DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Priority
Claim of priority to provisional patent application 63/484676 is acknowledged.
Status of the Claims
Claims 1-20 are currently pending and have been considered below.
Drawings
The drawings are objected to because some of the text in Figs. 3B, 6-8, 10, and 12A-16K is too small to be legible. 37 C.F.R. 1.84(p)(3) specifies that numbers, letters, and reference characters must measure at least .32 cm (1/8 inch) in height. Corrected drawing sheets in compliance with 37 CFR 1.121(d) are required in reply to the Office action to avoid abandonment of the application. Any amended replacement drawing sheet should include all of the figures appearing on the immediate prior version of the sheet, even if only one figure is being amended. The figure or figure number of an amended drawing should not be labeled as “amended.” If a drawing figure is to be canceled, the appropriate figure must be removed from the replacement sheet, and where necessary, the remaining figures must be renumbered and appropriate changes made to the brief description of the several views of the drawings for consistency. Additional replacement sheets may be necessary to show the renumbering of the remaining figures. Each drawing sheet submitted after the filing date of an application must be labeled in the top margin as either “Replacement Sheet” or “New Sheet” pursuant to 37 CFR 1.121(d). If the changes are not accepted by the examiner, the applicant will be notified and informed of any required corrective action in the next Office action. The objection to the drawings will not be held in abeyance.
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 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1
In the instant case, claims 1-11 are directed to a method (i.e. a process) and claims 12-20 are directed to a system (i.e. a 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 and 12 recite steps that, under their broadest reasonable interpretations, cover certain methods of organizing human activity, e.g. managing personal behavior, relationships, or interactions between people. Specifically, claim 1 recites:
A method of evaluating outcomes of a hematopoietic stem cell transplant (HSCT) and other cellular therapies for a patient, the method comprising:
receiving a set of patient characteristics;
defining one or more treatment options;
generating data based on the set of patient characteristics and one or more treatment options, including at least one actionable variable;
determining a risk score associated with one or more defined outcomes, based on the data;
evaluating an impact on the outcome associated with at least one actionable variable;
characterizing the impact of at least one actionable variable on the outcome over a range of associated values associated with at least one actionable variable to produce a patient set of treatment options based on one or more treatment options;
identifying a nondominated option from among the patient set of treatment options; and
generating a visualization of the nondominated option.
The italicized functions, when considered as a whole, describe a patient treatment planning operation that could be achieved by a human actor such as a clinician or other medical professional managing their personal behavior and/or interactions with others. For example, a patient could provide information (e.g. characteristics) about themselves to a clinician during an appointment or other interaction, and the clinician might use their medical expertise to come up with a set of potential treatment options, then use the patient characteristics to generate certain adjustable parameters (i.e. actionable variables) related to different treatment options (e.g. age of a stem cell donor, blood type of a blood donor, medication type or dosage, etc.). The clinician could then use their medical expertise to calculate risk scores for different outcomes, and evaluate/characterize the impact on the outcomes caused by changing the adjustable parameter over a given range of values. Finally, the clinician could identify an optimal treatment option based on their previous analysis, and generate a visualization (e.g. a report, chart, graphical representation, etc.) of the identified treatment option for sharing with the patient. Accordingly, claim 1 recites an abstract idea in the form of a certain method of organizing human activity.
Similarly, claim 12 recites:
A system for evaluating one or more outcomes of a hematopoietic stem cell transplant (HSCT) for a patient, the system comprising: a processor; a non-transitory memory in communication with the processor and storing instructions that, when executed, cause the processor to generate:
a risk assessor including one or more machine learning models, each of the one or more machine learning models associated with an outcome and configured to generate a risk score associated with the outcome based on a patient dataset, wherein the patient dataset is associated with a set of treatment options and includes at least one actionable variable; and
a trade-off analyzer including a multi-objective optimization framework, wherein the trade-off analyzer uses the multi-object optimization framework to: characterize an impact on the risk score associated with the actionable variable across a range of values for the actionable variable; and identify a non-dominated treatment option based on the set of treatment options, the range of values, and the impact.
