Detailed Notice
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 .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 08/28/2025 has been entered.
Status of Claims
Claims 1-23 are currently pending.
Claims 1, 3, 13, 15, and 20-23 are amended.
Claims 1-23 are rejected.
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 1-23 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception (i.e., a law of nature, a natural phenomenon, or an abstract idea) without significantly more.
Step 1:
In the instant case, claims 1-12 and 21-23 are directed toward a non-transitory computer readable medium (i.e., manufacture), claims 13-19 are directed toward a method (i.e., process), and claims 20 is directed toward a system (i.e., machine). Thus, each of the claims falls within one of the four statutory categories. Nevertheless, the claims fall within the judicial exception of an abstract idea.
Step 2A—Prong 1:
Independent claims 1, 13, and 20 recites steps that, under their broadest reasonable interpretations, cover performance of the limitations of a mental process and/or certain method of organizing human activity but for the recitation of generic computer components.
Claim 1 recites: “A non-transitory computer readable medium comprising instructions which, when executed by one or more hardware processors, causes performance of operations comprising: training a first machine learning model to predict characteristics of healthcare reimbursements for healthcare services for patients, the training comprising: obtaining a first training data set of a plurality of training data sets, the first training data set including: a first set of healthcare attributes corresponding to a first healthcare service for a first patient; and a first reimbursement amount approved for the first healthcare service; obtaining a second training data set of a plurality of training data sets, the second training data set including: a second set of healthcare attributes corresponding to a second healthcare service for a second patient; and a second reimbursement amount approved for the first healthcare service; training the first machine learning model using the plurality of training data sets to generate a trained machine learning model; obtaining, at a first point in time prior to generating a medical claim corresponding to a target healthcare service for a particular patient, a first target set of healthcare attributes corresponding to a condition of the target healthcare service for the particular patient; and applying the trained machine learning model to the first target set of healthcare attributes to estimate characteristics of a predicted reimbursement for the target healthcare service corresponding to the condition of the particular patient; applying the characteristics of the predicted reimbursement for the target healthcare service to a second machine learning model to generate a set of key features corresponding to healthcare attributes input to the trained machine learning model to estimate the characteristics of the predicted reimbursement for the target healthcare service, wherein the second machine learning model is trained to predict key features based on an impact of the set of key features on predicting, by the trained machine learning model, the characteristics of the healthcare reimbursements for healthcare services for patients; and based on the characteristics of the predicted reimbursement predicted by the first machine learning model and the set of key features predicted by the second machine learning model:[[,]] identifying a first treatment to improve the predicted reimbursement; and generating, in real-time, a first recommendation to a healthcare provider to perform the first treatment, wherein including the first treatment in a treatment plan administered for the particular patient by the healthcare provider improves at least one of (a) an amount of the predicted reimbursement, and (b) a time for payment of the predicted reimbursement for one or more treatments in the treatment plan”.
The limitations of obtaining a first training data set of a plurality of training data sets, the first training data set including: a first set of healthcare attributes corresponding to a first healthcare service for a first patient; and a first reimbursement amount approved for the first healthcare service; obtaining a second training data set of a plurality of training data sets, the second training data set including: a second set of healthcare attributes corresponding to a second healthcare service for a second patient; and a second reimbursement amount approved for the first healthcare service; training the first… model using the plurality of training data sets to generate a trained… model; obtaining, at a first point in time prior to generating a medical claim corresponding to a target healthcare service for a particular patient, a first target set of healthcare attributes corresponding to a condition of the target healthcare service for the particular patient; and applying the trained… model to the first target set of healthcare attributes to estimate characteristics of a predicted reimbursement for the target healthcare service corresponding to the condition of the particular patient; applying the characteristics of the predicted reimbursement for the target healthcare service to a second… model to generate a set of key features corresponding to healthcare attributes input to the trained… model to estimate the characteristics of the predicted reimbursement for the target healthcare service, wherein the second… model is trained to predict key features based on an impact of the set of key features on predicting, the characteristics of the healthcare reimbursements for healthcare services for patients; and based on the characteristics of the predicted reimbursement predicted and the set of key features predicted: identifying a first treatment to improve the predicted reimbursement; and generating, in real-time, a first recommendation to a healthcare provider to perform the first treatment, wherein including the first treatment in a treatment plan administered for the particular patient by the healthcare provider improves at least one of (a) an amount of the predicted reimbursement, and (b) a time for payment of the predicted reimbursement for one or more treatments in the treatment plan, given the broadest reasonable interpretation, cover the abstract idea of a certain method of organizing human activity because they recite managing personal behavior or relationships or interactions between people, (i.e. social activities, teaching, and following rules or instructions—in this case the aforementioned steps recite a process of training, obtaining, generating, applying, estimate, predict, identifying, and administered, which is properly interpreted as a “personal behavior”), fundamental economic principles or practices and commercial or legal interactions (i.e., reimbursements) (see MPEP 2106.04(a)(2)(II)), and/or a mental process that a doctor/insurer/person should determine when predicting characteristics of healthcare reimbursements for healthcare services, but instead automates the process via a computer model, e.g. see MPEP 2106.04(a)(2). Any limitations not identified above as part of the abstract idea are deemed “additional elements”, and will be discussed in further detail below.
Further, the abstract idea of claims 13 and 20 are identical as the abstract idea of claim 1. This limitation, given the broadest reasonable interpretation, also falls under the abstract idea of a certain method of organizing human activity because it recites managing personal behavior or relationships or interactions between people. Additionally, the limitations may also fall under the abstract idea of a mental process because doctor/insurer/person should determine when predicting characteristics of healthcare reimbursements for healthcare services.
Dependent claims 2-12, 14-19, and 21-23 include other limitations, as well as specific step of data to be processed, received, and applied, but these only serve to further limit the abstract idea and do not add and additional elements, and hence are nonetheless directed towards fundamentally the same abstract idea as independent claims 1, 13, and 20. However, recitation of an abstract idea is not the end of the 35 U.S.C. 101 analysis. Each of the claims must be analyzed for additional elements that indicate the abstract idea is integrated into a practical application to determine whether the claim is considered to be “directed to” an abstract idea.
Step 2A—Prong 2:
Claims 1-23 are not integrated into a practical application because the additional elements (i.e. any limitations that are not identified as part of the abstract idea) amount to no more than limitations which:
Amount to mere instructions to apply an exception—for example, the recitation of “non-transitory computer readable medium”, “hardware processors”, “memory”, “system”, and “machine learning model”, which amount to merely invoking a computer as a tool to perform the abstract idea, e.g. see FIG. 1, FIG. 4, and [0021]-[0025], of the present specification, and see further MPEP 2106.05(f);
Generally linking the abstract idea to a particular technological environment or field of use, for example, “training the first machine learning model using the plurality of training data sets to generate a trained machine learning model”, “ applying the trained machine learning model to”, “to a second machine learning model to”, “input to the trained machine learning model to”, “wherein the second machine learning model is trained to”, “by the trained machine learning model”, “by the first machine learning model”, and “by the second machine learning model”, which amounts to limiting the abstract idea to the field of technology/the environment of computers, see MPEP 2106.05(h); and/or
Merely acquiring information for further analysis by the system and the particular manner of acquisition is not described or shown to be important, for example, “obtaining a first training data set of a plurality of training data sets”, “obtaining a second training data set of a plurality of training data sets”, and “obtaining, at a first point in time prior to generating a medical claim corresponding to a target healthcare service for a particular patient, a first target set of healthcare attributes corresponding to a condition of the target healthcare service for the particular patient”, which amounts to insignificant extra-solution activity in the form of mere data gathering because it merely functions tangentially to the main idea of the invention and serves only to bring in the data necessary for the inventions main analysis, see MPEP 2106.05(g).
Additionally, dependent claims 2-12, 14-19, and 21-23 include other limitations, but as stated above, the limitations recited by these claims do not include any additional elements beyond those already recited in independent claims 1, 13, and 20, and hence also do not integrate the aforementioned abstract idea into a practical application.
Step 2B:
The claims do not include additional elements that are sufficient to amount to “significantly more” than the judicial exception because the additional elements (i.e. the elements other than the abstract idea), as stated above, are directed towards no more than limitations that amount to mere instructions to apply the exception, and/or generally link the abstract idea to a particular technological environment or field of use, which even when reevaluated under the considerations of Step 2B of the analysis, do not amount to “significantly more” than the abstract idea.
Dependent claims 2-12, 14-19, and 21-23 include other limitations, but none of these limitations are deemed significantly more than the abstract idea because, as stated above, the aforementioned dependent claims do not recite any additional elements not already recited in independent claims 1, 13, and 20, and hence do not amount to “significantly more” than the abstract idea.
Additionally, the additional elements (i.e., “obtaining a first training data set of a plurality of training data sets”, “obtaining a second training data set of a plurality of training data sets”, and “obtaining, at a first point in time prior to generating a medical claim corresponding to a target healthcare service for a particular patient, a first target set of healthcare attributes corresponding to a condition of the target healthcare service for the particular patient”), add extra solution activity, which comprises limitations which amount to elements that have been recognized as well-understood, routine, and conventional activity in a particular field as demonstrated by:
Relevant court decisions (See MPEP 2106.05(d)(II)):
Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network); but see DDR Holdings, LLC v. Hotels.com, L.P., 773 F.3d 1245, 1258, 113 USPQ2d 1097, 1106 (Fed. Cir. 2014) (“Unlike the claims in Ultramercial, the claims at issue here specify how interactions with the Internet are manipulated to yield a desired result‐‐a result that overrides the routine and conventional sequence of events ordinarily triggered by the click of a hyperlink.” (emphasis added)).
Thus, taken alone, the additional elements do not amount to significantly more than the abstract idea identified above. Furthermore, looking at the limitations as an ordered combination adds nothing that is not already present when looking at the elements taken individually, and there is no indication that the combination of elements improves the functioning of a computer or improves any other technology, and their collective functions merely provide conventional computer implementation.
Therefore, whether taken individually or as an ordered combination, claims 1-23 are nonetheless rejected under 35 U.S.C. 101 as being directed to non-statutory subject matter.
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, 3-5, 12-13, 15-17, 20, and 22-23 are rejected under 35 U.S.C. 103 as being unpatentable over Bostic et al. (US 20210202099 A1), hereinafter Bostic, in view Zahora et al. (US 20220309592 A1), hereinafter Zahora.
