DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Amendment
This action is in response to the amendments filed on 09/04/2025. Claims 1, 5, 24, and 35 are amended, claims 2, 3, 6, 10-21, 23, 25, 27, 29, 30, 31-34, and 36-55 were previously cancelled. Claims 1, 4, 5, 7-9, 22, 24, 26, 28, and 35 are currently pending.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1, 4-5, 7-9, 22, 24, 26, 28, and 35 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Regarding claim 1, in light of the Applicant’s most recent arguments, the limitation “accessing data from the at least one database, the data comprising: a longitudinal opioid-benzodiazepine dosage pattern over time of the patient” is indefinite. In Applicant’s arguments regarding the specification having support for “accessing data from the at least one database, the data comprising: a longitudinal opioid-benzodiazepine dosage pattern over time of the patient” the Applicant states “As described in FIG. 1A, the opioid use management system “may be configured to obtain[] subject data” –in myriad ways, including “from another system”. This seems to state that the “longitudinal opioid-benzodiazepine dosage pattern” is the obtained “subject data”. The Examiner points to paragraph [0050] which states, in part, “The subject data 132 may include values of various different predictors such as those shown in Table 1 and Table 2 above.” (emphasis added) Based on this, it is demonstrated that the “longitudinal opioid-benzodiazepine dosage pattern” is found in Table 1 or Table 2. The Examiner notes that despite the Tables of prior art Lo Ciganic being identical to that of the instant invention, the Applicant has argued that the tables of the prior art do not disclose a “longitudinal opioid-benzodiazepine dosage pattern”.
The Applicant’s arguments further point to Fig. 12 in attempting to demonstrate that the prior art differs from the instant invention. The Examiner points to paragraph [0094] which state, in part, “Accordingly, the inventors have developed techniques that use information about opioid and BZD dosage combination use patterns over time in a subject to predict a risk of OUD and/or an opioid overdose episode. The techniques use a statistical model to determine a longitudinal opioid-BZD dosage pattern of a subject, and determine a risk of OUD and/or an opioid overdose episode within a period of time based on the identified patterns. A longitudinal dosage pattern over time may also be referred to herein as a "pattern" or a "trajectory".”, and [0102] which states, in part, “In some embodiments, the opioid use management system 100 described herein with reference to FIGs. lA-lB may be configured to obtain measurements of opioid and BZD dosage of a subject over a period of time ( e.g., 1 month, 2 months, 3 months, 4 months, 5 months, 6 months, 7 months, 8 months, 9 months, 10 months, 11 months, or 12 months). The system 100 may be configured to categorize the subject into one of the trajectories ( e.g., illustrated in FIG. 12).” (emphasis added) Based on this, it appears as if the “longitudinal opioid-benzodiazepine dosage pattern” is determined by the system and displayed in Fig. 12.
In conclusion it is unclear if the “longitudinal opioid-benzodiazepine dosage pattern” refers to the data being obtained from Tables 1 and 2 (which is disclosed by the prior art), or if it is in reference to the data being determined by the system based on data obtained from Tables 1 and 2. For sake of examination, the Examiner shall assume the former since the claim limitation calls for accessing data from a database.
Dependent claims are rejected as well since they inherit the limitations of the independent claim.