But for the recitation of generic computer components like a processor executing instructions stored in a non-transitory memory and high-level machine learning, the italicized functions, when considered as a whole, describe a patient treatment planning operation that could be achieved by a human actor such as a clinician or other medical professional managing their personal behavior and/or interactions with others. For example, a clinician could generate various risk scoring models each associated with a clinical outcome and use the models to generate risk scores associated with each outcome based on a patient dataset associated with a set of treatment options and including at least one adjustable parameter (as described above). The clinician could also come up with a multi-objective optimization framework and use the framework to evaluate/characterize the impact on the outcomes caused by changing the adjustable parameter over a given range of values and identify an optimal treatment option for the patient based on the framework analysis. Accordingly, claim 12 recites an abstract idea in the form of a certain method of organizing human activity.
Dependent claims 2-11 and 13-20 inherit the limitations that recite an abstract idea from their dependence on claims 1 and 12, respectively, and thus these claims also recite an abstract idea under the Step 2A – Prong 1 analysis. In addition, claims 2-11 and 13-20 recite additional limitations that further describe the abstract idea identified in the independent claims.
Specifically, claim 2 specifies that the range of values is set according to the one or more treatment options, which a clinician could achieve by assigning an appropriate range of values to the actionable variable based on their knowledge of the treatment options under consideration.
Claim 3 recites training a model to determine the risk score associated with each outcome of the one or more outcomes, which a clinician could achieve by fitting data to an outcome prediction model for use in determining a risk score for each outcome.
Claim 4 specifies that the model is a survival outcome model or a competing risks outcome model, each of which are types of statistical models that a clinician would be capable of fitting data to and using for outcome risk prediction.
Claim 5 recites training two or more models for at least one outcome of the one or more outcomes, and training an ensemble model for the at least one outcome based on output generated using the two or more models. A clinician could accomplish these functions by fitting data to at least two different kinds of models (e.g. a regression model and a decision tree model) and then using the output of those models to fit an ensemble model (e.g. a weighted summation or voting schema).
Claim 6 specifies that characterizing the impact of the at least one actionable variable on the outcome over a range of values comprises performing a Pareto optimization, which is a mathematical operation and thus fits into the “mathematical concepts” category of abstract idea.
Claim 7 specifies that identifying the nondominated option is based on the Pareto optimization, which a clinician could achieve by using the output of the mathematical Pareto optimization process as a basis for identifying the treatment.
Claim 8 recites generating a self-organizing map based on the patient dataset and the risk score, which is a mathematical process that involves mapping an input space to a lower-dimensional output space to organize data, such that it fits into the “mathematical concepts” category of abstract idea.
Claim 9 specifies that identifying the nondominated option is based on the self-organizing map, which a clinician could achieve by using the output of the mathematical self-organizing map generation process as a basis for identifying the treatment.
Claim 10 recites that the impact is evaluated using one or more of variable importance scores, dependence analysis, and exponentiated coefficient estimates, each of which are mathematical operations that a human actor such as a clinician could perform and use as a basis for evaluating the impact of differences in values of an actionable variable on risk of a clinical outcome.
Claim 11 specifies that the outcome is one or more of overall survival, effect free survival, rejection, relapse, or graft-versus-host disease, each of which are types of outcomes about which a clinician would be capable of performing analysis.
Claim 13 specifies that the multi-objective optimization framework comprises a Pareto optimization, which is a mathematical operation and thus fits into the “mathematical concepts” category of abstract idea.
Claim 14 specifies that the multi-objective optimization framework comprises a self-organizing map, which is a mathematical operation and thus fits into the “mathematical concepts” category of abstract idea.