Regarding claim 1 Bostic teaches a non-transitory computer readable medium comprising instructions which, when executed by one or more hardware processors, causes performance of operations comprising: training a first machine learning model to predict characteristics of healthcare reimbursements for healthcare services for patients ([0040]-[0042]), the training comprising: obtaining a first training data set of a plurality of training data sets ([0022], [0033], [0040], and [0045]), the first training data set including: a first set of healthcare attributes corresponding to a first healthcare service for a first patient ([0023], [0041], [0068], [0086], and [0089]); and a first reimbursement amount approved for the first healthcare service (Abstract, [0040], [0041], [0072]); obtaining a second training data set of a plurality of training data sets ([0022], [0033], [0040], and [0045]), the second training data set including: a second set of healthcare attributes corresponding to a second healthcare service for a second patient ([0023], [0041], [0068], [0086], and [0089]); and a second reimbursement amount approved for the first healthcare service (Abstract, [0040], [0041], [0072]); training the first machine learning model using the plurality of training data sets to generate a trained machine learning model ([0039]-[0040]); obtaining¸ at a first point in time prior to generating a medical claim corresponding to a target healthcare service for a particular patient ([0023]: “matching the simulated future health state to a predicted patient medical service need; matching the predicted patient medical service need to at least one of the patient's healthcare providers” and [0041]: “the method includes monitoring insurance billing events, that includes ingesting patient data received from one of a plurality of patient data providers and healthcare services data relating to the patient data; ingesting data relating to insurance reimbursement criteria and insurance reimbursement records relating to the healthcare services data; determining one or more relationships between the ingested patient data, healthcare services data, insurance reimbursement criteria and insurance reimbursement records, and data and previously ingested patient data, healthcare services data, insurance reimbursement criteria and insurance reimbursement records wherein at least one new enriched data set is created based on the determined one or more relationships”), a first target set of healthcare attributes corresponding to a condition of the target healthcare service for the particular patient ([0023], [0041], [0086], and [0089]); and applying the trained machine learning model to the first target set of healthcare attributes to estimate characteristics of a predicted reimbursement for the target healthcare service ([0037] and [0040]-[0041]) corresponding to the condition of the particular patient ([0115], [0181], [0185], and [0245]).
Bostic does not teach applying the characteristics of the predicted reimbursement for the target healthcare service to a second machine learning model to generate a set of key features corresponding to healthcare attributes input to the trained machine learning model to estimate the characteristics of the predicted reimbursement for the target healthcare service, wherein the second machine learning model is trained to predict key features based on an impact of the set of key features on predicting, by the trained machine learning model, the characteristics of the healthcare reimbursements for healthcare services for patients; and based on the characteristics of the predicted reimbursement predicted by the first machine learning model and the set of key features predicted by the second machine learning model; identifying a first treatment to improve the predicted reimbursement; and generating, in real-time, a first recommendation to a healthcare provider to perform the first treatment, wherein including the first treatment plan administered for the particular patient by the healthcare provider at least one of (a) an amount of the predicted reimbursement, and (b) a time for payment of the predicted reimbursement for one or more treatments in the treatment plan.
However, Zahora teaches applying the characteristics of the predicted reimbursement for the target healthcare service to a second machine learning model (Zahora, FIG. 1A, FIG. 1B, FIG. 8, [0057-[0058]]: “FIG. 1A is a process flow diagram of an example process for using a predictive analytics platform to perform machine learning analysis on a data universe of information resources to produce complex predictive intelligence… FIG. 1B is a flow chart of an example method for performing machine learning analysis on a data universe of information resources to calculate a payment estimation by applying historic payment patterns to a present medical claim”, [0065]: “FIG. 8 is a flow diagram of an example process for predicting future revenue through performing machine learning analysis of historic medical procedures to identify and extrapolate commonly paired and/or follow-on procedures to procedures captured in recent claims data”, and [0092]-[0093]: “The platform as described herein creates a predictive model based on a hybrid of data for a particular patient with data for payers associated with the particular patient to provide a prediction and estimation for a medical claim for the particular patient… The historic information, in some examples, may include services rendered, products sold, patient demographic information, patient insurance information, and remittance received from various payers to reimburse provision of medical services and products. The data universe may further include information regarding the various payers and the payer plans, such as copayment levels, deductible amounts, reimbursement eligibility requirements, and reimbursement amounts”) to generate a set of key features corresponding to healthcare attributes input to the trained machine learning model to estimate the characteristics of the predicted reimbursement for the target healthcare service (Zahora, [0016]: “The payment pattern application engine may be configured to estimate a supplemental reimbursement from the medical payment assistance program”, [0021]: “the medical claim includes a pre-service estimate for medical services and the payment pattern application engine is configured to calculate a payment estimation and confidence score for the pre-service estimate for the medical services. The medical services may include at least one of medical transport and a prescription”, [0032]: “The payment pattern application engine may be configured to estimate a supplemental reimbursement from the medical payment assistance program.”, and [0129]: “he patient matching engine 230 may access the patient data 240 to match features to insured persons in the data repository 210. The patient matching engine 230, in some embodiments, returns exact matches. The patient matching engine 230, in other embodiments, returns both exact matches and similar matches. Similar matches may be qualified with a closeness rating or percentage similarity to the patient. The patient matching engine 230, for example, may return a set of patient identifiers for referencing the de-identified patient data 240”), wherein the second machine learning model is trained to predict key features based on an impact of the set of key features on predicting, by the trained machine learning model, the characteristics of the healthcare reimbursements for healthcare services for patients (Zahora, [0021]: “The payment pattern application engine may be configured to calculate a payment estimation and confidence score for one or more candidate medical procedure codes. The payment pattern application engine may be configured to provide alternative or preferred medical procedure codes based on the calculated payment estimations… The payment pattern application engine may be configured to calculate payment estimations and confidence scores and provide a recommended patient pre-payment amount based on the calculated payment estimations and confidence scores” and [0269]: “In some implementations, the predictive analytics platform 102 includes a patient pre-payment calculation engine 1092 to determine the recommended pre-payment amount 1092 based on the balance due 1090 and the likelihood of patient payment 1088. For example, although the patient's likelihood to pay $1,000 is medium 1088 a with a confidence score of 98/100 1088 b, the patient may have a higher likelihood of covering a portion of the balance due 1090. Further, the likelihood of the patient repaying any portion of the balance due 1090 may drop significantly after leaving the facility (e.g., when being billed to a home address 1094). Thus, the patient pre-payment calculation engine 1092 may analyze remittance data 268 related to similar patients (e.g., identified by the patient matching engine 230) to identify an amount of funds up to the balance due 1090 that is associated with a high likelihood of payment. The patient pre-payment calculation engine 1092 may further identify the amount based on at least a threshold confidence score of the high likelihood that the patient will pay the amount. The patient pre-payment calculation engine 1092 may perform many of the same functions of the patient likelihood to pay analysis engine 228, but with a goal of a likelihood rating rather than a goal of a particular reimbursement amount”); and based on the characteristics of the predicted reimbursement predicted by the first machine learning model and the set of key features predicted by the second machine learning model (Zahora, [0016]: “The payment pattern application engine may be configured to estimate a supplemental reimbursement from the medical payment assistance program”, [0021]: “the medical claim includes a pre-service estimate for medical services and the payment pattern application engine is configured to calculate a payment estimation and confidence score for the pre-service estimate for the medical services. The medical services may include at least one of medical transport and a prescription”, [0032]: “The payment pattern application engine may be configured to estimate a supplemental reimbursement from the medical payment assistance program.”, and [0129]: “he patient matching engine 230 may access the patient data 240 to match features to insured persons in the data repository 210. The patient matching engine 230, in some embodiments, returns exact matches. The patient matching engine 230, in other embodiments, returns both exact matches and similar matches. Similar matches may be qualified with a closeness rating or percentage similarity to the patient. The patient matching engine 230, for example, may return a set of patient identifiers for referencing the de-identified patient data 240”, and [0269]: “In some implementations, the predictive analytics platform 102 includes a patient pre-payment calculation engine 1092 to determine the recommended pre-payment amount 1092 based on the balance due 1090 and the likelihood of patient payment 1088. For example, although the patient's likelihood to pay $1,000 is medium 1088 a with a confidence score of 98/100 1088 b, the patient may have a higher likelihood of covering a portion of the balance due 1090. Further, the likelihood of the patient repaying any portion of the balance due 1090 may drop significantly after leaving the facility (e.g., when being billed to a home address 1094). Thus, the patient pre-payment calculation engine 1092 may analyze remittance data 268 related to similar patients (e.g., identified by the patient matching engine 230) to identify an amount of funds up to the balance due 1090 that is associated with a high likelihood of payment. The patient pre-payment calculation engine 1092 may further identify the amount based on at least a threshold confidence score of the high likelihood that the patient will pay the amount. The patient pre-payment calculation engine 1092 may perform many of the same functions of the patient likelihood to pay analysis engine 228, but with a goal of a likelihood rating rather than a goal of a particular reimbursement amount”); identifying a first treatment to improve the predicted reimbursement (Zahora, FIG. 17A through 17C, [0083], [0092]: “The platform may utilize demographic and insurance coverage discovery to further improve the accuracy and timeliness of the predictions and estimations. “Discovery” refers to a process of finding and identifying information relevant to a medical claim that may be unknown and/or unavailable to a patient, payer, and/or provider prior to the discovery process or information of which the patient, payer, and/or provider was unaware of prior to the discovery process”, [0094]: “applying machine learning algorithms and data analytics to historic information, medical providers may be supplied with complex predictive intelligence, such as dependable revenue stream estimates and revenue forecasting. Thus, systems and methods described herein provide a technical solution to the technical problem of reliably predicting revenue streams from an abundance of payer sources each having different reimbursement levels and requirements for coverage eligibility. Further, the inventors recognized that in combining data sources and applying machine learning analysis, the resultant data analytics may support real-time or near-real-time decision making that had not previously been possible, such as rapid patient pre-approval for medical procedures and prescription products as well as complex dynamic scheduling of mobile medical services. Thus, systems and methods described herein additionally provide a technical solution to the technical problem of overcoming resource availability obstacles and reimbursement obstacles to automatically deploy resources to patients in need in a timely manner, thereby improving patient outcomes