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 .
Claims 1-16 are pending and have been examined.
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 1-16 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Claims 1-16 are directed to a method, system or product, which are statutory categories of invention. (Step 1: YES).
The Examiner has identified method Claim 1 as the claim that represents the claimed invention for analysis and is similar to system Claim 14 and product Claim 15.
Claim 1 recites the limitations of:
A computer-implemented method for determining optimized patient symptom-mitigation predictions, the method comprising:
receiving profile information and user symptoms from a user;
matching the profile information and the user symptoms to one of a plurality of profile clusters that are within one or more symptom-mitigation clusters created using textual data, the profile clusters being grouped within the one or more symptom-mitigation clusters based on profile information of authors of the textual data; and
predicting a symptom-mitigation strategy based on the matching profile cluster
These above limitations, under their broadest reasonable interpretation, cover performance of the limitation as certain methods of organizing human activity. The claim recites elements, highlighted in bold above, which covers performance of the limitation as a managing personal behavior and managing relationships between people. Receiving profile information and user symptoms from a user (following rules/instructions), matching profile information and user symptoms to a plurality of profile clusters created using textual data based on profile information of authors of the textual data (social activities and following rules or instructions) and predicting a symptom-mitigation strategy (teaching) are managing personal behavior and relationships between people. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation as managing personal behavior or managing relationships between people, then it falls within the “Certain Methods of Organizing Human Activity” grouping of abstract ideas. Accordingly, the claim recites an abstract idea. Claims 14 and 15 are also abstract for similar reasons. (Step 2A-Prong 1: YES. The claims are abstract)
In as much as a person in their mind and/or with pen and paper can perform the steps, the claim is also abstract as a mental process. A person can receive (read) profile information and user symptoms, a person can match (in their mind or with pen and paper) profile information and user symptoms to a plurality of profile clusters, a person can predict a symptom-mitigation strategy based on matching (analyze the matched data to predict). Also, using a computer has been shown to fall under mental processes grouping of abstract ideas (see MPEP 2106.04(a)(2) III C)
This judicial exception is not integrated into a practical application. In particular, the claims only recite: computer (Claim 1); computer system, hardware processors (Claim 14); non-transitory computer readable medium, processor (Claim 15). The computer hardware is recited at a high-level of generality (i.e., as a generic processor performing a generic computer function) such that it amounts no more than mere instructions to apply the exception using a generic computer component. Accordingly, these additional elements, when considered separately and as an ordered combination, do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. Therefore claims 1, 14, and 15 are directed to an abstract idea without a practical application. (Step 2A-Prong 2: NO. The additional claimed elements are not integrated into a practical application)
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception because, when considered separately and as an ordered combination, they do not add significantly more (also known as an “inventive concept”) to the exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional element of using a computer hardware amounts to no more than mere instructions to apply the exception using a generic computer component. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. Accordingly, these additional elements, when considered separately and as an ordered combination, do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. Steps such as receiving are steps that are considered insignificant extra solution activity and mere instructions to apply the exception using general computer components (see MPEP 2106.05(d), II). Thus claims 1, 14, and 15 are not patent eligible. (Step 2B: NO. The claims do not provide significantly more)
Dependent claims 2-13 and 16 further define the abstract idea that is present in independent claim 1 and thus correspond to Certain Methods of Organizing Human Activity and Mental Processes and hence are abstract for the reasons presented above. The dependent claims do not include any additional elements that integrate the abstract idea into a practical application or are sufficient to amount to significantly more than the judicial exception when considered both individually and as an ordered combination. Claims 2, 3, and 4 recite a generic database at a high level of generality. Claim 3 recites a clustering algorithm, which is a mathematical concept and abstract for that reason also. Claim 7 recites averaging a distance which is abstract as a mathematical concept. Claim 16 recites machine learning or artificial intelligence system which are generic systems claimed at a high level of generality. Therefore, the claims 2-13 and 16 are directed to an abstract idea. Thus, the claims 1-16 are not patent-eligible.
Examiner Request
The Applicant is requested to indicate where in the specification there is support for amendments to claims should Applicant amend. The purpose of this is to reduce potential 35 U.S.C. §112(a) or §112 1st paragraph issues that can arise when claims are amended without support in the specification. The Examiner thanks the Applicant in advance.
Claim Rejections - 35 USC § 103
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 (i.e., changing from AIA to pre-AIA ) 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.
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.
Claims 1-4 and 7-10 are rejected under 35 U.S.C. 103 as being unpatentable over Pub. No. US 2022/0406405 to Durham et al. in view of Pub. No. US 2020/0335225 to Quintero Padron et al.