Claim 15 recites receiving the range of values for the actionable variable, which could be achieved as part of a certain method of organizing human activity by a clinician thinking about which values of the actionable variable are desired to be evaluated, communicating with a colleague to obtain a preset range of values, etc.
Claim 16 specifies that the actionable variable is associated with a HSCT donor characteristic, which is a type of parameter that a clinician would be capable of evaluating over a range of values to make predictions about risk of different outcomes.
Claim 17 specifies that the patient dataset comprises a plurality of predictor variables, which are types of data that a clinician would be capable of analyzing.
Claim 18 recites evaluating an impact value on the risk score for each predictor variable of the plurality of predictor variables, which a clinician would be capable of achieving by making an impact evaluation for each of at least two predictor variables in a dataset.
Claim 19 specifies that the outcome comprises a first outcome and a second outcome, and that a first impact value and second impact value are evaluated for each of the first outcome and second outcome, respectively, for each predictor variable of the plurality of predictor variables, which a clinician would be capable of achieving by performing determination/analysis processes for each of two outcomes.
Claim 20 recites generating a visualization comparing the first impact value with the second impact value, which a clinician would be capable of achieving by drawing or otherwise creating a visual or graphical representation of the two impact values, e.g. in a graph, chart, written report, etc.
However, recitation of an abstract idea is not the end of the 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
The judicial exception is not integrated into a practical application. In particular, independent claims 1 and 12 do not include additional elements that integrate the abstract idea into a practical application. There are no additional elements recited in claim 1 beyond the abstract idea itself; however, for purposes of further eligibility analysis, Examiner will assume that the method is intended to be performed via computer hardware as indicated in paras. [0077]-[0083] of Applicant’s specification, such that claim 1 is considered to include the additional elements of a computer processor executing the various steps. The additional elements of claim 12 include a processor; a non-transitory memory in communication with the processor and storing instructions that, when executed, cause the processor to generate the risk assessor and trade-off analyzer; as well as the fact that the models are machine learning models. These additional elements, when considered in the context of each claim as a whole, merely serve to automate interactions that could be achieved by and among human actors (as described above), and thus amount to instructions to “apply” the abstract idea using generic computer components (see MPEP 2106.05(f)). For example, a clinician could interact with a patient to obtain characteristic data or other values for a dataset and perform various mathematical analyses on the data to determine risk scores for each outcome and characterize the impacts on the risk scores resulting from changing a certain adjustable input parameter associated with a treatment so that a most optimal treatment may be identified, and performing such analyses and determinations with high-level computing components like a processor and machine learning models merely digitizes and/or automates these otherwise-abstract functions such that they are implemented in an automated environment. Accordingly, claims 1 and 12 as a whole are each directed to an abstract idea without integration into a practical application.
The judicial exception recited in dependent claims 2-11 and 13-20 is also not integrated into a practical application under a similar analysis as above. The functions of claims 2-11, 13-14, and 16-20 are performed with the same additional elements introduced in the independent claims, without introducing any new additional elements of their own, and accordingly also amount to mere instructions to apply the abstract idea in the same manner explained above for these same additional elements. Claim 15 recites a user interface via which the trade-off analyzer can receive the range of values for the actionable variable, which also amounts to instructions to “apply” the abstract idea because the otherwise-abstract function of receiving a range of variables for an actionable variable is merely being digitized with a high-level computing interface such that this input is assigned electronically.
Accordingly, the additional elements of claims 1-20 do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. Claims 1-20 are directed to an abstract idea.