while improving the bottom line”, [0116]: “a type of medical service (e.g., a life-changing service such as cancer treatment, limb amputation, etc. versus a non-life-changing service such as tonsil removal) and/or a cost to patient range (e.g., hundreds, thousands, over ten thousand, etc.) may be considered in matching the first patient to the one or more second patients. The patient payment pattern, for example, may include one or more likelihoods (e.g., percentage likelihood, likelihood on a scale of one to ten), such as a first likelihood related to a patient most demographically similar but with a more dissimilar cost to patient range and a second likelihood related to a patient less demographically similar but within a same cost to patient range as the medical claim for the first patient” and [0280]: “the charting service 1220 may utilize the supplemented and/or corrected demographic information to obtain medical records for the victim that may enable improved care for the victim”); and generating, in real-time, a first recommendation to a healthcare provider to perform the first treatment (Zahora, FIG. 17A through 17C, [0083], and [0092]: “these systems and methods enable the patient to better plan their treatments with evaluations of care providers, treatment options, and associated costs”, [0094]: “applying machine learning algorithms and data analytics to historic information, medical providers may be supplied with complex predictive intelligence, such as dependable revenue stream estimates and revenue forecasting. Thus, systems and methods described herein provide a technical solution to the technical problem of reliably predicting revenue streams from an abundance of payer sources each having different reimbursement levels and requirements for coverage eligibility. Further, the inventors recognized that in combining data sources and applying machine learning analysis, the resultant data analytics may support real-time or near-real-time decision making that had not previously been possible, such as rapid patient pre-approval for medical procedures and prescription products as well as complex dynamic scheduling of mobile medical services. Thus, systems and methods described herein additionally provide a technical solution to the technical problem of overcoming resource availability obstacles and reimbursement obstacles to automatically deploy resources to patients in need in a timely manner, thereby improving patient outcomes while improving the bottom line”, [0116]: “a type of medical service (e.g., a life-changing service such as cancer treatment, limb amputation, etc. versus a non-life-changing service such as tonsil removal) and/or a cost to patient range (e.g., hundreds, thousands, over ten thousand, etc.) may be considered in matching the first patient to the one or more second patients. The patient payment pattern, for example, may include one or more likelihoods (e.g., percentage likelihood, likelihood on a scale of one to ten), such as a first likelihood related to a patient most demographically similar but with a more dissimilar cost to patient range and a second likelihood related to a patient less demographically similar but within a same cost to patient range as the medical claim for the first patient”, and [0278]: “The ePCR may include a detailed record of treatments and interventions provided to the patient by the caregivers along with patient transport information. The predictive analytics platform 102 may receive the ePCR from the charting service 1220 and may populate a medical claim with information recorded in the ePCR and/or utilize the medical procedure codes indicated by and/or recorded in the ePCR to estimate payments as described above. In an implementation, the medical device(s) 1296 may communicatively couple to the charting service 1220 and/or the mobile device 1280 to provide medical treatment information for automatic recordation in the ePCR”), wherein including the first treatment plan administered for the particular patient by the healthcare provider at least one of (a) an amount of the predicted reimbursement, (Zahora, FIG. 17A through 17C, [0013]: “calculating the payment estimation includes applying at least one deductible amount corresponding to a deductible level of an active payer plan of first patient. Calculating the payment estimation may include automatically determining a first deductible amount of the at least one deductible amount by contacting an external computing system of the payer via a network to request a current balance of a remaining maximum deductible”, [0021]: “the medical claim includes a pre-service estimate for medical services and the payment pattern application engine is configured to calculate a payment estimation and confidence score for the pre-service estimate for the medical service”, [0022]: “The payment pattern application engine may be configured to receive a patient medical condition from the CAD, calculate payment estimations based on the patient medical condition and the identified insurance coverages, and provide the payment estimations to one or more of the CAD and the ePCR. The payment pattern application engine may be configured to provide one or more of a recommended transport destination and medical procedure codes based on the payment estimations and identified insurance coverages. The payment pattern application engine may be configured to receive updated medical condition information from the ePCR application, update the payment estimations based on the updated medical condition information, and provide the updated payment estimations to the ePCR application”, [0116]: “a type of medical service (e.g., a life-changing service such as cancer treatment, limb amputation, etc. versus a non-life-changing service such as tonsil removal) and/or a cost to patient range (e.g., hundreds, thousands, over ten thousand, etc.) may be considered in matching the first patient to the one or more second patients. The patient payment pattern, for example, may include one or more likelihoods (e.g., percentage likelihood, likelihood on a scale of one to ten), such as a first likelihood related to a patient most demographically similar but with a more dissimilar cost to patient range and a second likelihood related to a patient less demographically similar but within a same cost to patient range as the medical claim for the first patient”, and [0278]: “The ePCR may include a detailed record of treatments and interventions provided to the patient by the caregivers along with patient transport information. The predictive analytics platform 102 may receive the ePCR from the charting service 1220 and may populate a medical claim with information recorded in the ePCR and/or utilize the medical procedure codes indicated by and/or recorded in the ePCR to estimate payments as described above. In an implementation, the medical device(s) 1296 may communicatively couple to the charting service 1220 and/or the mobile device 1280 to provide medical treatment information for automatic recordation in the ePCR”) and (b) a time for payment of the predicted reimbursement for one or more treatments in the treatment plan (Zahora, FIG. 17A through 17C, [0083], and [0094]: “applying machine learning algorithms and data analytics to historic information, medical providers may be supplied with complex predictive intelligence, such as dependable revenue stream estimates and revenue forecasting. Thus, systems and methods described herein provide a technical solution to the technical problem of reliably predicting revenue streams from an abundance of payer sources each having different reimbursement levels and requirements for coverage eligibility. Further, the inventors recognized that in combining data sources and applying machine learning analysis, the resultant data analytics may support real-time or near-real-time decision making that had not previously been possible, such as rapid patient pre-approval for medical procedures and prescription products as well as complex dynamic scheduling of mobile medical services. Thus, systems and methods described herein additionally provide a technical solution to the technical problem of overcoming resource availability obstacles and reimbursement obstacles to automatically deploy resources to patients in need in a timely manner, thereby improving patient outcomes while improving the bottom line”).
It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Bostic to incorporate the teachings of Zahora and account for streamlining the complexities of claims processing and records keeping on behalf of the medical provider and determining the correct billing amount can involve numerous factors, such as patient co-pay amount, patient deductible, multiple insurers and/or liability insurance, and location where the medical services were rendered (Zahora, Abstract and [0002]-[0004]).
Regarding claims 3 and 15 Bostic further teaches the operations further comprise: applying the trained machine learning model to a plurality of sets of healthcare attributes corresponding to a plurality of target healthcare services ([0041] and [0075]) to estimate characteristics of respective predicted reimbursements for the plurality of target healthcare services (Bostic, [0040]-[0042]); and prioritizing the plurality of target healthcare services for billing actions to be executed by users (Bostic, [0040]-[0042]).
Regarding claims 4 and 16 Bostic further teaches applying the trained machine learning model to the target healthcare service comprises applying the trained machine learning model to reimbursement submission data for requesting a reimbursement for the target healthcare service (Bostic, [0040]-[0042]).
Regarding claim 12 Bostic further teaches the first training data set comprises a data of a data type that is not used for generating medical claims (Bostic, [0193], [0287], [0289], and [0290]).
Regarding claim 13 Bostic teaches a method comprising: training a first machine learning model to predict characteristics of healthcare reimbursements for healthcare services for patients ([0040]-[0042]), the training comprising: obtaining a first training data set of a plurality of training data sets ([0022], [0033], [0040], and [0045]), the first training data set including: a first set of healthcare attributes corresponding to a first healthcare service for a first patient ([0023], [0041], [0068], [0086], and [0089]); and a first reimbursement amount approved for the first healthcare service (Abstract, [0040], [0041], [0072]); obtaining a second training data set of a plurality of training data sets ([0022], [0033], [0040], and [0045]), the second training data set including: a second set of healthcare attributes corresponding to a second healthcare service for a second patient ([0023], [0041], [0068], [0086], and [0089]); and a second reimbursement amount approved for the first healthcare service (Abstract, [0040], [0041], [0072]); training the first machine learning model using the plurality of training data sets to generate a trained machine learning model ([0039]-[0040]); obtaining¸ at a first point in time prior to generating a medical claim corresponding to a target healthcare service for a particular patient ([0023]: “matching the simulated future health state to a predicted patient medical service need; matching the predicted patient medical service need to at least one of the patient's healthcare providers” and [0041]: “the method includes monitoring insurance billing events, that includes ingesting patient data received from one of a plurality of patient data providers and healthcare services data relating to the patient data; ingesting data relating to insurance reimbursement criteria and insurance reimbursement records relating to the healthcare services data; determining one or more relationships between the ingested patient data, healthcare services data, insurance reimbursement criteria and insurance reimbursement records, and data and previously ingested patient data, healthcare services data, insurance reimbursement criteria and insurance reimbursement records wherein at least one new enriched data set is created based on the determined one or more relationships”), a first target set of healthcare attributes corresponding to a condition of the target healthcare service for the particular patient ([0023], [0041], [0086], and [0089]); and applying the trained machine learning model to the first target set of healthcare attributes to estimate characteristics of a predicted reimbursement for the target healthcare service ([0037] and [0040]-[0041]) corresponding to the condition of the particular patient ([0115], [0181], [0185], and [0245]).
Bostic does not teach applying the characteristics of the predicted reimbursement for the target healthcare service to a second machine learning model to generate a set of key features corresponding to healthcare attributes input to the trained machine learning model to estimate the characteristics of the predicted reimbursement for the target healthcare service, wherein the second machine learning model is trained to predict key features based on an impact of the set of key features on predicting, by the trained machine learning model, the characteristics of the healthcare reimbursements for healthcare services for patients; and based on the characteristics of the predicted reimbursement predicted by the first machine learning model and the set of key features predicted by the second machine learning model; identifying a first treatment to improve the predicted reimbursement; and generating, in real-time, a first recommendation to a healthcare provider to perform the first treatment, wherein including the first treatment plan administered for the particular patient by the healthcare provider at least one of (a) an amount of the predicted reimbursement, and (b) a time for payment of the predicted reimbursement for one or more treatments in the treatment plan.