Regarding claims 1, 14, and 15
(claim 1) A computer-implemented method for determining optimized patient symptom-mitigation predictions, the method comprising:
receiving profile information and user symptoms from a user;
Durham et al. teaches:
Fig. 1, ref. 106 (Data Sources) and 114 (Co-Occurring Conditions, therefore symptoms)..
PNG
media_image1.png
446
566
media_image1.png
Greyscale
Fig. 1, ref. 128 and profiles…
“The one or more characterization systems 128 can determine one or more profiles for individuals. In one or more implementations, the one or more profiles can be used to determine amounts of similarity between features of the individuals and features associated with one or more classifications. The one or more profiles can include genetic profiles, profiles of morphological features of individuals, profiles of levels of analytes present in individuals, profiles of electrical activity data of individuals, profiles of imaging data of individuals, or one or more combinations thereof. The one or more profiles determined by the one or more characterization systems 128 can be generated using information obtained from the one or more data sources 106.” [0048]
matching the profile information and the user symptoms to one of a plurality of profile clusters that are within one or more symptom-mitigation clusters created using textual data, the profile clusters being grouped within the one or more symptom-mitigation clusters based on profile information of authors of the textual data; and
Condition (symptom) with treatment (mitigation), that is group of individuals (cluster)…
“In the illustrative example of FIG. 1, the one or more therapeutic repositioning systems 134 can determine that at least a first therapeutic 136 can be a recommendation to treat a neurodevelopmental condition with respect to the first group of individuals 130. Additionally, the one or more therapeutic repositioning systems 134 can determine that at least an Nth therapeutic 138 can be a recommendation to treat a neurodevelopmental condition with respect to the Nth group of individuals 132. In various examples, the first therapeutic 136 and the Nth therapeutic 138 can be recommended as treatments for the same neurodevelopmental condition. However, due to different features being associated with the first group of individuals 130 and the Nth group of individuals 132, the first therapeutic 136 can have a greater probability of reducing symptoms of the neurodevelopmental condition with respect to the first group of individuals 130 than the Nth therapeutic 138. Additionally, the Nth therapeutic 138 can have a greater probability of reducing symptoms of the neurodevelopmental condition with respect to the second group of individuals 132 than the first therapeutic 136 based on differences between the features of the Nth group of individuals 132 and the first group of individuals 130.” [0056]
Correlation (matching) based on features of individuals (profile), where condition is present based on classification of individuals…
“The computational biology system 102 can include one or more classification systems 118. The one or more classification systems 118 can determine classifications in which to group individuals in which a neurodevelopmental condition is present. The one or more classification systems 118 can analyze data obtained from the one or more data sources 106 and determine features of individuals in which a neurodevelopmental condition is present that can be used in the classification of the individuals. In one or more implementations, the one or more classification systems 118 can implement at least one of one or more machine learning techniques or one or more statistical techniques to determine amounts of correlation between features of individuals in which a neurodevelopmental condition is present. To illustrate, the one or more classification systems 118 can determine features related to classification of individuals in which a neurodevelopmental condition is present based on the features having at least a threshold amount of correlation with respect to one another. The threshold amount of correlation can be based on a number of the individuals in which a neurodevelopmental condition is present that exhibit the features in relation to a broader group of individuals in which the neurodevelopmental condition is present. In various examples, the one or more classification systems 118 can implement one or more combinations of multiple machine learning techniques and/or one or more combinations of statistical techniques to determine features to include in each of a number of classifications of individuals in which a neurodevelopmental condition is present.” [0044]
Example of features…
“The implementations described herein are directed to developing frameworks, architectures, systems, methods, and techniques for characterizing individuals diagnosed with neurodevelopmental conditions according to characteristics that are not merely behavioral characteristics. The implementations described herein can include a computational biology system that identifies subgroups of individuals in which a neurodevelopmental condition is present. The individuals included in each sub-group can have a common set of features that can include at least one or more morphological features, one or more genetic features, one or more co-occurring conditions, disruption of one or more biological pathways, one or more metabolic features, levels of one or more analytes, or one or more combinations thereof. The characteristics of the individuals included in each sub-group can also be predictive of the responsiveness of the individuals to one or more treatments for a given neurodevelopmental condition.” [0031]
Assess how close (mathing) features (text of users symptoms) and individuals (profile from the user), therefore combined, are to the centroid…