Step 2B
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional elements of a processor executing instructions stored in a non-transitory memory for performing the various steps/functions of the invention and specifying that the predictive models are machine learning models amount to mere instructions to apply the exception using generic computer components. As evidence of the generic nature of the above recited additional elements,
Examiner notes paras. [0077]-[0082] of Applicant’s specification, where an example computing system is disclosed in terms of generic processor and memory elements that one of ordinary skill in the art would recognize as known in the art, as well as paras. [0091]-[0093], disclosing many examples of known machine learning model types that may be utilized to implement the machine learning models of the invention such that one of ordinary skill in the art would understand that no improvements to any specific machine learning model architectures are being made, and many different known model types are instead utilized as a means to digitize and/or automate the data analysis functions. Examiner further notes that the user interface (as in claim 15) is described in para. [0158] in highly generic terms and as being “responsive to any known input feature such as a mouse, a touch screen, a keyboard, voice commands, etc.,” leading one of ordinary skill in the art to understand that the user interface may encompass any generic user interface that permits known user input types such that receiving data from a user is digitized.
Analyzing these additional elements as an ordered combination adds nothing that is not already present when considering the elements individually; the overall effect of the computer implementation, high-level machine learning, and user interface in combination is to digitize and/or automate a patient treatment planning operation that could otherwise be achieved as a certain method of organizing human activity. Thus, when considered as a whole and in combination, claims 1-20 are not patent eligible.
Claim Rejections - 35 USC § 103
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-7, 10-13, and 16-20 are rejected under 35 U.S.C. 103 as being unpatentable over Chang et al. (US 20250125055 A1) in view of Choi et al. (Reference U on the accompanying PTO-892).
Claim 1
Note: the language of the preamble specifying that the method evaluates outcomes of a hematopoietic stem cell transplant (HSCT) and other cellular therapies for a patient does not structurally or functionally limit the positively recited steps of the claim, such that this language is not considered patentably limiting. In other words, there is no indication in the positively recited steps of the claim that the treatments or outcomes must be specific to HSCT or other cellular therapies, and thus their invocation in the preamble amounts to specifying an intended use in this field of medicine. See MPEP 2111.02(II). However, in the interest of compact prosecution, this aspect of the claim will be addressed with prior art below.
Chang teaches a method of evaluating outcomes of (Chang abstract, Fig. 1, noting a method for recommending a patient treatment based on predicted outcomes of the treatments), the method comprising:
receiving a set of patient characteristics (Chang Fig. 1, [0026], noting the system receives patient feature information (i.e. characteristics));
defining one or more treatment options (Chang [0059]-[0060], noting the system can search for similar patients with different patterns of treatment options for use in predicting outcomes for each treatment option, considered equivalent to defining one or more potential treatment options);
generating data based on the set of patient characteristics and one or more treatment options, including at least one actionable variable (Chang [0090], noting a treatment variable (e.g. a continuous or discrete “actionable variable”) can be determined and varied (e.g. different dosages as in [0056]) when determining outcomes associated with each treatment level. See also [0065]-[0067], noting the time of treatment can be varied when determining outcomes for each treatment, such that time is also considered to be an “actionable variable”);
determining a risk score associated with one or more defined outcomes, based on the data (Chang Fig. 1, [0030], [0067], [0072]-[0076], [0079]-[0088], noting various methods of calculating risk of one or more different defined outcomes for different treatment options based on the data);
evaluating an impact on the outcome associated with at least one actionable variable; characterizing the impact of at least one actionable variable on the outcome over a range of associated values associated with at least one actionable variable to produce a patient set of treatment options based on one or more treatment options (Chang [0035]-[0037], [0072]-[0074], [0089]-[0091], noting a multi-objective optimization process is implemented to compare outcomes/impacts of different levels of a set of treatment options given at different times (i.e. over a range of values associated with at least one actionable variable));
identifying a nondominated option from among the patient set of treatment options (Chang Fig. 1, [0035]-[0037], noting the system identifies a nondominated option from the optimization process as the optimal treatment); and
generating a visualization of the nondominated option (Chang Fig. 1, [0038], noting the recommended treatment option is displayed to a user via a user interface, i.e. a visualization of the nondominated option is generated).