However, Zahora teaches applying the characteristics of the predicted reimbursement for the target healthcare service to a second machine learning model (Zahora, FIG. 1A, FIG. 1B, FIG. 8, [0057-[0058]]: “FIG. 1A is a process flow diagram of an example process for using a predictive analytics platform to perform machine learning analysis on a data universe of information resources to produce complex predictive intelligence… FIG. 1B is a flow chart of an example method for performing machine learning analysis on a data universe of information resources to calculate a payment estimation by applying historic payment patterns to a present medical claim”, [0065]: “FIG. 8 is a flow diagram of an example process for predicting future revenue through performing machine learning analysis of historic medical procedures to identify and extrapolate commonly paired and/or follow-on procedures to procedures captured in recent claims data”, and [0092]-[0093]: “The platform as described herein creates a predictive model based on a hybrid of data for a particular patient with data for payers associated with the particular patient to provide a prediction and estimation for a medical claim for the particular patient… The historic information, in some examples, may include services rendered, products sold, patient demographic information, patient insurance information, and remittance received from various payers to reimburse provision of medical services and products. The data universe may further include information regarding the various payers and the payer plans, such as copayment levels, deductible amounts, reimbursement eligibility requirements, and reimbursement amounts”) to generate a set of key features corresponding to healthcare attributes input to the trained machine learning model to estimate the characteristics of the predicted reimbursement for the target healthcare service (Zahora, [0016]: “The payment pattern application engine may be configured to estimate a supplemental reimbursement from the medical payment assistance program”, [0021]: “the medical claim includes a pre-service estimate for medical services and the payment pattern application engine is configured to calculate a payment estimation and confidence score for the pre-service estimate for the medical services. The medical services may include at least one of medical transport and a prescription”, [0032]: “The payment pattern application engine may be configured to estimate a supplemental reimbursement from the medical payment assistance program.”, and [0129]: “he patient matching engine 230 may access the patient data 240 to match features to insured persons in the data repository 210. The patient matching engine 230, in some embodiments, returns exact matches. The patient matching engine 230, in other embodiments, returns both exact matches and similar matches. Similar matches may be qualified with a closeness rating or percentage similarity to the patient. The patient matching engine 230, for example, may return a set of patient identifiers for referencing the de-identified patient data 240”), wherein the second machine learning model is trained to predict key features based on an impact of the set of key features on predicting, by the trained machine learning model, the characteristics of the healthcare reimbursements for healthcare services for patients (Zahora, [0021]: “The payment pattern application engine may be configured to calculate a payment estimation and confidence score for one or more candidate medical procedure codes. The payment pattern application engine may be configured to provide alternative or preferred medical procedure codes based on the calculated payment estimations… The payment pattern application engine may be configured to calculate payment estimations and confidence scores and provide a recommended patient pre-payment amount based on the calculated payment estimations and confidence scores” and [0269]: “In some implementations, the predictive analytics platform 102 includes a patient pre-payment calculation engine 1092 to determine the recommended pre-payment amount 1092 based on the balance due 1090 and the likelihood of patient payment 1088. For example, although the patient's likelihood to pay $1,000 is medium 1088 a with a confidence score of 98/100 1088 b, the patient may have a higher likelihood of covering a portion of the balance due 1090. Further, the likelihood of the patient repaying any portion of the balance due 1090 may drop significantly after leaving the facility (e.g., when being billed to a home address 1094). Thus, the patient pre-payment calculation engine 1092 may analyze remittance data 268 related to similar patients (e.g., identified by the patient matching engine 230) to identify an amount of funds up to the balance due 1090 that is associated with a high likelihood of payment. The patient pre-payment calculation engine 1092 may further identify the amount based on at least a threshold confidence score of the high likelihood that the patient will pay the amount. The patient pre-payment calculation engine 1092 may perform many of the same functions of the patient likelihood to pay analysis engine 228, but with a goal of a likelihood rating rather than a goal of a particular reimbursement amount”); and based on the characteristics of the predicted reimbursement predicted by the first machine learning model and the set of key features predicted by the second machine learning model (Zahora, [0016]: “The payment pattern application engine may be configured to estimate a supplemental reimbursement from the medical payment assistance program”, [0021]: “the medical claim includes a pre-service estimate for medical services and the payment pattern application engine is configured to calculate a payment estimation and confidence score for the pre-service estimate for the medical services. The medical services may include at least one of medical transport and a prescription”, [0032]: “The payment pattern application engine may be configured to estimate a supplemental reimbursement from the medical payment assistance program.”, and [0129]: “he patient matching engine 230 may access the patient data 240 to match features to insured persons in the data repository 210. The patient matching engine 230, in some embodiments, returns exact matches. The patient matching engine 230, in other embodiments, returns both exact matches and similar matches. Similar matches may be qualified with a closeness rating or percentage similarity to the patient. The patient matching engine 230, for example, may return a set of patient identifiers for referencing the de-identified patient data 240”, and [0269]: “In some implementations, the predictive analytics platform 102 includes a patient pre-payment calculation engine 1092 to determine the recommended pre-payment amount 1092 based on the balance due 1090 and the likelihood of patient payment 1088. For example, although the patient's likelihood to pay $1,000 is medium 1088 a with a confidence score of 98/100 1088 b, the patient may have a higher likelihood of covering a portion of the balance due 1090. Further, the likelihood of the patient repaying any portion of the balance due 1090 may drop significantly after leaving the facility (e.g., when being billed to a home address 1094). Thus, the patient pre-payment calculation engine 1092 may analyze remittance data 268 related to similar patients (e.g., identified by the patient matching engine 230) to identify an amount of funds up to the balance due 1090 that is associated with a high likelihood of payment. The patient pre-payment calculation engine 1092 may further identify the amount based on at least a threshold confidence score of the high likelihood that the patient will pay the amount. The patient pre-payment calculation engine 1092 may perform many of the same functions of the patient likelihood to pay analysis engine 228, but with a goal of a likelihood rating rather than a goal of a particular reimbursement amount”); identifying a first treatment to improve the predicted reimbursement (Zahora, FIG. 17A through 17C, [0083], [0092]: “The platform may utilize demographic and insurance coverage discovery to further improve the accuracy and timeliness of the predictions and estimations. “Discovery” refers to a process of finding and identifying information relevant to a medical claim that may be unknown and/or unavailable to a patient, payer, and/or provider prior to the discovery process or information of which the patient, payer, and/or provider was unaware of prior to the discovery process”, [0094]: “applying machine learning algorithms and data analytics to historic information, medical providers may be supplied with complex predictive intelligence, such as dependable revenue stream estimates and revenue forecasting. Thus, systems and methods described herein provide a technical solution to the technical problem of reliably predicting revenue streams from an abundance of payer sources each having different reimbursement levels and requirements for coverage eligibility. Further, the inventors recognized that in combining data sources and applying machine learning analysis, the resultant data analytics may support real-time or near-real-time decision making that had not previously been possible, such as rapid patient pre-approval for medical procedures and prescription products as well as complex dynamic scheduling of mobile medical services. Thus, systems and methods described herein additionally provide a technical solution to the technical problem of overcoming resource availability obstacles and reimbursement obstacles to automatically deploy resources to patients in need in a timely manner, thereby improving patient outcomes while improving the bottom line”, [0116]: “a type of medical service (e.g., a life-changing service such as cancer treatment, limb amputation, etc. versus a non-life-changing service such as tonsil removal) and/or a cost to patient range (e.g., hundreds, thousands, over ten thousand, etc.) may be considered in matching the first patient to the one or more second patients. The patient payment pattern, for example, may include one or more likelihoods (e.g., percentage likelihood, likelihood on a scale of one to ten), such as a first likelihood related to a patient most demographically similar but with a more dissimilar cost to patient range and a second likelihood related to a patient less demographically similar but within a same cost to patient range as the medical claim for the first patient” and [0280]: “the charting service 1220 may utilize the supplemented and/or corrected demographic information to obtain medical records for the victim that may enable improved care for the victim”); and generating, in real-time, a first recommendation to a healthcare provider to perform the first treatment (Zahora, FIG. 17A through 17C, [0083], and [0092]: “these systems and methods enable the patient to better plan their treatments with evaluations of care providers, treatment options, and associated costs”, [0094]: “applying machine learning algorithms and data analytics to historic information, medical providers may be supplied with complex predictive intelligence, such as dependable revenue stream estimates and revenue forecasting. Thus, systems and methods described herein provide a technical solution to the technical problem of reliably predicting revenue streams from an abundance of payer sources each having different reimbursement levels and requirements for coverage eligibility. Further, the inventors recognized that in combining data sources and applying machine learning analysis, the resultant data analytics may support real-time or near-real-time decision making that had not previously been possible, such as rapid patient pre-approval for medical procedures and prescription products as well as complex dynamic scheduling of mobile medical services. Thus, systems and methods described herein additionally provide a technical solution to the technical problem of overcoming resource availability obstacles and reimbursement obstacles to automatically deploy resources to patients in need in a timely manner, thereby improving patient outcomes while improving the bottom line”, [0116]: “a type of medical service (e.g., a life-changing service such as cancer treatment, limb amputation, etc. versus a non-life-changing service such as tonsil removal) and/or a cost to patient range (e.g., hundreds, thousands, over ten thousand, etc.) may be considered in matching the first patient to the one or more second patients. The patient payment pattern, for example, may include one or more likelihoods (e.g., percentage likelihood, likelihood on a scale of one to ten), such as a first likelihood related to a patient most demographically similar but with a more dissimilar cost to patient range and a second likelihood related to a patient less demographically similar but within a same cost to patient range as the medical claim for the first patient”, and [0278]: “The ePCR may include a detailed record of treatments and interventions provided to the patient by the caregivers along with patient transport information. The predictive analytics platform 102 may receive the ePCR from the charting service 1220 and may populate a medical claim with information recorded in the ePCR and/or utilize the medical procedure codes indicated by and/or recorded in the ePCR to estimate payments as described above. In an implementation, the medical device(s) 1296 may communicatively couple to the charting service 1220 and/or the mobile device 1280 to provide medical treatment information for automatic recordation in the ePCR”), wherein including the first treatment plan administered for the particular patient by the healthcare provider at least one of (a) an amount of the predicted reimbursement (Zahora, FIG. 17A through 17C, [0013]: “calculating the payment estimation includes applying at least one deductible amount corresponding to a deductible level of an active payer plan of first patient. Calculating the payment estimation may include automatically determining a first deductible amount of the at least one deductible amount by contacting an external computing system of the payer via a network to request a current balance of a remaining maximum deductible”, [0021]: “the medical claim includes a pre-service estimate for medical services and the payment pattern application engine is configured to calculate a payment estimation and confidence score for the pre-service estimate for the medical service”, [0022]: “The payment pattern application engine may be configured to receive a patient medical condition from the CAD, calculate payment estimations based on the patient medical condition and the identified insurance coverages, and provide the payment estimations to one or more of the CAD and the ePCR. The payment pattern application engine may be configured to provide one or more of a recommended transport destination and medical procedure codes based on the payment estimations and identified insurance coverages. The payment pattern application engine may be configured to receive updated medical condition information from the ePCR application, update the payment estimations based on the updated medical condition information, and provide the updated payment estimations to the ePCR application”, [0116]: “a type of medical service (e.g., a life-changing service such as cancer treatment, limb amputation, etc. versus a non-life-changing service such as tonsil removal) and/or a cost to patient range (e.g., hundreds, thousands, over ten thousand, etc.) may be considered in matching the first patient to the one or more second patients. The patient payment pattern, for example, may include one or more likelihoods (e.g., percentage likelihood, likelihood on a scale of one to ten), such as a first likelihood related to a patient most demographically similar but with a more dissimilar cost to patient range and a second likelihood related to a patient less demographically similar but within a same cost to patient range as the medical claim for the first patient”, and [0278]: “The ePCR may include a detailed record of treatments and interventions provided to the patient by the caregivers along with patient transport information. The predictive analytics platform 102 may receive the ePCR from the charting service 1220 and may populate a medical claim with information recorded in the ePCR and/or utilize the medical procedure codes indicated by and/or recorded in the ePCR to estimate payments as described above. In an implementation, the medical device(s) 1296 may communicatively couple to the charting service 1220 and/or the mobile device 1280 to provide medical treatment information for automatic recordation in the ePCR”), and (b) a time for payment of the predicted reimbursement for one or more treatments in the treatment plan (Zahora, FIG. 17A through 17C, [0083], and [0094]: “applying machine learning algorithms and data analytics to historic information, medical providers may be supplied with complex predictive intelligence, such as dependable revenue stream estimates and revenue forecasting. Thus, systems and methods described herein provide a technical solution to the technical problem of reliably predicting revenue streams from an abundance of payer sources each having different reimbursement levels and requirements for coverage eligibility. Further, the inventors recognized that in combining data sources and applying machine learning analysis, the resultant data analytics may support real-time or near-real-time decision making that had not previously been possible, such as rapid patient pre-approval for medical procedures and prescription products as well as complex dynamic scheduling of mobile medical services. Thus, systems and methods described herein additionally provide a technical solution to the technical problem of overcoming resource availability obstacles and reimbursement obstacles to automatically deploy resources to patients in need in a timely manner, thereby improving patient outcomes while improving the bottom line”).