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, 4, 5, 7-9, 22, 24, 26, 28, and 35 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. Representative claim 1 recites (additional elements crossed out):
A method for predicting risk of incident opioid use disorder (OUD) and/or of an opioid overdose episode for a patient
accessing data from the at least one database the data comprising:
a longitudinal opioid-benzodiazepine dosage pattern over time of the patient;
values for at least 50 predictors from shown in Table 1 associated with the patient; and
values for one or more social determinants of health associated with the patient,
generating machine learning model input features using the data accessed from the at least one database; and
processing the machine learning model input features
The above limitations as drafted, is a process that, under its broadest reasonable interpretation covers mathematical calculations and/or mental processes. That is, other than reciting the steps as being performed by a “software platform” comprising a “software application program” and at least one “database”, at least one “computer hardware processor” and a “trained machine learning model” nothing in the claim precludes the steps as being described as mathematical calculations, mental processes, and following rules or instructions. For example, but for the “software platform” comprising a “software application program” and at least one “database”, and at least one “computer hardware processor” language, the limitations describe a method for determining data to be put into a model, and inputting the data into a model to obtain an output indicating a risk of OUD and/or opioid overdose episode. The limitations of “accessing data from the at least one database, the data comprising…” and, “generating machine learning model input features using the data accessed from the at least one database” describe actions that may be performed mentally or with pen and paper. Further, but for the “GUI” recitation, the limitations of “…displaying a risk categorization of the patient…”, “…displaying one or more recommended actions determined using the risk categorization of the patient”, and “…allowing a user of the software application program to select at least one recommended action” describe actions that may be performed with pen and paper. If a claim limitation, under its broadest reasonable interpretation, describes steps that may be performed mentally or with pen and paper, then it falls within the “Mental Processes” grouping of abstract ideas. The limitation of “processing the machine learning model input features with a trained regularized multivariable logistic regression machine learning model trained based on diagnosis data indicating diagnosis of OUD and/or occurrence of opioid overdose for a plurality of subjects to obtain output indicative of the risk of OUD and/or the opioid overdose episode for the patient” describe the performance of mathematical calculations (i.e., entering values into a mathematical model). If a claim limitation, under its broadest reasonable interpretation, describes mathematical calculations then it falls within the “Mathematical Concepts” grouping of abstract ideas. Further, all of the limitations describe following rules or instructions. If a claim limitation, under its broadest reasonable interpretation, describes following rules or instructions then it falls within the “Certain Methods of Organizing Human Activity” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
The judicial exception is not integrated into a practical application. In particular, the claims recite the additional elements of a “software platform” comprising a “software application program” and at least one “database”, and at least one “computer hardware processor” to perform the steps. This additional elements are recited at a high level of generality (see at least Paras. [0104]-[0110]) such that it amounts to no more than mere instructions to apply the exception using generic computing components. Furthermore, the claims recite the additional elements of a “trained regularized multivariable logistic regression machine learning model”. However, the functionality intended to be performed by the “trained regularized multivariable logistic regression machine learning model” appears to be based on very rudimentary constraints (e.g., input features associated with values for predictors). Without some prohibition in the claims regarding scalability, computation load, etc., this “trained regularized multivariable logistic regression machine learning model” could reasonably be considered an additional abstract idea in the “mental process” and/or “mathematical concepts” category, but for which is simply automated (i.e., “apply it”).
Accordingly, the additional elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. More specifically, the additional elements fail to include (1) improvements to the functioning of a computer or to any other technology or technical field (see MPEP 2106.05(a)), (2) applying or using a judicial exception to effect a particular treatment or prophylaxis for a disease or medical condition (see Vanda memo), (3) applying the judicial exception with, or by use of, a particular machine (see MPEP 2106.05(b)), (4) effecting a transformation or reduction of a particular article to a different state or thing (see MPEP 2106.05(c)), or (5) applying or using the judicial exception in some other meaningful way beyond generally linking the use of the judicial exception to a particular technological environment, such that the claim as a whole is more than a drafting effort designed to monopolize the exception (see MPEP 2106.05(e) and Vanda memo).
Rather, the limitations merely add the words “apply it” (or an equivalent) with the judicial exception, or 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)) or generally link the use of the judicial exception to a particular technological environment or field of use (see MPEP 2106.05(h)), particularly as it relates to the recited “software platform” comprising a “software application program” and at least one “database”, at least one “computer hardware processor”, and “trained regularized multivariable logistic regression machine learning model” elements. The claims are therefore still directed to an abstract idea.
The claims do 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 additional elements of using a software platform” comprising a “software application program” and at least one “database”, at least one “computer hardware processor”, and “trained regularized multivariable logistic regression machine learning model” to perform the steps amounts to no more than mere instructions to apply the exception using generic computer components. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept.
Claims 4,5, 7-9, 22, 24, 26, 28, and 35 are dependent on claim 1 and include all the limitations of claim 1. Therefore, they are also directed to the same abstract idea. The remaining dependent claims have not been found to integrate the judicial exception into a practical application, or provide significantly more than the abstract idea since they merely further narrow the abstract idea. Therefore, the dependent claims are found to be directed to an abstract idea without significantly more.
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.
