught by Chaudhuri et al. with the motivation of identifying subtypes of ARDS to guideline clinical treatment and trial design (Zhang et al., pg. 1).
RE: Claim 91 (Previously Presented) Wilson, Sinha et al., and Chaudhuri et al. teach the claimed:
91. (Previously Presented) The method of claim 72,
Wilson et al., Sinha et al., and Chaudhuri fail to explicitly teach, but Zhang et al. teaches the claimed:
wherein the therapy is trophic enteral feeding, and wherein determining whether the subject is a candidate subject comprises determining that the subject is likely to be a responder responsive to the classification of the subject comprising subphenotype B from the two or more subphenotypes ((Zhang et al., pg. 2, lines 32-39) (The study was a secondary analysis of the early versus delayed enteral feeding to treat people with acute lung injury or acute respiratory distress syndrome (EDEN) study; The study randomized 1,000 patients within 48 h of developing ARDS requiring mechanical ventilation in approximately equal numbers to receive either trophic or full enteral feeding for the first six days; The effectiveness of both interventions was comparable with regard to clinical outcomes such as 60-day mortality, ventilator-free days and infectious complications)).
One of ordinary skill in the art at the time of the effective filing date would have found it obvious to combine the analysis for early versus delated enteral feeding to treat people with ARDS as taught by Zhang et al. within the method of subphenotyping of ARDS patients as taught by Wilson et al., the method of developing models to classify ARDS phenotypes using readily available clinical data only as taught by Sinha et al. and the using a trained machine learning model utilizing mean blood pressure values of a patient to ARDS stage classification as taught by Chaudhuri et al. with the motivation of identifying subtypes of ARDS to guideline clinical treatment and trial design (Zhang et al., pg. 1).
Response to Arguments
Applicant's arguments filed 06/11/2025 have been fully considered but they are not persuasive. Applicant’s arguments will be addressed herein below in the order in which they appear in the response filed on 06/11/2025.
In the remarks, Applicant argues in substance that:
Regarding the 101 rejection of claims 1-2, 8, 14-15, 61-64, 68-76, and 90-91, Applicant argues the claims are note directed to an abstract idea, the claims are integrated into a practical application, and amount to significantly more than an abstract idea; and
Regarding the 103 rejection of claims 1-2, 8, 14-15, 61-64, 68-76, and 90-91, Applicant argues the cited prior art fails to disclose the newly amended feature of a mean arterial pressure and further the combination of Wilson and Sinha is improper.
In response to Applicant’s argument that (a) regarding the 101 rejection of claims 1-2, 8, 14-15, 61-64, 68-76, and 90-91, Examiner respectfully disagrees.
First, Applicant argues under “Pathway A” in the Remarks that the claims recite a technological improvement through the advantage of the instant claims ability to classify ARDS without the need for incorporation of biomarkers, which would otherwise problematically increase the complexity of the model. Examiner respectfully disagrees and submits that the analysis of classifying ARDS without the need for incorporation of biomarkers is not a technological solution rooted in technology solving a technical problem, but rather, is an alleged improvement of the Mental Process step of evaluation. That is, Applicant’s arguments are directed to allegedly improving an abstract idea, i.e. classifying ARDS without the use of biomarkers, but this is not a technical problem as it relates to the 101 analysis.
Second, under “Pathway B”, Applicant argues that the claims are not directed to a Mental process because the claim “as currently recited is directed to improving the process of classification of patients to subphenotypes.” See Remarks at pg. 9. Examiner respectfully submits that Applicant’s argument here shows the exact rationale for concluding that the claim is directed to a Mental Process i.e. classification of patients to subphenotypes, through the recitation of at least the steps of obtaining or having obtained electronic health record (EHR) data for a subject exhibiting acute respiratory distress syndrome (ARDS); and determining a classification of the subject selected from two or more subphenotypes from the EHR data for the subject without analyzing biomarker levels of the subject. That is, but for the recitation of a general purpose computer component of a trained subphenotype classifier, the claim is directed to an abstract idea under Step 2A, Prong 1.