It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Bostic to incorporate the teachings of Zahora and account for streamlining the complexities of claims processing and records keeping on behalf of the medical provider and determining the correct billing amount can involve numerous factors, such as patient co-pay amount, patient deductible, multiple insurers and/or liability insurance, and location where the medical services were rendered (Zahora, Abstract and [0002]-[0004]).
Regarding claim 20 Bostic teaches a system comprising: one or more processors; and memory storing instructions that, when executed by the one or more processors, cause the system to perform operations comprising: training a first machine learning model to predict characteristics of healthcare reimbursements for healthcare services for patients ([0040]-[0042]), the training comprising: obtaining a first training data set of a plurality of training data sets ([0022], [0033], [0040], and [0045]), the first training data set including: a first set of healthcare attributes corresponding to a first healthcare service for a first patient ([0023], [0041], [0068], [0086], and [0089]); and a first reimbursement amount approved for the first healthcare service (Abstract, [0040], [0041], [0072]); obtaining a second training data set of a plurality of training data sets ([0022], [0033], [0040], and [0045]), the second training data set including: a second set of healthcare attributes corresponding to a second healthcare service for a second patient ([0023], [0041], [0068], [0086], and [0089]); and a second reimbursement amount approved for the first healthcare service (Abstract, [0040], [0041], [0072]); training the first machine learning model using the plurality of training data sets to generate a trained machine learning model ([0039]-[0040]); obtaining¸ at a first point in time prior to generating a medical claim corresponding to a target healthcare service for a particular patient ([0023]: “matching the simulated future health state to a predicted patient medical service need; matching the predicted patient medical service need to at least one of the patient's healthcare providers” and [0041]: “the method includes monitoring insurance billing events, that includes ingesting patient data received from one of a plurality of patient data providers and healthcare services data relating to the patient data; ingesting data relating to insurance reimbursement criteria and insurance reimbursement records relating to the healthcare services data; determining one or more relationships between the ingested patient data, healthcare services data, insurance reimbursement criteria and insurance reimbursement records, and data and previously ingested patient data, healthcare services data, insurance reimbursement criteria and insurance reimbursement records wherein at least one new enriched data set is created based on the determined one or more relationships”), a first target set of healthcare attributes corresponding to a condition of the target healthcare service for the particular patient ([0023], [0041], [0086], and [0089]); and applying the trained machine learning model to the first target set of healthcare attributes to estimate characteristics of a predicted reimbursement for the target healthcare service ([0037] and [0040]-[0041]) corresponding to the condition of the particular patient ([0115], [0181], [0185], and [0245]).
Bostic does not teach applying the characteristics of the predicted reimbursement for the target healthcare service to a second machine learning model to generate a set of key features corresponding to healthcare attributes input to the trained machine learning model to estimate the characteristics of the predicted reimbursement for the target healthcare service, wherein the second machine learning model is trained to predict key features based on an impact of the set of key features on predicting, by the trained machine learning model, the characteristics of the healthcare reimbursements for healthcare services for patients; and based on the characteristics of the predicted reimbursement predicted by the first machine learning model and the set of key features predicted by the second machine learning model; identifying a first treatment to improve the predicted reimbursement; and generating, in real-time, a first recommendation to a healthcare provider to perform the first treatment, wherein including the first treatment plan administered for the particular patient by the healthcare provider at least one of (a) an amount of the predicted reimbursement, and (b) a time for payment of the predicted reimbursement for one or more treatments in the treatment plan.
However, Zahora teaches applying the characteristics of the predicted reimbursement for the target healthcare service to a second machine learning model (Zahora, FIG. 1A, FIG. 1B, FIG. 8, [0057-[0058]]: “FIG. 1A is a process flow diagram of an example process for using a predictive analytics platform to perform machine learning analysis on a data universe of information resources to produce complex predictive intelligence… FIG. 1B is a flow chart of an example method for performing machine learning analysis on a data universe of information resources to calculate a payment estimation by applying historic payment patterns to a present medical claim”, [0065]: “FIG. 8 is a flow diagram of an example process for predicting future revenue through performing machine learning analysis of historic medical procedures to identify and extrapolate commonly paired and/or follow-on procedures to procedures captured in recent claims data”, and [0092]-[0093]: “The platform as described herein creates a predictive model based on a hybrid of data for a particular patient with data for payers associated with the particular patient to provide a prediction and estimation for a medical claim for the particular patient… The historic information, in some examples, may include services rendered, products sold, patient demographic information, patient insurance information, and remittance received from various payers to reimburse provision of medical services and products. The data universe may further include information regarding the various payers and the payer plans, such as copayment levels, deductible amounts, reimbursement eligibility requirements, and reimbursement amounts”) to generate a set of key features corresponding to healthcare attributes input to the trained machine learning model to estimate the characteristics of the predicted reimbursement for the target healthcare service (Zahora, [0016]: “The payment pattern application engine may be configured to estimate a supplemental reimbursement from the medical payment assistance program”, [0021]: “the medical claim includes a pre-service estimate for medical services and the payment pattern application engine is configured to calculate a payment estimation and confidence score for the pre-service estimate for the medical services. The medical services may include at least one of medical transport and a prescription”, [0032]: “The payment pattern application engine may be configured to estimate a supplemental reimbursement from the medical payment assistance program.”, and [0129]: “he patient matching engine 230 may access the patient data 240 to match features to insured persons in the data repository 210. The patient matching engine 230, in some embodiments, returns exact matches. The patient matching engine 230, in other embodiments, returns both exact matches and similar matches. Similar matches may be qualified with a closeness rating or percentage similarity to the patient. The patient matching engine 230, for example, may return a set of patient identifiers for referencing the de-identified patient data 240”), wherein the second machine learning model is trained to predict key features based on an impact of the set of key features on predicting, by the trained machine learning model, the characteristics of the healthcare reimbursements for healthcare services for patients (Zahora, [0021]: “The payment pattern application engine may be configured to calculate a payment estimation and confidence score for one or more candidate medical procedure codes. The payment pattern application engine may be configured to provide alternative or preferred medical procedure codes based on the calculated payment estimations… The payment pattern application engine may be configured to calculate payment estimations and confidence scores and provide a recommended patient pre-payment amount based on the calculated payment estimations and confidence scores” and [0269]: “In some implementations, the predictive analytics platform 102 includes a patient pre-payment calculation engine 1092 to determine the recommended pre-payment amount 1092 based on the balance due 1090 and the likelihood of patient payment 1088. For example, although the patient's likelihood to pay $1,000 is medium 1088 a with a confidence score of 98/100 1088 b, the patient may have a higher likelihood of covering a portion of the balance due 1090. Further, the likelihood of the patient repaying any portion of the balance due 1090 may drop significantly after leaving the facility (e.g., when being billed to a home address 1094). Thus, the patient pre-payment calculation engine 1092 may analyze remittance data 268 related to similar patients (e.g., identified by the patient matching engine 230) to identify an amount of funds up to the balance due 1090 that is associated with a high likelihood of payment. The patient pre-payment calculation engine 1092 may further identify the amount based on at least a threshold confidence score of the high likelihood that the patient will pay the amount. The patient pre-payment calculation engine 1092 may perform many of the same functions of the patient likelihood to pay analysis engine 228, but with a goal of a likelihood rating rather than a goal of a particular reimbursement amount”); and based on the characteristics of the predicted reimbursement predicted by the first machine learning model and the set of key features predicted by the second machine learning model (Zahora, [0016]: “The payment pattern application engine may be configured to estimate a supplemental reimbursement from the medical payment assistance program”, [0021]: “the medical claim includes a pre-service estimate for medical services and the payment pattern application engine is configured to calculate a payment estimation and confidence score for the pre-service estimate for the medical services. The medical services may include at least one of medical transport and a prescription”, [0032]: “The payment pattern application engine may be configured to estimate a supplemental reimbursement from the medical payment assistance program.”, and [0129]: “he patient matching engine 230 may access the patient data 240 to match features to insured persons in the data repository 210. The patient matching engine 230, in some embodiments, returns exact matches. The patient matching engine 230, in other embodiments, returns both exact matches and similar matches. Similar matches may be qualified with a closeness rating or percentage similarity to the patient. The patient matching engine 230, for example, may return a set of patient identifiers for referencing the de-identified patient data 240”, and [0269]: “In some implementations, the predictive analytics platform 102 includes a patient pre-payment calculation engine 1092 to determine the recommended pre-payment amount 1092 based on the balance due 1090 and the likelihood of patient payment 1088. For example, although the patient's likelihood to pay $1,000 is medium 1088 a with a confidence score of 98/100 1088 b, the patient may have a higher likelihood of covering a portion of the balance due 1090. Further, the likelihood of the patient repaying any portion of the balance due 1090 may drop significantly after leaving the facility (e.g., when being billed to a home address 1094). Thus, the patient pre-payment calculation engine 1092 may analyze remittance data 268 related to similar patients (e.g., identified by the patient matching engine 230) to identify an amount of funds up to the balance due 1090 that is associated with a high likelihood of payment. The patient pre-payment calculation engine 1092 may further identify the amount based on at least a threshold confidence score of the high likelihood that