Claim(s) 1, 4, 5, 7-9, 22, 24, 26, 28, and 35 is/are rejected under 35 U.S.C. 103 as being unpatentable over Donaldson (US 2021/0249141)1 in view of “Evaluation of Machine-Learning Algorithms for Predicting Opioid Overdose Risk Among Medicare Beneficiaries With Opioid Prescriptions” by Wei-Hsuan Lo-Ciganic, available March 22, 2019, hereinafter referred to as Lo-Ciganic.
Regarding claim 1, Donaldson discloses A method for predicting risk of incident opioid use disorder (OUD) and/or of an opioid overdose episode for a patient using a software platform for monitoring and treatment of subjects, the software platform comprising a software application program and at least one database communicatively coupled to the software application program, the method comprising:
executing the software application program, using at least one computer hardware processor to perform:
accessing data from the at least one database (see Para. [0030] –The machine readable storage medium 207 may include instructions, when executed by the processor 205, to receive a plurality of physical and/or clinical data from a plurality of EMR associated to a plurality of test subjects and subjects not diagnosed with a substance use disorder from a user device 221 or patient record database 224.”, and Para. [0010] – “In an embodiment, the clinical data includes age, sex, race, and ethnicity. In an embodiment, the clinical data includes demographic data, socioeconomic data, and any data that about a subject that can be obtained by observation or oral or written communication.” Also see Provisional, Table 3 – “Clinical data includes age, sex, nicotine and/or alcohol abuse or dependence, cannabinoid abuse or dependence, and no illicit drug use.”
However, Donaldson does not explicitly disclose the data comprising:
a longitudinal opioid-benzodiazepine dosage pattern over time of the patient
values for at least 50 predictors shown in Table 1 associated with the patient; and
values for one or more social determinants of health associated with the patient
(see Lo Ciganic pages 3-4 – “The predictor candidates also included a series of variables related to prescription opioid and relevant medication use: (1) total and mean daily morphine milligram equivalent (MME),17 (2) cumulative and continuous duration of opioid use (ie, no gap >32 days between fills),45 (3) total number of opioid prescriptions overall and by active ingredient, (4) type of opioid based on the US Drug Enforcement Administration’s Controlled Substance Schedule (I-IV) and duration of action, (5) number of opioid prescribers, (6) number of pharmacies providing opioid prescriptions,11,17,23 (7) number of early opioid prescription refills (refilling opioid prescriptions >3 days before the previous prescription runs out),46 (8) cumulative days of early opioid prescription refills, (9) cumulative days of concurrent benzodiazepines and/or muscle relaxant use, (10) number and duration of other relevant prescriptions (eg, gabapentinoids), and (11) receipt of methadone hydrochloride or buprenorphine hydrochloride for opioid use disorder.”, and Lo-Ciganic, page 3 – “We compiled 268 predictor candidates, informed by the literature (eTable 4 in the Supplement)2.” It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify Donaldson to utilize the particular data of Lo-Ciganic since both Donaldson and Lo-Ciganic are within the same field of endeavor (i.e. determining risk of opium use disorder), and all of the claimed elements were known in the prior art and one skilled in the art could have combined the elements as claimed by known methods with no change in their respective functions and the combination would have yielded predictable results to one of ordinary skill in the art at the time of the invention.
Donaldson also discloses generating machine learning model input features using the data accessed form the at least one database; (see Para. [0031] – “The data (as in, the training data 102) used to train the one or more predictive machine learning models (e.g., a stacked model 210) may include a substance use disorder profile. The substance use disorder profile may include one or more of a plurality of physical or clinical data from a plurality of EMR associated to a plurality of test subjects and subjects not diagnosed with a substance use disorder, a plurality of SNP profiles associated to the plurality of test subjects and subjects not diagnosed with a substance use disorder, a plurality of medical diagnoses regarding substance use disorders associated to the plurality of test subjects and subjects not diagnosed with a substance use disorder, and/or a plurality of treatment regimen recommendations.” Also see Provisional, Para. [0005] – “The method further includes the step of generating an ensemble machine learning model responsive to the two or more initial machine learning models that meet the predetermined sensitivity, the predetermined specificity, and the predetermined accuracy; and supplying a plurality of physical characteristics of a patient and a corresponding SNP expression profile of the patient to the ensemble machine learning model as inputs to predict a risk of a substance use disorder.”)