In summary, Chang teaches a method for predicting outcomes of different treatment options with different treatment variables and identifying a nondominated option based on analysis of the predicted outcomes. Chang does not appear to specifically limit the type of medical condition being evaluated or the type of corresponding different treatment options and outcomes, though it does provide an example embodiment specific to sepsis treatment (see [0056]). Accordingly, though this reference appears to contemplate broad application of the method to different types of medical conditions and corresponding medical treatments and outcomes, it fails to explicitly disclose evaluating outcomes specific to a hematopoietic stem cell transplant (HSCT) and other cellular therapies.
However, Choi teaches that cellular therapies like hematopoietic stem cell transplant are in need of improved systems for predicting HSCT-specific outcomes related to different treatment variables so that the most optimal treatment option can be proactively recommended for a patient (Choi Abstract on Pg 1, Introduction on Pg 2, Application of the Algorithm for Donor Selection on Pg 7, Discussion on Pg 8). It therefore would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to apply the non-specific treatment outcome prediction and recommendation method of Chang to the specific field of evaluating HSCT cellular therapy outcomes as in Choi in order to improve existing HSCT treatment planning methodologies so that better patient outcomes are achieved in this specific complication-prone field of medicine (as suggested by Choi Introduction on Pg 2 & Discussion on Pg 8).
Claim 2
Chang in view of Choi teaches the method of claim 1, and the combination further teaches wherein the range of values is set according to the one or more treatment options (Chang [0056], noting treatments of different dosages of a substance, considered equivalent to a range of values set according to the one or more treatment options).
Claim 3
Chang in view of Choi teaches the method of claim 1, and the combination further teaches training, for each outcome of the one or more outcomes, a machine learning model, to determine the risk score associated with the outcome (Chang [0083], [0088], [0092], noting a machine learning model is trained for each outcome).
Claim 4
Chang in view of Choi teaches the method of claim 3, and the combination further teaches wherein the machine learning model is a survival outcome model or a competing risks outcome model (Chang [0056], [0069], noting patient survival is a selected outcome for which a machine learning model can be trained as explained above for claim 3, such that a model trained to predict the outcome of survival would be a survival outcome model).
Claim 5
Chang in view of Choi teaches the method of claim 1, and the combination further teaches training for at least one outcome of the one or more outcomes, two or more machine learning models; (Chang [0083], [0088], [0092], noting machine learning models are trained for each of multiple outcomes, such that two or more machine learning models may be trained for two or more outcomes).
Though the present combination teaches training at least two machine learning models for at least two defined outcomes, it fails to explicitly disclose training for the at least one outcome an ensemble machine learning model based on output generated using the two or more machine learning models. However, Choi further teaches that a gradient boosting machine (GBM) learning algorithm, which is a type of ensemble model that combines the outputs of several classifiers, performed most optimally at outcome prediction for HSCT (Choi Selection of a Predicting Model on Pg 3). It therefore would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the multiple machine learning model training method of the combination to include training of an ensemble model as in Choi in order to utilize an outcome prediction method shown to outperform other machine learning methods (as suggested by Choi Selection of a Predicting Model on Pg 3)
Claim 6
Chang in view of Choi teaches the method of claim 1, and the combination further teaches wherein characterizing the impact of the at least one actionable variable on the outcome over a range of values comprises performing a Pareto optimization (Chang [0036], [0072]-[0076], [0079], [0089]-[0092]).
Claim 7
Chang in view of Choi teaches the method of claim 6, and the combination further teaches wherein identifying the nondominated option is based on the Pareto optimization (Chang [0036]-[0037]).