the patient will pay the amount. The patient pre-payment calculation engine 1092 may perform many of the same functions of the patient likelihood to pay analysis engine 228, but with a goal of a likelihood rating rather than a goal of a particular reimbursement amount”); identifying a first treatment to improve the predicted reimbursement (Zahora, FIG. 17A through 17C, [0083], [0092]: “The platform may utilize demographic and insurance coverage discovery to further improve the accuracy and timeliness of the predictions and estimations. “Discovery” refers to a process of finding and identifying information relevant to a medical claim that may be unknown and/or unavailable to a patient, payer, and/or provider prior to the discovery process or information of which the patient, payer, and/or provider was unaware of prior to the discovery process”, [0094]: “applying machine learning algorithms and data analytics to historic information, medical providers may be supplied with complex predictive intelligence, such as dependable revenue stream estimates and revenue forecasting. Thus, systems and methods described herein provide a technical solution to the technical problem of reliably predicting revenue streams from an abundance of payer sources each having different reimbursement levels and requirements for coverage eligibility. Further, the inventors recognized that in combining data sources and applying machine learning analysis, the resultant data analytics may support real-time or near-real-time decision making that had not previously been possible, such as rapid patient pre-approval for medical procedures and prescription products as well as complex dynamic scheduling of mobile medical services. Thus, systems and methods described herein additionally provide a technical solution to the technical problem of overcoming resource availability obstacles and reimbursement obstacles to automatically deploy resources to patients in need in a timely manner, thereby improving patient outcomes while improving the bottom line”, [0116]: “a type of medical service (e.g., a life-changing service such as cancer treatment, limb amputation, etc. versus a non-life-changing service such as tonsil removal) and/or a cost to patient range (e.g., hundreds, thousands, over ten thousand, etc.) may be considered in matching the first patient to the one or more second patients. The patient payment pattern, for example, may include one or more likelihoods (e.g., percentage likelihood, likelihood on a scale of one to ten), such as a first likelihood related to a patient most demographically similar but with a more dissimilar cost to patient range and a second likelihood related to a patient less demographically similar but within a same cost to patient range as the medical claim for the first patient” and [0280]: “the charting service 1220 may utilize the supplemented and/or corrected demographic information to obtain medical records for the victim that may enable improved care for the victim”); and generating, in real-time, a first recommendation to a healthcare provider to perform the first treatment (Zahora, FIG. 17A through 17C, [0083], and [0092]: “these systems and methods enable the patient to better plan their treatments with evaluations of care providers, treatment options, and associated costs”, [0094]: “applying machine learning algorithms and data analytics to historic information, medical providers may be supplied with complex predictive intelligence, such as dependable revenue stream estimates and revenue forecasting. Thus, systems and methods described herein provide a technical solution to the technical problem of reliably predicting revenue streams from an abundance of payer sources each having different reimbursement levels and requirements for coverage eligibility. Further, the inventors recognized that in combining data sources and applying machine learning analysis, the resultant data analytics may support real-time or near-real-time decision making that had not previously been possible, such as rapid patient pre-approval for medical procedures and prescription products as well as complex dynamic scheduling of mobile medical services. Thus, systems and methods described herein additionally provide a technical solution to the technical problem of overcoming resource availability obstacles and reimbursement obstacles to automatically deploy resources to patients in need in a timely manner, thereby improving patient outcomes while improving the bottom line”, [0116]: “a type of medical service (e.g., a life-changing service such as cancer treatment, limb amputation, etc. versus a non-life-changing service such as tonsil removal) and/or a cost to patient range (e.g., hundreds, thousands, over ten thousand, etc.) may be considered in matching the first patient to the one or more second patients. The patient payment pattern, for example, may include one or more likelihoods (e.g., percentage likelihood, likelihood on a scale of one to ten), such as a first likelihood related to a patient most demographically similar but with a more dissimilar cost to patient range and a second likelihood related to a patient less demographically similar but within a same cost to patient range as the medical claim for the first patient”, and [0278]: “The ePCR may include a detailed record of treatments and interventions provided to the patient by the caregivers along with patient transport information. The predictive analytics platform 102 may receive the ePCR from the charting service 1220 and may populate a medical claim with information recorded in the ePCR and/or utilize the medical procedure codes indicated by and/or recorded in the ePCR to estimate payments as described above. In an implementation, the medical device(s) 1296 may communicatively couple to the charting service 1220 and/or the mobile device 1280 to provide medical treatment information for automatic recordation in the ePCR”), wherein including the first treatment plan administered for the particular patient by the healthcare provider at least one of (a) an amount of the predicted reimbursement (Zahora, FIG. 17A through 17C, [0013]: “calculating the payment estimation includes applying at least one deductible amount corresponding to a deductible level of an active payer plan of first patient. Calculating the payment estimation may include automatically determining a first deductible amount of the at least one deductible amount by contacting an external computing system of the payer via a network to request a current balance of a remaining maximum deductible”, [0021]: “the medical claim includes a pre-service estimate for medical services and the payment pattern application engine is configured to calculate a payment estimation and confidence score for the pre-service estimate for the medical service”, [0022]: “The payment pattern application engine may be configured to receive a patient medical condition from the CAD, calculate payment estimations based on the patient medical condition and the identified insurance coverages, and provide the payment estimations to one or more of the CAD and the ePCR. The payment pattern application engine may be configured to provide one or more of a recommended transport destination and medical procedure codes based on the payment estimations and identified insurance coverages. The payment pattern application engine may be configured to receive updated medical condition information from the ePCR application, update the payment estimations based on the updated medical condition information, and provide the updated payment estimations to the ePCR application”, [0116]: “a type of medical service (e.g., a life-changing service such as cancer treatment, limb amputation, etc. versus a non-life-changing service such as tonsil removal) and/or a cost to patient range (e.g., hundreds, thousands, over ten thousand, etc.) may be considered in matching the first patient to the one or more second patients. The patient payment pattern, for example, may include one or more likelihoods (e.g., percentage likelihood, likelihood on a scale of one to ten), such as a first likelihood related to a patient most demographically similar but with a more dissimilar cost to patient range and a second likelihood related to a patient less demographically similar but within a same cost to patient range as the medical claim for the first patient”, and [0278]: “The ePCR may include a detailed record of treatments and interventions provided to the patient by the caregivers along with patient transport information. The predictive analytics platform 102 may receive the ePCR from the charting service 1220 and may populate a medical claim with information recorded in the ePCR and/or utilize the medical procedure codes indicated by and/or recorded in the ePCR to estimate payments as described above. In an implementation, the medical device(s) 1296 may communicatively couple to the charting service 1220 and/or the mobile device 1280 to provide medical treatment information for automatic recordation in the ePCR”), and (b) a time for payment of the predicted reimbursement for one or more treatments in the treatment plan (Zahora, FIG. 17A through 17C, [0083], and [0094]: “applying machine learning algorithms and data analytics to historic information, medical providers may be supplied with complex predictive intelligence, such as dependable revenue stream estimates and revenue forecasting. Thus, systems and methods described herein provide a technical solution to the technical problem of reliably predicting revenue streams from an abundance of payer sources each having different reimbursement levels and requirements for coverage eligibility. Further, the inventors recognized that in combining data sources and applying machine learning analysis, the resultant data analytics may support real-time or near-real-time decision making that had not previously been possible, such as rapid patient pre-approval for medical procedures and prescription products as well as complex dynamic scheduling of mobile medical services. Thus, systems and methods described herein additionally provide a technical solution to the technical problem of overcoming resource availability obstacles and reimbursement obstacles to automatically deploy resources to patients in need in a timely manner, thereby improving patient outcomes while improving the bottom line”).
It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Bostic to incorporate the teachings of Zahora and account for streamlining the complexities of claims processing and records keeping on behalf of the medical provider and determining the correct billing amount can involve numerous factors, such as patient co-pay amount, patient deductible, multiple insurers and/or liability insurance, and location where the medical services were rendered (Zahora, Abstract and [0002]-[0004]).
Regarding claim 22 Bostic further teaches training the first machine learning model comprises: using a machine learning algorithm to apply a first iteration of the first machine learning model to the first training data set to generate a first set of results (Bostic, [0050]-[0053], [0067], [0083], [0085], and [0172]); adjusting parameters of the first iteration of the first machine learning model based on the first set of results to generate a second iteration of the first machine learning model (Bostic, [0050]-[0053], [0067], [0083], [0085], and [0172]); and using the first machine learning algorithm to apply the second iteration of the machine learning model to the second training data set (Bostic, [0050]-[0053], [0067], [0083], [0085], and [0172]).
Regarding claim 23 Bostic further teaches the first training data set comprises the first set of attributes that is different from the second set of attributes comprised in the second training data set, and wherein the operations further comprise: generating a first vector representing the first training data set; and generating a second vector representing the second training data set, wherein the first machine learning model is configured to receive input data comprising a first set of features, wherein training the first machine learning model comprises providing the first vector and the second vector to the first machine learning model, and wherein the first vector and the second vector each comprise the first set of features (Bostic, [0050]-[0053], [0067], [0083], [0085], [0172], [0214], and [0231]).