Donaldson partially discloses processing the machine learning model input features with a trained regularized multivariable logistic regression machine learning model trained based on diagnosis data indicating diagnosis of OUD and/or occurrence of opioid overdose for a plurality of subjects to obtain output indicative of the risk of OUD and/or the opioid overdose episode for the patient (see Para. [0031] – “The data (as in, the training data 102) used to train the one or more predictive machine learning models (e.g., a stacked model 210) may include a substance use disorder profile. The substance use disorder profile may include one or more of a plurality of physical or clinical data from a plurality of EMR associated to a plurality of test subjects and subjects not diagnosed with a substance use disorder, a plurality of SNP profiles associated to the plurality of test subjects and subjects not diagnosed with a substance use disorder, a plurality of medical diagnoses regarding substance use disorders associated to the plurality of test subjects and subjects not diagnosed with a substance use disorder, and/or a plurality of treatment regimen recommendations.” and Para. [0032] – “The machine-readable storage medium 207 may include instructions to assign a plurality of weight values to the training data 102 within a plurality of substance use disorder profiles, such as to a plurality of associated SNP profiles of the training data, determine the one or more weighted data inputs of the training data, thereby to form an initial set of decision making rules. The machine-readable storage medium 207 may include instructions to compare the initial set of decision-making rules to the one or more weighted data inputs of the training data and re-weight the initial set of decision-making rules based on the comparison to create a re-weighted set of decision-making rules. The re-weighted set of decision-making rules and the one or more weighted data inputs of the training data may be used to update the initial predictive model to generate an ensemble predictive model for determining risk of substance use disorder.” Also see Provisional, Para. [0005], and Para. [0016] – “The processor may further be configured to assign a plurality of weight values to the plurality of physical characteristics and associated SNP expression profiles of the training data, determine the one or more weighted data inputs of the training data, thereby to form an initial set of decision making rules.” However, Donaldson does not fully disclose a “regularized multivariable logistic regression machine learning model”. Donaldson discloses the training of models using regularization techniques. See Donaldson, Para. [0065] – “This case study utilized a larger data set and advanced machine learning to build a model that more accurately predicts OUD risk based on genotype alone (i.e., expression of alleles associated with reward, self-control, and affect) or in combination with clinical data (i.e., sex, age, and other substance abuse or dependence). Genetic data was modeled with and without clinical data using a machine learning platform that utilized a random forest model 105, a gradient boosted tree model 107, an elastic net model, a SVM model 109, or some combination thereof.” Also see Provisional Para. [0026] – “This case control study utilized a larger data set and advanced machine learning to build a model that more accurately predicts OUD risk based on genotype alone (i.e., expression of alleles associated with reward, self-control, and affect) or in combination with clinical characteristics (i.e., sex, age, and other substance abuse or dependence). Genetic data were modeled with and without clinical data using a machine learning platform that utilized random forest, gradient boosted trees, and elastic net classifiers.”) However, Donaldson does not explicitly disclose multivariable logistic regression. See Lo-Ciganic, page 4, “In both the primary and sensitivity analyses (eFigure 2 in the Supplement), we developed and tested prediction algorithms for opioid overdose using 5 commonly used machine-learning approaches: multivariate logistic regression, least absolute shrinkage and selection operator–type regression (LASSO), random forest (RF), gradient boosting machine (GBM), and deep neural network (DNN).” It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify Donaldson to utilize the specific model of Lo-Ciganic since both Donaldson and Lo-Ciganic are within the same field of endeavor (i.e. determining risk of opium use disorder based on a model), and all of the claimed elements were known in the prior art and one skilled in the art could have combined the elements as claimed by known methods with no change in their respective functions and the combination would have yielded predictable results to one of ordinary skill in the art at the time of the invention.)