Claim 10
Chang in view of Choi teaches the method of claim 1, and the combination further contemplates evaluating outcome risk based on importance sampling methods (Chang [0081]). However, the present combination fails to explicitly disclose wherein the impact is evaluated using one or more of variable importance scores, dependence analysis, and exponentiated coefficient estimates. However, Choi further teaches that evaluating outcomes for different treatments so that the optimal treatment can be selected includes calculating scores reflecting the importance of each variable to the outcome (Choi Abstract on Pg 1, Application of the Algorithm for Donor Selection on Pg 7). It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the method of the combination to include evaluating the impact of each variable on the outcome using variable importance scores as in Choi in order to facilitate an explainable visualization of the impact of each variable on the outcome so that more appropriate treatments may be prioritized for selection (as suggested by Choi Application of the Algorithm for Donor Selection on Pg 7 & Discussion on Pg 8).
Claim 11
Chang in view of Choi teaches the method of claim 1, and the combination further teaches wherein the outcome is one or more of overall survival, effect free survival, rejection, relapse, or graft-versus-host disease (Chang [0056], [0069], noting an evaluated outcome can include patient survival or mortality, i.e. overall survival. See also Choi abstract, noting that outcomes evaluated in the context of HSCT can include overall death (i.e. overall survival) and relapse incidence).
Claim 12
Note: the language of the preamble specifying that the system evaluates outcomes of a hematopoietic stem cell transplant (HSCT) for a patient does not structurally or functionally limit the positively recited elements of the claim, such that this language is not considered patentably limiting. In other words, there is no indication in the positively recited limitations of the claim that the treatments or outcomes must be specific to HSCT, and thus its invocation in the preamble amounts to specifying an intended use in this field of medicine. See MPEP 2111.02(II). However, in the interest of compact prosecution, this aspect of the claim will be addressed with prior art below.
Chang teaches a system for evaluating one or more outcomes (Chang Figs. 1-2, [0025], [0042]-[0049] noting a system including a processor and memory executing instructions to implement various modules for recommending a patient treatment based on predicted outcomes of the treatments):
a risk assessor including one or more machine learning models, each of the one or more machine learning models associated with an outcome and configured to generate a risk score associated with the outcome based on a patient dataset, wherein the patient dataset is associated with a set of treatment options and includes at least one actionable variable (Chang Fig. 1, [0030], [0067], [0072]-[0076], [0079]-[0092], noting the system trains and stores at least one time-varying effect model including outcome-specific machine learning models that predict risk of respective outcomes for different treatment options over varying treatment variables (i.e. an actionable variable) based on patient data); and
a trade-off analyzer including a multi-objective optimization framework, wherein the trade-off analyzer uses the multi-object optimization framework to: characterize an impact on the risk score associated with the actionable variable across a range of values for the actionable variable; and identify a non-dominated treatment option based on the set of treatment options, the range of values, and the impact (Chang [0035]-[0037], [0072]-[0074], [0089]-[0091], noting the system implements a multi-objective optimization process (i.e. a trade-off analyzer) to compare outcomes/impacts of different levels of a set of treatment options given at different times (i.e. over a range of values associated with at least one actionable variable) and ultimately identifies a nondominated option from the optimization process as the optimal treatment).
In summary, Chang teaches a system for predicting outcomes of different treatment options with different treatment variables and identifying a nondominated option based on analysis of the predicted outcomes. Chang does not appear to specifically limit the type of medical condition being evaluated or the type of corresponding different treatment options and outcomes, though it does provide an example embodiment specific to sepsis treatment (see [0056]). Accordingly, though this reference appears to contemplate broad application of the system to different types of medical conditions and corresponding medical treatments and outcomes, it fails to explicitly disclose evaluating outcomes specific to a hematopoietic stem cell transplant (HSCT).
However, Choi teaches that the field of hematopoietic stem cell transplant is in need of improved systems for predicting HSCT-specific outcomes related to different treatment variables so that the most optimal treatment option can be proactively recommended for a patient (Choi Abstract on Pg 1, Introduction on Pg 2, Application of the Algorithm for Donor Selection on Pg 7, Discussion on Pg 8). It therefore would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to apply the non-specific treatment outcome prediction and recommendation system of Chang to the specific field of evaluating HSCT outcomes as in Choi in order to improve existing HSCT treatment planning methodologies so that better patient outcomes are achieved in this specific complication-prone field of medicine (as suggested by Choi Introduction on Pg 2 & Discussion on Pg 8).