Claim(s) 2 and 14 are rejected under 35 U.S.C. 103 as being unpatentable over Bostic and Zahora, in view of Lintereur et al. (US 20220176039 A1), hereinafter Lintereur and Poteet et al. (US 2021020253 B1), hereinafter Poteet.
Regarding claims 2 and 14 Bostic teaches the non-transitory computer readable medium of claim 1, the method of claim 13.
Bostic and Zahora do not teach obtaining data identifying an actual reimbursement for the target healthcare service.
However, Lintereur teaches obtaining data identifying an actual reimbursement for the target healthcare service (Lintereur, [0003], [0028], [0065], and [0082]).
It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Bostic to incorporate the teachings of Lintereur and account for obtain data identifying an actual reimbursement used in modeling in order to allow for immediate reimbursement to the physician for the patient treatment provided (Lintereur, [0003] and [0006]).
Bostic, Zahora, and Lintereur do not teach updating the trained machine learning model based on the actual reimbursement.
However, Poteet teaches updating the trained machine learning model based on the actual reimbursement (Poteet, Abstract and [0061]).
It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Bostic, Zahora, and Lintereur to incorporate the teachings of Poteet and account for updating a machine learning model based on the reimbursement (Poteet, Abstract and [0061]).
Claim(s) 5, 8, and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Bostic and Zahora, in view of Morrison et al. (US 11244767 B1), hereinafter Morrison.
Regarding claims 5 and 17 Bostic teaches the non-transitory computer readable medium of claim 4 and the method of claim 16.
Bostic and Zahora do not teach operations further comprise generating a recommendation for modifying the reimbursement submission data based on the characteristics of the predicted reimbursement for the target healthcare service.
However, Morrison teaches operations further comprise generating a recommendation for modifying the reimbursement submission data based on the characteristics of the predicted reimbursement for the target healthcare service (Morrison, (101) Col. 37, lines 6-67).
It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Bostic and Zahora to incorporate the teachings of Morrison and account for modify a reimbursement claim based on a reimbursement prediction of the health service (Morrison, Abstract, (6), (7), and (14)).
Regarding claim 8 Bostic teaches the non-transitory computer readable medium of claim 1.
Bostic and Zahora do not teach the characteristics of the predicted reimbursement comprise a predicted time of payment for at least a portion of a reimbursement for the target healthcare service.
However, Morrison teaches the characteristics of the predicted reimbursement comprise a predicted time of payment for at least a portion of a reimbursement for the target healthcare service (Morrison, (101), Col. 37, lines 6-67).
It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Bostic and Zahora to incorporate the teachings of Morrison and account for predicting a time a reimbursement will be issued for a health service (Morrison, Abstract, (6), (7), and (14)).
Claim(s) 7, 10, and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Bostic and Zahora, in view of Lintereur et al. (US 20220176039 A1), hereinafter Lintereur.
Regarding claims 7 and 19 Bostic teaches the non-transitory computer readable medium of claim 1 and the method of claim 13.
Bostic and Zahora do not teach the characteristics of the predicted reimbursement comprise a reimbursement amount for the target healthcare service.
However, Lintereur teaches the characteristics of the predicted reimbursement comprise a reimbursement amount for the target healthcare service (Lintereur, [0021]-[0022] and [0081]-[0082]).
It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Bostic and Zahora to incorporate the teachings of Lintereur and account for a reimbursement amount for a health service is part of the predicted reimbursement (Lintereur, Abstract and [0006]).
Regarding claim 10 Bostic teaches the non-transitory computer readable medium of claim 1, wherein the first set of attributes corresponding to the first healthcare service (Bostic, [0023], [0041], [0068], [0086], and [0089]) comprise one or more of; a type of the first healthcare service (Bostic, [0101]-[0102]); an age of the first patient (Bostic, [0077]); an ordering provider that provided the first healthcare service (Bostic, [0101]-[0102]); a drug code for a drug administered during the first healthcare service (Bostic, [0101]-[0102]); and a monetary value representing a requested reimbursement amount associated with the first healthcare service (Bostic, [0040]-[0042]).
Bostic and Zahora do not teach a healthcare service code corresponding to the first healthcare service.
However, Lintereur teaches a healthcare service code corresponding to the first healthcare service (Lintereur, [0023] and [0082]).
It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Bostic and Zahora to incorporate the teachings of Lintereur and account for a healthcare service code for the health service (Lintereur, Abstract, [0023], and [0082]).
Claim(s) 6 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Bostic and Zahora, in view of Padmanabh et al. (US 20180322556 A1), hereinafter Padmanabh.
Regarding claims 6 and 18 Bostic teaches the non-transitory computer readable medium of claim 5 and method of claim 17.
Bostic and Zahora do not teach the reimbursement submission data comprises vendor-agnostic data.
However, Padmanabh teaches the reimbursement submission data comprises vendor-agnostic data (Padmanabh, [0078]).
It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Bostic and Zahora to incorporate the teachings of Padmanabh and account for the reimbursement data also contains vendor-agnostic data (Padmanabh, Abstract and [0078]).
Claim 9 is rejected under 35 U.S.C. 103 as being unpatentable over Bostic et al. (US 20210202099
A1), hereinafter Bostic, in view of Ketchel et al. (US 20220036413 A1), hereinafter Ketchel.
Regarding claim 9 Bostic teaches the non-transitory computer readable medium of claim 1.
Bostic does not teach the predicted characteristics of the reimbursement comprise a reimbursement percentage of an amount billed for the target healthcare service.
However, Ketchel teaches the predicted characteristics of the reimbursement comprise a reimbursement percentage of an amount billed for the target healthcare service (Ketchel, [0216]).
It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Bostic to incorporate the teachings of Ketchel and account for a percentage a person received in reimbursements for a health service (Ketchel, Abstract and [0216]).
Claim 11 is rejected under 35 U.S.C. 103 as being unpatentable over Bostic and Zahora, in view of Menard et al. (US 20230214455 A1), hereinafter Menard.
Regarding claim 11 Bostic teaches the non-transitory computer readable medium of claim 1.
Bostic and Zahora do not teach the first training data set further comprises a payer that approved the first reimbursement amount.
However, Menard teaches the first training data set further comprises a payer that approved the first reimbursement amount (Menard, [0004]-[0005], [0007]-[0009], and [0044]).
It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Bostic and Zahora to incorporate the teachings of Menard and account for approving the reimbursement amount to be paid (Lintereur, Abstract and [0002]).
Claim 21 is rejected under 35 U.S.C. 103 as being unpatentable over Bostic and Zahora, in view of Dunn et al. (US 20200286616A1), hereinafter Dunn.
Regarding claim 21 Bostic teaches the non-transitory computer readable medium of claim 1.
Bostic and Zahora do not teach training the machine learning model using the plurality of training data sets comprises: (a) applying a machine learning algorithm to a particular training data set to generate a first set of labels, comprising at least one of a set of reimbursement amounts and a set of treatments to improve the set of reimbursement amounts; (b) applying an error function to determine a set of adjustments to one or both of weights and offsets of the machine learning model; (c) adjusting the one or both of the weights and offsets of the machine learning model based on the error function; and (d) iteratively performing (a)-(c) to training data sets until a performance metric is achieved for the trained machine learning model.
However, Dunn teaches training the first machine learning model using the plurality of training data sets comprises: (a) applying a machine learning algorithm to a particular training data set to generate a first set of labels, comprising at least one of a set of reimbursement amounts and a set of treatments to improve the set of reimbursement amounts; (b) applying an error function to determine a set of adjustments to one or both of weights and offsets of the machine learning model; (c) adjusting the one or both of the weights and offsets of the machine learning model based on the error function; and (d) iteratively performing (a)-(c) to training data sets until a performance metric is achieved for the trained machine learning model (Dunn, [0025], [0094], [0131], [0147]-[0150], [0156], and [0216]-[0220]).
It would have been obvious to one of ordinary skill in the art before the effective filing date to modify Bostic and Zahora to incorporate the teachings of Dunn and account for insurance carriers to verify, evaluate, and adjudicate these claims efficiently (Dunn, Abstract, and [0002]-[0004]).
Response to Arguments
Applicant's arguments filed 08/28/2025 have been fully considered but they are not persuasive.
Regarding the 35 U.S.C. 101 Rejection, Applicant argues the claims are not directed to an abstract idea. Applicant also argues if a claim is directed to a specific improvement in computer performance, the claim is not directed to an abstract idea. Examiner respectfully disagrees. The limitations of “obtaining a first training data set…”, “obtaining a second training data set…”, “obtaining, at a first point in time prior to generating a medical claim…”, “applying the characteristics of the predict reimbursement…”, “identifying a first treatment…”, “generating, in real-time, a first recommendation…”, and “administered for the particular patient” are all step that can be performed by a person, persons, or person via computer tools (see MPEP 2106.04(a)(2) states “the sub-groupings encompass both activity of a single person (for example, a person following a set of instructions or a person signing a contract online) and activity that involves multiple people (such as a commercial interaction), and thus, certain activity between a person and a computer (for example a method of anonymous loan shopping that a person conducts using a mobile phone) may fall within the “certain methods of organizing human activity” grouping”). Additionally, an improvement in the abstract idea itself is not eligible. Only an improvement in the technical field/technology via the additional elements is 101 eligible. An improvement to the technology would be assessed in Step 2A, Prong 2. An improvement would need to be provided to the additional elements, which would then integrate the abstract idea into a practical application. However, the claims do not recite a technological improvement, but merely applies the abstract idea to a computer environment/tool (i.e., machine learning model) (see MPEP 2106.04(d) states “Merely reciting the words “apply it” (or an equivalent) with the judicial exception, or merely including instructions to implement an abstract idea on a computer, or merely using a computer as a tool to perform an abstract idea, as discussed in MPEP § 2106.05(f)”).
Applicant argues the specification recites applying a machine learning model to patient data prior to generating a medical claim to identify in real-time medical treatment to improve imbursements. Applicant then argues the claims recite an improvement of the functions of a computer by improving a computer’s recommendation for a healthcare treatment and improvement in technology of generating healthcare treatment recommendations based on machine learning models, and improvement in the technical field of medicine, medical reimbursement technology, and healthcare treatment recommendation technology. Examiner respectfully disagrees. Improving imbursements is not a technical problem, but again, a business practice/abstract idea improvement. And applying a machine learning model to patient data to generate a medical claim to identify a treatment is not a technological improvement. The technology (i.e., machine learning model) is being applied to the abstract idea (generating medical claim to identify real-time medical treatment). Also, improving a recommendation is not an improvement to the technology, but again, applying the prediction (abstract idea) to the machine learning (additional element/technology). Neither the machine learning model or the computer is improved when reimbursements are made, or when the amount of reimbursements increases or is improved. Collecting money owed or due has no bearing on the functionality of the computer or technology.