Donaldson also discloses generating a graphical user interface (GUI) comprising a plurality of GUI elements including one or more selectable GUI elements, the plurality of GUI elements including:
a first GUI element displaying a risk categorization of the patient determined using the output, generated using the trained machine learning model, indicative of the risk of OUD and/or of the opioid overdose for the patient;
a second GUI element displaying one or more recommended actions determined using the risk categorization of the patient; and
a third GUI element allowing a user of the software application program to select at least one recommended action; and
displaying the GUI to the user
See Para. [0049] – “In response to a determination that the score indicator is greater than a threshold, at step 320, a score, prediction, and/or indicator indicating a potential or likelihood for development of a substance use disorder by a patient may be returned or transmitted to the user device 220. Additionally, the method 300 may include identifying one or more associated recommendations related to the determined score, prediction, and/or indicator, and automatically transmitting (along with the score or prediction) the determined initial risk and the one or more associated treatment regimen recommendations (e.g., a recommendation to utilize an alternate drug or prescription, such as a non-opioid, or a reduced dose opioid drug-based treatment regimen) to the user device 220 or display 216 of a health care professional via the electronic communication network in real time or in a timeframe of hours to days. Thus, the healthcare professional may have relevant information on treatment regimens to discuss with the patient, and the treatment regimens would be based on a potential for or risk of developing a substance use disorder and/or the treatment regimen recommendation.” The Examiner asserts that by presenting the recommendation to a user, the user is allowed to select the recommendation (ex. system recommends a doctor prescribe a pill, and the doctor chooses to prescribe the pill). The Examiner further asserts that that claimed number of “GUI elements” merely recites a design choice. The sole difference between the invention and the prior art is that the invention features an element displaying one or more recommended actions, and an element allowing selection of a recommended action. This is akin to a mere duplication of parts. See MPEP §2144 VI B, and In re Harza, 274 F.2d 669, 671; 124 USPQ 378, 380 (CCPA 1960), where it is stated that "the mere duplication of parts has no patentable significance unless a new and unexpected result is produced…". As no unexpected result is produced by providing two separate elements for the display and selection of recommended actions, there is no patentable significance in doing so. Thus, it would have been obvious to modify Donaldson to feature the number of GUI elements recited by the instant claims since all of the claimed elements were known in the prior art and one skilled in the art could have combined the elements as claimed by known methods with no change in their respective functions and the combination would have yielded predictable results of selecting a displayed recommended action.
Regarding claim 4, Donaldson and Lo-Ciganic disclose the method of claim 3, wherein the trained regularized multivariable logistic regression machine learning model is trained using Elastic Net regularization. (See Donaldson, Para. [0065] and Provisional Para. [0026])
Regarding claim 5, Donaldson discloses The method of claim 1, further comprising training a machine learning model using training data and a supervised learning to technique to obtain the regularized multivariable logistic regression machine learning model, wherein the training data comprises paired data comprising input-output pairs, each input-output pair having input values for the at least 50 predictors and a corresponding output value indicative of a risk of OUD and/or the opioid overdose episode, wherein the corresponding output value indicative of the risk of OUD is set based on an indication of OUD diagnosis, and/or initiation of methadone or buprenorphine. (See Lo-Ciganic, Page 4 – “In both the primary and sensitivity analyses (eFigure 2 in the Supplement), we developed and tested prediction algorithms for opioid overdose using 5 commonly used machine-learning approaches: multivariate logistic regression, least absolute shrinkage and selection operator–type regression (LASSO), random forest (RF), gradient boosting machine (GBM), and deep neural network (DNN). Previous studies consistently showed that these methods yield the best prediction results57,58; the eAppendix in the Supplement describes the details for each approach used. Given that beneficiaries may have multiple opioid overdose episodes, we present the results from a patient-level random subset (ie, using one 3-month period with predictor candidates measured to predict risk in the subsequent 3 months for each patient) from the validation data for ease of interpretation.” It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify Donaldson to utilize the features of Lo-Ciganic since both Donaldson and Lo-Ciganic are within the same field of endeavor (i.e. determining risk of opium use disorder using a model), and all of the claimed elements were known in the prior art and one skilled in the art could have combined the elements as claimed by known methods with no change in their respective functions and the combination would have yielded predictable results to one of ordinary skill in the art at the time of the invention.)