Claim 13
Chang in view of Choi teaches the system of claim 12, and the combination further teaches wherein the multi-objective optimization framework comprises a Pareto optimization (Chang [0036], [0072]-[0076], [0079], [0089]-[0092]).
Claim 16
Chang in view of Choi teaches the system of claim 12, but the present combination fails to explicitly disclose wherein the actionable variable is associated with a HSCT donor characteristic. However, Choi further teaches that in the context of predicting and comparing HSCT-specific outcomes for different treatment options, donor characteristics may be the differing inputs for the different evaluated treatments (Choi Application of the Algorithm for Donor Selection on Pg 7, noting donor age and donor HLA type may be different actionable variables dictating different predicted outcomes for a recipient so that a most optimized treatment donor may be selected). It therefore would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the evaluated treatment variable of the combination to include HSCT donor characteristics as in Choi in order to evaluate impacts of this specific type of treatment variable which is known to be associated with HSCT outcomes (as suggested by Choi Introduction on Pg 2 & Principal Findings on Pg 7).
Claim 17
Chang in view of Choi teaches the system of claim 12, and the combination further teaches wherein the patient dataset comprises a plurality of predictor variables (Chang [0026], noting the system receives a plurality of patient features (i.e. a plurality of predictor variables) used to make the outcome and treatment predictions).
Claim 18
Chang in view of Choi teaches the system of claim 17, showing a system that characterizes impacts of different values of a treatment variable given a set of patient predictor variables on a predicted outcome risk. However, the present combination does not appear to describe evaluating a specific impact value for each predictor variable, and thus fails to explicitly disclose wherein the risk assessor evaluates, for each predictor variable of the plurality of predictor variables, an impact value on the risk score. However, Choi further teaches that evaluating outcomes for different treatments so that the optimal treatment can be selected includes calculating scores reflecting the importance of each predictor variable to the outcome (Choi Abstract on Pg 1, Application of the Algorithm for Donor Selection on Pg 7), considered equivalent to evaluating an impact value on the risk score for each predictor variable of a plurality of predictor variables. It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the method of the combination to include evaluating an impact value of each predictor variable on the predicted outcome as in Choi in order to facilitate an explainable visualization of the numerical value representing an impact of each variable on the outcome so that more appropriate treatments may be prioritized for selection (as suggested by Choi Application of the Algorithm for Donor Selection on Pg 7 & Discussion on Pg 8).
Claim 19
Chang in view of Choi teaches the system of claim 18, and the combination further teaches wherein the outcome comprises a first outcome and a second outcome; and wherein the risk assessor evaluates, for each predictor variable of the plurality of predictor variables, a first impact value associated with the first outcome and a second impact value associated with the second outcome (Chang [0003], noting multiple outcomes corresponding to different organ systems may be evaluated; see also Choi Application of the Algorithm for Donor Selection on Pg 7, noting impact values may be calculated for each predictor variable for a given outcome prediction. When considered together in the context of the combination explained above for claim 18, these teachings suggest that a first set of impact values could be calculated for a first outcome, and a second set of impact values could be calculated for a second outcome, such that each predictor variable would be associated with respective first and second impact values corresponding to their impacts on the first and second outcomes).
Claim 20
Chang in view of Choi teaches the system of claim 19, and the combination further teaches wherein the risk assessor generates a visualization comparing the first impact value with the second impact value (Chang [0038], [0055], noting a recommended treatment option, the input data for the patient (i.e. predictor variables), and other data is displayed to a user via a user interface, i.e. via a generated visualization; see also Choi Application of the Algorithm for Donor Selection on Pg 7, Discussion on Pg 8, noting calculated impact values for predictor variables may be visualized in two instances for comparison. When considered together in the context of the combination explained above for claims 18-19, these teachings suggest that the first and second sets of impact values corresponding to first and second predicted outcomes could be visualized for comparison as in Fig. 3 of Choi).