Applicant argues the claims do not recite a mental process, therefore, does not recite an abstract idea. Examiner respectfully disagrees. Estimating the characteristics of the predicted reimbursement for the target healthcare service is a step that can be done in the mind or pen and paper. MPEP 2106.04(a)(2) states “The courts do not distinguish between mental processes that are performed entirely in the human mind and mental processes that require a human to use a physical aid (e.g., pen and paper or a slide rule) to perform the claim limitation. See, e.g., Benson, 409 U.S. at 67, 65, 175 USPQ at 674-75, 674” and “Nor do the courts distinguish between claims that recite mental processes performed by humans and claims that recite mental processes performed on a computer. As the Federal Circuit has explained, “[c]ourts have examined claims that required the use of a computer and still found that the underlying, patent-ineligible invention could be performed via pen and paper or in a person’s mind.” Versata Dev. Group v. SAP Am., Inc., 793 F.3d 1306, 1335, 115 USPQ2d 1681, 1702 (Fed. Cir. 2015)”.
Applicant argues the claims are similar to Example #47, claim 3 and Examiner’s previous evaluation was in error because the Guidance identifies a process for determining when claims that recite machine learning models (such as neural networks) are patent eligible. Examiner respectfully disagrees. Example #47, claims are directed to a neural network that improves network intrusion detection, which is not the same a machine learning model used to identify, treat, or predict medical treatment and reimbursement. Furthermore, the Example #47, claim 3 is eligible because “steps (d)-(f) provide for improved network security using the information from the detection to enhance security by taking proactive measures to remediate the danger by detecting the source address associated with the potentially malicious packets. Specifically, the claim reflects the improvement in step (d), dropping potentially malicious packets in step (e), and blocking future traffic from the source address in step (f). These steps reflect the improvement described in the background” (See MPEP 2106.04(d)(1) and 2106.05(a)). The improvement in Example #47, claim 3 was to the functioning of the computer itself, which is not similar to the alleged improvement in the present application.
Applicant argues determining medical treatment plan that improve reimbursement results by combining a reimbursement prediction generated by a first model with a key features prediction generated by a second model is an improvement in the technological field of determining medical treatment plans. Examiner respectfully disagrees. Determining a medical treatment plan to improve reimbursement results… generated by models is not a technological improvement, but again, a business practice improvement. The method/abstract idea of determining a medical treatment plan is being applied to the model/technology. MPEP 2106.05(f) recites “The recitation of claim limitations that attempt to cover any solution to an identified problem with no restriction on how the result is accomplished and no description of the mechanism for accomplishing the result, does not integrate a judicial exception into a practical application or provide significantly more because this type of recitation is equivalent to the words “apply it”. See Electric Power Group, LLC v. Alstom, S.A., 830 F.3d 1350, 1356, 119 USPQ2d 1739, 1743-44 (Fed. Cir. 2016); Intellectual Ventures I v. Symantec, 838 F.3d 1307, 1327, 120 USPQ2d 1353, 1366 (Fed. Cir. 2016); Internet Patents Corp. v. Active Network, Inc., 790 F.3d 1343, 1348, 115 USPQ2d 1414, 1417 (Fed. Cir. 2015)” and “A claim having broad applicability across many fields of endeavor may not provide meaningful limitations that integrate a judicial exception into a practical application or amount to significantly more. For instance, a claim that generically recites an effect of the judicial exception or claims every mode of accomplishing that effect, amounts to a claim that is merely adding the words “apply it” to the judicial exception. See Internet Patents Corporation v. Active Network, Inc., 790 F.3d 1343, 1348, 115 USPQ2d 1414, 1418 (Fed. Cir. 2015)”.
Applicant argues the claims recite additional elements that amount to significantly more. Applicant further argues the claims recite an unconventional technological solution to the technological problem by applying a first machine learning model to patient data to determine reimbursement information, applying the output from the first model to a second model to identify key features for the first model, and then using the combined outputs of the models to identify a treatment to include a treatment plan that improves a reimbursement result. Examiner respectfully disagrees. Using multiple models does not amount to significantly more, but again, merely applies the abstract idea to the additional elements/technology (see MPEP 2106.05(h) recites “For claim limitations that generally link the use of the judicial exception to a particular technological environment or field of use, examiners should explain in an eligibility rejection why they do not meaningfully limit the claim. For example, an examiner could explain that employing generic computer functions to execute an abstract idea, even when limiting the use of the idea to one particular environment, does not add significantly more, similar to how limiting the abstract idea in Flook to petrochemical and oil-refining industries was insufficient”).
Applicant argues the claims are not well-understood, routine, and conventional by the prior art reference, by a court decision, or by another publication. Examiner respectfully disagrees. The current application has a current U.S.C. 103 Rejection and references to the MPEP and several court cases that show “receiving or transmitting data over a network” is not 101 eligible. Additionally, even if there wasn’t a prior art rejection or if it was overcome, MPEP 2106.05 recites “Specifically, lack of novelty under 35 U.S.C. 102 or obviousness under 35 U.S.C. 103 of a claimed invention does not necessarily indicate that additional elements are well-understood, routine, conventional elements. Because they are separate and distinct requirements from eligibility, patentability of the claimed invention under 35 U.S.C. 102 and 103 with respect to the prior art is neither required for, nor a guarantee of, patent eligibility under 35 U.S.C. 101”. Also, even if the prior art rejection was overcome, the rejection clearly articulates (a) citation a statement made by an applicant during prosecution that demonstrates the well-understood, routine, conventional nature of the additional element(s); (b) A citation to one or more of the court decisions discussed in Subsection II below as noting the well-understood, routine, conventional nature of the additional element(s); (c) A citation to a publication that demonstrates the well-understood, routine, conventional nature of the additional element(s); and (d) A statement that the examiner is taking official notice of the well-understood, routine, conventional nature of the additional element(s) (see MPEP 2106.05(d)).
Applicant argues the limitations of “identifying a treatment based on the combination of a machine-learning prediction of a reimbursement and machine-learning prediction of key features used to generate the first prediction, and generating a real-time recommendation to a healthcare provider that improves a reimbursement result for a treatment plan”, goes beyond generally linking any abstract idea to a particular technological environment. Examiner respectfully disagrees. The limitation of “identifying a treatment based on the combination of a machine-learning prediction of a reimbursement and machine-learning prediction of key features used to generate the first prediction, and generating a real-time recommendation to a healthcare provider that improves a reimbursement result for a treatment plan”, is part of the abstract idea and the abstract idea cannot integrate itself into a practical application. Therefore, the 35 U.S.C. 101 Rejection is maintained.
Regarding the 35 U.S.C. 102 and 35 U.S.C. 103 Rejection, Applicant argues Bostic does not teach the limitations of “applying the insurance reimbursement score to generate a recommendation for a treatment to include in a patient treatment plan that would improve a reimbursement result” and “generating a healthcare treatment recommendation based on the combination of a reimbursement prediction generated by a first model and a key features prediction generated by a second model. The combination of the reimbursement prediction and the key features prediction ensures any healthcare treatment recommendation is grounded in key features that affect reimbursement outcomes for healthcare treatments” in independent claims 1, 13, and 20. Examiner respectfully disagrees. Firstly, the claims do not recite an insurance reimbursement, and even if it did, the prior art reference, Bostic, explicitly relates to insurance reimbursement. Additionally, Bostic discloses at [0041] “the method includes monitoring insurance billing events, that includes ingesting patient data received from one of a plurality of patient data providers and healthcare services data relating to the patient data; ingesting data relating to insurance reimbursement criteria and insurance reimbursement records relating to the healthcare services data; determining one or more relationships between the ingested patient data, healthcare services data, insurance reimbursement criteria and insurance reimbursement records, and data and previously ingested patient data, healthcare services data, insurance reimbursement criteria and insurance reimbursement records wherein at least one new enriched data set is created based on the determined one or more relationships; transmitting the enriched data set to an analytic engine; using the analytic engine to calculate an insurance reimbursement score, wherein the insurance reimbursement score is based at least in part on an association between the healthcare services data and insurance reimbursement records; and using the analytic engine to calculate an insurance reimbursement score for a future planned health service event based at least in part on a comparison to the plurality of calculated insurance reimbursement scores” and [0115] recites “the method includes determining, using a machine learning module, whether providing a first treatment to said patient and/or said population of patients rather than providing a second treatment to said patient and/or said population of patients may result in an improved return on investment metric. In embodiments, the method includes determining an effect of a pre-existing condition on the return on investment metric of one of said patient and said population of patients” (e.g., “applying the insurance reimbursement score to generate a recommendation for a treatment to include in a patient treatment plan that would improve a reimbursement result”). Zahora teaches at [0094]: “applying machine learning algorithms and data analytics to historic information, medical providers may be supplied with complex predictive intelligence, such as dependable revenue stream estimates and revenue forecasting. Thus, systems and methods described herein provide a technical solution to the technical problem of reliably predicting revenue streams from an abundance of payer sources each having different reimbursement levels and requirements for coverage eligibility. Further, the inventors recognized that in combining data sources and applying machine learning analysis, the resultant data analytics may support real-time or near-real-time decision making that had not previously been possible, such as rapid patient pre-approval for medical procedures and prescription products as well as complex dynamic scheduling of mobile medical services. Thus, systems and methods described herein additionally provide a technical solution to the technical problem of overcoming resource availability obstacles and reimbursement obstacles to automatically deploy resources to patients in need in a timely manner, thereby improving patient outcomes while improving the bottom line” (e.g., “generating a healthcare treatment recommendation based on the combination of a reimbursement prediction generated by a first model and a key features prediction generated by a second model. The combination of the reimbursement prediction and the key features prediction ensures any healthcare treatment recommendation is grounded in key features that affect reimbursement outcomes for healthcare treatments”).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to RACHAEL SOJIN STONE whose telephone number is (571)272-8798. The examiner can normally be reached Monday-Friday 7 AM - 7 PM (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, Peter Choi can be reached at (469) 295-9171. 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.
/R.S.S./Examiner, Art Unit 3681
/PETER H CHOI/Supervisory Patent Examiner, Art Unit 3681
Prosecution Insights