Regarding claim 7, Donaldson does not explicitly disclose the method of claim 5, wherein a corresponding output value indicative of the risk of the opioid overdose episode is set based on an indication of an opioid overdose episode diagnosis (see Lo-Ciganic page 3 – “We identified any occurrence of fatal or nonfatal opioid overdose (prescription opioids or other opioids, including heroin), defined in each 3-month window after the index prescription using the International Classification of Diseases, Ninth Revision, and International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10), codes for overdose (eTable 2 in the Supplement) from inpatient or emergency department settings.14,41-44 Overdose was defined with either an opioid overdose code as the primary diagnosis (80% of identified overdose episodes) or other drug overdose or substance use disorder code as the primary diagnosis (eTable 3 in the Supplement) and opioid overdose as the nonprimary diagnosis (20% of identified overdose episodes), as defined previously.14 Sensitivity analyses using opioid overdose as the primary diagnosis and capturing any opioid overdose diagnosis code in any position yielded similar results.” It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify Donaldson to utilize the teachings of Lo-Ciganic since both Donaldson and Lo-Ciganic are within the same field of endeavor (i.e. determining risk of opium use disorder), and all of the claimed elements were known in the prior art and one skilled in the art could have combined the elements as claimed by known methods with no change in their respective functions and the combination would have yielded predictable results to one of ordinary skill in the art at the time of the invention.)
Regarding claim 8, Donaldson discloses The method of claim 1, wherein the regularized multivariable logistic regression machine learning model comprises a deep neural network model, a random forest model, and/or a gradient boosting machine model. (see at least Para. [0033] – “Models and methods may include decision trees, random forest models, random forests utilizing bagging or boosting (as in, gradient boosting), neural network methods, support vector machines (SVM), other supervised learning models, other semi-supervised learning models, and/or other unsupervised learning models, as will be readily understood by one having ordinary skill in the art.” However, the explicit use of the particular models is not disclosed in the associated Provisional. The limitation is taught by Lo-Ciganic, page 4, “In both the primary and sensitivity analyses (eFigure 2 in the Supplement), we developed and tested prediction algorithms for opioid overdose using 5 commonly used machine-learning approaches: multivariate logistic regression, least absolute shrinkage and selection operator–type regression (LASSO), random forest (RF), gradient boosting machine (GBM), and deep neural network (DNN).” It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify Donaldson to include the specific models of Lo-Ciganic since both Donaldson and Lo-Ciganic are within the same field of endeavor (i.e. determining risk of opium use disorder using models), and all of the claimed elements were known in the prior art and one skilled in the art could have combined the elements as claimed by known methods with no change in their respective functions and the combination would have yielded predictable results to one of ordinary skill in the art at the time of the invention.)
Regarding claim 9, Donaldson does not explicitly disclose the method of claim 1, wherein the output from the regularized multivariable logistic regression machine learning model indicates the risk of OUD and/or the opioid overdose episode for the patient within 90 days of the patient receiving an opioid prescription. (See Lo-Ciganic page 3 – “In the primary analysis, we used the variables measured in each 3-month period (eg, the first) to predict overdose risk in each subsequent 3-month period (eg, the second) (eFigure 2A in the Supplement). In sensitivity analyses, instead of using a previous 3-month period to predict overdose in the next period, we included information collected in all of the historical 3-month windows to predict opioid risk for each 3-month period for each person (eFigure 2B in the Supplement).”, and page 10 – “To our knowledge, this study is the first to predict overdose risk in the subsequent 3-month period after initiation of treatment with prescription opioids as opposed to 1-year or longer period.” It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify Donaldson to predict within a specific timeframe as taught by Lo-Ciganic since it would allow for quicker intervention.
Regarding claim 22, Donaldson partially discloses The method of claim 1, wherein the output of the regularized multivariable logistic regression machine learning model is indicative of the risk of the opioid overdose episode for the patient, and wherein the data further comprises values for a predictor indicating whether the patient has a previous history of OUD and/or an opioid overdose episode. Donaldson discloses output indicative of the risk of the opioid overdose episode (see Para. [0021] – “The method further includes the step of supplying a SNP profile of a patient to the ensemble machine learning model as an input to obtain a score indicative of the patient's risk of a substance use disorder.”) However, Donaldson does not explicitly disclose the data comprises values for a predictor indicating whether the patient has a previous history of OUD and/or an opioid overdose episode. See Lo-Ciganic page 3 – “We compiled 268 predictor candidates, informed by the literature (eTable 4 in the Supplement).”3 It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify Donaldson to include data indicating whether the patient has a previous history of prescription opioid overdose since it may be used to further refine the prediction that the patient is susceptible to overdosing again.