Under Step 2A, Prong 2, Applicant argues that the claim limitations impose meaningful limitations that is more than a drafting effort to monopolize the judicial exception. Examiner respectfully disagrees. The abstract idea is not integrated into a practical application. In particular, the claim only recites the above bolded additional elements of “through a patient subphenotype classifier trained on training data,,” to perform the obtaining and analyzing limitations. The additional element is recited at a high-level of generality (i.e., a patient subphenotype classifier such as a machine learned model trained using a machine learning implemented method, such as any one of a linear regression algorithm, logistic regression algorithm, decision tree algorithm, support vector machine classification, Naive Bayes classification, K-Nearest Neighbor classification, random forest algorithm, deep learning algorithm, gradient boosting algorithm, and dimensionality reduction techniques such as manifold learning, principal component analysis, factor analysis, autoencoder regularization, and independent component analysis, or combinations thereof. In various embodiments, the predictive model is trained using supervised learning algorithms, unsupervised learning algorithms, semi-supervised learning algorithms (e.g., partial supervision), weak supervision, transfer, multi-task learning, or any combination thereof performed by a processor as it relates to a general purpose computer component (Application Specification [0024], [00130]-[00131]). As such, the limitations amount to no more than mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea. See MPEP 2106.05(f). Accordingly, these additional elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea.
Applicant argues that the claim recites an improvement to the functioning of a computer or to any other technology or technical field. Examiner respectfully disagrees. As discussed above under the “Pathway A” analysis, the claim limitations do not recite any improvement to technology or the functioning of a computer, but rather, an alleged improvement in the abstract idea itself i.e. ARDS classification.
Applicant argues that the claim does not fall under the MPEP 2106.05(f) Mere Instructions to Apply an Exception. Examiner respectfully disagrees. The additional element of the claim is recited at a high level of generality, determining “through a patient subphenotype classifier trained on training data”. The claim limitation “applies” a “patient subphenotype classifier” to perform the Mental Process step of determining i.e. evaluation/judgment/opinion a subphenotype classification. This is a mere instruction step to perform the abstract idea. The patient subphenotype classifier is recited at a high level of generality i.e. a machine learning model executed via instructions on a processor as it relates to a general purpose computer component, wherein the machine learning model is trained using any one of a linear regression algorithm, logistic regression algorithm, decision tree algorithm, support vector machine classification, Naive Bayes classification, K-Nearest Neighbor classification, random forest algorithm, deep learning algorithm, gradient boosting algorithm, and dimensionality reduction techniques such as manifold learning, principal component analysis, factor analysis, autoencoder regularization, and independent component analysis, or combinations thereof. In various embodiments, the predictive model is trained using supervised learning algorithms, unsupervised learning algorithms, semi-supervised learning algorithms (e.g., partial supervision), weak supervision, transfer, multi-task learning, or any combination thereof. See Application Specification [0024], [00130]-[00131]. For this reason, the claim invokes applying a general purpose computer component as a tool to perform the abstract idea.
Under “Pathway C” Applicant argues under Step 2B that the claims amount to significantly more than the abstract idea. Examiner respectfully disagrees. As discussed above under Step 2A, the additional elements of the claim i.e. determining “through a patient subphenotype classifier trained on training data”. invokes a general purpose computer component I.e. a processor, to perform machine learning at a high level of generality, such that the claim limitations amount to mere instructions to perform the Mental Process using a computer as a tool. See Application Specification [0024], [00130]-[00131]; MPEP 2106.05(f).
Accordingly, Examiner respectfully maintains the 101 rejection as applied in the above Office Action.
In response to Applicant’s argument that (b) regarding the 103 rejection of claims 1-2, 8, 14-15, 61-64, 68-76, and 90-91, Examiner re