Claims 8-9 and 14 are rejected under 35 U.S.C. 103 as being unpatentable over Chang and Choi as applied to claims 1 or 12 above, and further in view of Ziegler et al. (US 20120109675 A1).
Claim 8
Chang in view of Choi teaches the method of claim 1, but the present combination fails to explicitly disclose generating a self-organizing map based on the patient dataset and the risk score. However, Ziegler teaches an analogous clinical therapy and outcome optimization method that includes generating a self-organizing map based on patient data and predicted outcomes (i.e. risk scores) for clinical treatments (Ziegler [0063]-[0064]). It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the treatment outcome evaluation method of the combination to include generating a self-organizing map as in Ziegler in order to identify correlations between patient data and treatment outcomes that can be used to evaluate prospective clinical therapies based on their likelihood of success (as suggested by Ziegler [0063]-[0064]).
Claim 9
Chang in view of Choi and Ziegler teaches the method of claim 8, and the combination further teaches wherein identifying the nondominated option is based on the self-organizing map (Chang [0036]-[0037], noting the system identifies a nondominated option from the optimization process as the optimal treatment; see also Ziegler [0063]-[0064], noting an optimal treatment is identified based on the self-organizing map. When considered together in the context of the combination explained above for claim 8, these teachings show that the optimal treatment would be identified based on the self-organizing map optimization process).
Claim 14
Chang in view of Choi teaches the system of claim 12, but the present combination fails to explicitly disclose wherein the trade-off analyzer further comprises a self-organizing map. However, Ziegler teaches an analogous clinical therapy and outcome optimization system that includes generating a self-organizing map for clinical treatments (Ziegler [0063]-[0064]). It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the treatment outcome evaluation system of the combination to include generating a self-organizing map as in Ziegler in order to identify correlations between patient data and treatment outcomes that can be used to evaluate prospective clinical therapies based on their likelihood of success (as suggested by Ziegler [0063]-[0064]).
Claim 15 is rejected under 35 U.S.C. 103 as being unpatentable over Chang and Choi as applied to claim 12 above, and further in view of Pappada (US 20150227710 A1).
Claim 15
Chang in view of Choi teaches the system of claim 12, and the combination further teaches a user interface, wherein the trade-off analyzer can receive, via the user interface, (Chang [0025], [0044], noting the system includes a user interface that allows the user to convey information to the system).
Though the present combination includes a user interface by which a user can input data to the system, it fails to explicitly disclose that the user interface is used to specifically input the range of values for the actionable variable. However, Pappada teaches an analogous treatment outcome evaluation system that allows a user to input a range of values for an actionable variable of a potential treatment plan using a user interface so that the impact of varying the actionable variable on the outcome of interest can be interactively visually conveyed (Pappada Figs. 2B-C, [0096]-[0097], [0101]). It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the user interface of the combination to include functionality for allowing a user to manually specify a range of values for an actionable variable during a treatment impact comparison as in Pappada in order to permit an intuitive simulation capability whereby clinical staff can vary any desired actionable variable input to visualize the predicted impact to a patient (as suggested by Pappada [0097] & [0101]).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Hill, JR. et al. (US 20210174962 A1) describes systems for using machine learning optimization models for identifying optimal blood transfusion treatment parameters. Nagar et al. (Reference V on the accompanying PTO-892) describes methods for generating self-organizing maps to visualize the results of multi-objective optimization problems.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to KAREN A HRANEK whose telephone number is (571)272-1679. The examiner can normally be reached M-F 8:00-4:00 ET.
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, Shahid Merchant can be reached at 571-270-1360. 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.
/KAREN A HRANEK/ Primary Examiner, Art Unit 3684