“The reliability of the groups of features 312, 314, 316 and the subgroups of individuals in which a neurodevelopmental condition is present that are identified according to the groups of features 312, 314, 316 can be evaluated by the computational biology system 102. In various implementations, for an example subgrouping of individuals, the overall similarity within subgroups compared to the similarity within other subgroups can be measured using one or more validity indices, such as Dunn Index, Davies-Boulding index and/or Silhouette index. For example, the groups of features 312, 314, 316 and/or the subgroups of individuals generated using the groups of features 312, 314, 316 can be evaluated using a Davies-Bouldin index, assessing how close the members in a cluster are to the centroid of the cluster and how distant the centroids of different clusters are. In various examples, the features included in the group of features 312, 314, 316 can be identified based on Davies-Bouldin index values for the respective groups of features 312, 314, 316 being at most threshold Davies-Bouldin index values. Additionally, the groups of features 312, 314, 316 and/or the sub-groups of individuals generated using the groups of features 312, 314, 316 can be evaluated using a Silhouette index, assessing how close the members of the same cluster are to each other and how close they are to the members of other clusters…” [0095]
A distribution of distances of the therapeutics and determine candidate (most common) therapeutics…
“In one or more implementations, the computational biology system 102 can determine one or more nodes of the gene expression profile distance network 1114 having a minimum distance with respect to an inverted consensus gene expression profile 1148 or having less than a threshold distance with respect to an inverted one of the consensus gene expression profiles 1148. The nodes having less than the threshold distance can correspond to a candidate therapeutic for individuals included in the subgroup corresponding to one of the consensus gene expression profiles 1148. In various examples, the computational biology system 102 can determine a distribution of pairwise distances, such as profile similarity scores, of the therapeutics included in the gene expression profile distance network 1114 to determine a mean of the distribution and a standard deviation for the distribution. The computational biology system 102 can determine one or more candidate therapeutics corresponding to a consensus gene expression profile 1148 based on a threshold number of standard deviations from the mean for a particular pair of therapeutics having nodes included in the gene expression profile distance network 1114 in relation to an inverted consensus gene expression profile 1148. The threshold number of standard deviations can include no greater than about 3 standard deviations, no greater than about 2.5 standard deviations, no greater than about 2 standard deviations, no greater than 1.5 standard deviations, no greater than about 1 standard deviation, or no greater than about 0.5 standard deviations.” [0194]
See Authors below.
predicting a symptom-mitigation strategy based on the matching profile cluster.
Greater probability (predicting) decreasing symptoms (symptom-mitigation based on grouping of individuals…
“…Additionally, by grouping the individuals in which a neurodevelopmental condition is present according to implementations described herein, the computational biology system can have a higher probability of identifying therapeutics that can be used to effectively treat the neurodevelopmental condition. In particular, the computational biology system can determine subgroups for individuals in which a neurodevelopmental condition is present based on the predisposition of the individuals to respond to one or more therapeutics. Accordingly, the recommendations for candidate treatments determined by the computational biology system for individuals within a particular grouping can have a greater probability of decreasing the symptoms of the individuals with respect to the neurodevelopmental condition than the treatments of the neurodevelopmental condition that are determined according to conventional methods and techniques.” [0035]
Authors
Durham et al. teaches data with symptoms and therapies. They do not teach authors.
Quintero Padron et al. also in the business of symptoms and therapies teaches:
Social media with symptoms and therapies, and posts by a person (therefore author)…
“This social media analytics module displays all public posts in social media related to a specific disease, for example related to “Sjögren's syndrome”, for example on Facebook (registered trademark) and or on Twitter (registered trademark). This social media analytics module is accessible for researchers 41 and clinicians 42 and allows the user to include “free text” in order to refine the list of posts by adding time, location, symptoms, therapies, etc . . . , but only if the public post contains this information. Embedded Posts are a simple way to put public posts, by a page or by a person, into “HarmonicSS” platform. Only public posts are embedded. To access Facebook (registered trademark) post, an application programming interface is used to get data out of, and put data into, Facebook's platform (registered trademark). A low-level HTTP-based application programming interface is used to programmatically query data, post new stories, manage ads, upload photos, and perform other usual tasks that an application programming interface may usually implement. To access to Twitter (registered trademark) post, there is not a dedicated functionality, so this is worked around, for example, by using “HarmonicSS” Twitter (registered trademark) user and get mentions # hashtags, like # Sjögren for example.” [0073]
“All users, including patients 43, researchers 41 and clinicians 42, have all user access 10, authorizing them access to social media analytics function 25 in passive mode by consulting newly posted information on it, as well as to notifications and reporting function 26.” [0083]
It would have been obvious to one of ordinary skill in the art before the effective filing date to include in the method and system of Durham et al. the ability to use social media as taught by Quintero Padron et al. since the claimed invention is merely a combination of old elements and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. Further motivation is provided by Quintero Padron et al. who teaches the advantages of using post written by persons for finding symptom and related therapies. Durham benefits as they also are interested in finding information about symptoms and therapies from data sources.