Regarding claim 24, Donaldson does not explicitly disclose The method of claim 1 wherein the one or more recommended actions comprise selecting the patient for enrollment in a lock-in program, making an outreach call to the patient, referring the patient to a use disorder specialist, prescribing an opioid antagonist therapy, administering an opioid antagonist therapy to the patient, and/or initiating an evidence-based intervention. (see Lo-Ciganic page 3 – “In response, health systems, payers, and policymakers have developed programs to identify and intervene in individuals at high risk of problematic opioid use and overdose. These programs, whether outreach calls from case managers, prior authorizations, referrals to substance use disorder specialists, dispensing of naloxone hydrochloride, or enrollment in lock-in programs, can be expensive to payers and burdensome to patients.” It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify Donaldson to include the specific interventions of Lo-Ciganic since both Donaldson and Lo-Ciganic are within the same field of endeavor (i.e. providing interventions to at-risk subjects), and all of the claimed elements were known in the prior art and one skilled in the art could have combined the elements as claimed by known methods with no change in their respective functions and the combination would have yielded predictable results to one of ordinary skill in the art at the time of the invention.)
Regarding claim 26, Donaldson discloses The method of claim 24, wherein initiating the evidence-based intervention comprises initiating use of medication used to treat OUD, wherein the medication comprises buprenorphine and/or naltrexone. (see Para. [0079] – “For individuals prospectively identified as genetically at-risk for OUD, a treatment regimen may avoid or reduce opioid exposure, even in the context of perioperative pain management. The treatment regimen for such an individual would be involve non-opioid modalities, such as regional anesthesia and analgesic techniques, and multimodal analgesia. Additionally, knowledge regarding the genetic predisposition for OUD informs treatment paradigms for at-risk individuals who have been exposed to opioids, as such individuals may be candidates for buprenorphine-assisted dose reduction or behavioral interventions.”)
Regarding claim 28, Donaldson does not explicitly disclose The method of claim 24, wherein the opioid antagonist therapy comprises naloxone. (see Lo Ciganic page 3 – “In response, health systems, payers, and policymakers have developed programs to identify and intervene in individuals at high risk of problematic opioid use and overdose. These programs, whether outreach calls from case managers, prior authorizations, referrals to substance use disorder specialists, dispensing of naloxone hydrochloride, or enrollment in lock-in programs, can be expensive to payers and burdensome to patients.) It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify Donaldson to include the specific intervention of Lo-Ciganic since both Donaldson and Lo-Ciganic are within the same field of endeavor (i.e. providing interventions to at-risk subjects), and all of the claimed elements were known in the prior art and one skilled in the art could have combined the elements as claimed by known methods with no change in their respective functions and the combination would have yielded predictable results to one of ordinary skill in the art at the time of the invention.)
Regarding claim 35, Donaldson does not explicitly disclose the method of claim 1, wherein the longitudinal opioid-benzodiazepine dosage pattern over time of the patient indicates one of a plurality of predetermined opioid-BZD trajectories. (see Lo Ciganic pages 3-4 – “The predictor candidates also included a series of variables related to prescription opioid and relevant medication use: (1) total and mean daily morphine milligram equivalent (MME),17 (2) cumulative and continuous duration of opioid use (ie, no gap >32 days between fills),45 (3) total number of opioid prescriptions overall and by active ingredient, (4) type of opioid based on the US Drug Enforcement Administration’s Controlled Substance Schedule (I-IV) and duration of action, (5) number of opioid prescribers, (6) number of pharmacies providing opioid prescriptions,11,17,23 (7) number of early opioid prescription refills (refilling opioid prescriptions >3 days before the previous prescription runs out),46 (8) cumulative days of early opioid prescription refills, (9) cumulative days of concurrent benzodiazepines and/or muscle relaxant use, (10) number and duration of other relevant prescriptions (eg, gabapentinoids), and (11) receipt of methadone hydrochloride or buprenorphine hydrochloride for opioid use disorder.” It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to modify Donaldson to utilize the particular data of Lo-Ciganic since both Donaldson and Lo-Ciganic are within the same field of endeavor (i.e. determining risk of opium use disorder), and all of the claimed elements were known in the prior art and one skilled in the art could have combined the elements as claimed by known methods with no change in their respective functions and the combination would have yielded predictable results to one of ordinary skill in the art at the time of the invention.)
Response to Arguments
Applicant's arguments regarding claims rejected under 35 U.S.C. 101 have been fully considered but they are not persuasive. Applicant argues with substance:
Applicant argues that the claims “does not set forth or describe any mathematical relationships, calculations, formulas, or equations using words or mathematical symbols”. The Examiner respectfully disagrees. The claims recite a “trained regularized multivariable logistic regression machine learning model”. Therefore the claims recite the performance of mathematical calculations.
Applicant argues that the claims do not fall within “Certain Methods of Organizing Human Activity” due to allegedly being dissimilar to the examples presented in the MPEP. This is not persuasive as the groupings of abstract ideas are not limited to the examples presented in the MPEP.
Applicant makes a conclusory statement that the claims recite multiple limitations that do not encompass mental steps (page 13 of response). Conclusory statements are not persuasive. . The Examiner further notes that the limitation “processing the machine learning model input features with a trained regularized multivariable logistic regression machine learning model” was not found to fall under “Mental Processes”, but was instead found to fall under “Mathematical Concepts”.
Applicant argues that the claims solve the problem of predicting risk of incident opioid use disorder by the use of a particular machine learning model. Even if so, predicting risk of incident opioid use disorder is not a technological problem.
Applicant argues that the claims do not perform an existing process. This is not persuasive. Novelty is not a consideration as to whether or not the claims are directed to an abstract idea.
Applicant argues that the claims are “narrowed to the particular application of prediction of a particular range of disorders in a particular way within a particular architecture”. This is not persuasive. Even if “narrow”, the claims are still directed to an abstract idea. "[D]efining the excluded categories, the Court has ruled that the exclusion applies if a claim involves a natural law or phenomenon or abstract idea, even if the particular natural law or phenomenon or abstract idea at issue is narrow. Mayo, 132 S. Ct. at 1303" buySAFE, Inc. v. Google, Inc., 112 USPQ2d 1093 (Fed. Cir. 2014).
Applicant argues that the claims present a technological solution to a technological problem. This is not persuasive. As stated above, predicting risk of incident opioid use disorder is not a technological problem. Further the claims introduce no improvement to the functioning to any of the involved computing elements. The claims merely apply machine learning to opioid use disorder prediction. No improvements have been made in the realm of software applications, nor are any improvements made in the realm of machine learning.
Based on at least the above, the 101 rejection is maintained.
Applicant's arguments regarding claims rejected under 35 U.S.C. 103 have been fully considered but they are not persuasive. Applicant argues with substance:
Applicant argues that the prior art does not disclose a longitudinal opioid-benzodiazepine dosage pattern over time of the patient. Applicant states that “the claimed dosage pattern is therefore a time-series of multiple values, not a single number such as the cumulative days of concurrent BZD and opioid use, as described in Lo-Ciganic”. The Examiner respectfully disagrees. The category under which “cumulative overlapping days of concurrent opioid and BZD use” is titled “Patterns of non-opioid prescription” . Based on this, as well as the 112(b) rejection above, it is apparent that Lo-Ciganic discloses longitudinal opioid-benzodiazepine dosage pattern over time of the patient.
Based on at least the above, the 103 rejection is maintained.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to KYLE G ROBINSON whose telephone number is (571)272-9261. The examiner can normally be reached Monday - Thursday, 7:00 - 4:30 EST; Friday 7:00-11:00 EST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Kambiz Abdi can be reached on 571-272-6702. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/KYLE G ROBINSON/Examiner, Art Unit 3685
/KAMBIZ ABDI/Supervisory Patent Examiner, Art Unit 3685
1 Also see Provisional 62/892,212 filed on August 27, 2019 which is incorporated by reference
2 eTable4 appears to be identical to Table 1. Also, the elements of “Receipt of low-income subsidy” and “Anxiety disorders” equate to “economic stability” and “community context”, respectively. Further note that language “Patterns of prescription opioid use” and “Patterns of non-opioid prescription use”
3 eTable 4 features a Health status factor of “History of prescription opioid overdose”