Regarding claim 2
The computer-implemented method of claim 1, further comprising obtaining the textual data at least in part from social media posts and/or crowd-sourced data, and
Durham et al. teaches:
Obtaining public and private data…
“FIG. 2 is a diagram illustrating an example framework 200 to generate a number of curated databases 202 using information obtained from a number of data sources 204. At least a portion of the number of data sources 204 can include publicly accessible information. Information can be publicly accessible when in can be obtained without a specified form of authentication and/or credentials. In various examples, publicly accessible information can be obtained via one or more websites and/or via a physical repository of information that is accessible to the public. Publicly accessible information can be included in publications, such as research articles, medical literature, clinical trials information, one or more combinations thereof, and the like. In addition, at least a portion of the number of data sources 204 can store privately accessible information. In one or more implementations, the privately accessible information can be obtained using specified authentication information and/or credentials. Privately accessible information can be obtained electronically or from a physical location that has restricted access.” [0058]
Web crawlers to obtain data from websites and extract text…
“The framework 200 can include the computational biology system 102. The computational biology system 102 can perform operations to obtain data from the number of data sources 204 and to generate the number of curated databases 202. For example, at operation 214, the computational biology system 102 can obtain data from the one or more data sources 204. To illustrate, the computational biology system 102 can utilize one or more web crawlers to obtain data from one or more websites included in the number of data sources 204. Additionally, the computational biology system 102 can extract text from one or more documents. The one or more documents can include electronic documents. The one or more documents can also include physical documents. In various examples, the computational biology system 102 can perform at least one of optical character recognition (OCR) operations or natural language process operations to extract text information from documents included in the number of data sources 204. In one or more implementations, information can be extracted from the one or more data sources 204 in relation to one or more search criteria provided by the computational biology system 102, such as at least one of one or more keywords, one or more categories, or one or more classifications. Information can be sent from at least one of the data sources 206, 208, 210, 212 to the computational biology system 102 in response to one or more requests for information provided by the computational biology system 102. Information can also be obtained from the one or more data sources 204 through manual entry of the information via at least one of one or more user interfaces or one or more input devices of one or more computing devices that are coupled to and/or in electronic communication with the computational biology system 102.” [0064]
building a profile database including the profile information about the authors of the social media posts and/or the crowd-sourced data.
Determine (build) profiles….
“The one or more characterization systems 128 can determine one or more profiles for individuals. In one or more implementations, the one or more profiles can be used to determine amounts of similarity between features of the individuals and features associated with one or more classifications. The one or more profiles can include genetic profiles, profiles of morphological features of individuals, profiles of levels of analytes present in individuals, profiles of electrical activity data of individuals, profiles of imaging data of individuals, or one or more combinations thereof. The one or more profiles determined by the one or more characterization systems 128 can be generated using information obtained from the one or more data sources 106.” [0048]
One example of profile database…
“… The one or more data stores can include one or more databases that are electronically accessible via a network. In one or more illustrative examples, the computational biology system 102 can obtain the gene expression profile 1302 from a database that is accessible via a website.” [0216]
The combined references teach social media. They do not teach social media posts.
Quintero Padron et al. also in the business of social media teaches:
Social media posts and related to (classifying) disease (content)….
“This social media analytics module displays all public posts in social media related to a specific disease, for example related to “Sjögren's syndrome”, for example on Facebook (registered trademark) and or on Twitter (registered trademark). This social media analytics module is accessible for researchers 41 and clinicians 42 and allows the user to include “free text” in order to refine the list of posts by adding time, location, symptoms, therapies, etc . . . , but only if the public post contains this information. Embedded Posts are a simple way to put public posts, by a page or by a person, into “HarmonicSS” platform. Only public posts are embedded. To access Facebook (registered trademark) post, an application programming interface is used to get data out of, and put data into, Facebook's platform (registered trademark). A low-level HTTP-based application programming interface is used to programmatically query data, post new stories, manage ads, upload photos, and perform other usual tasks that an application programming interface may usually implement. To access to Twitter (registered trademark) post, there is not a dedicated functionality, so this is worked around, for example, by using “HarmonicSS” Twitter (registered trademark) user and get mentions # hashtags, like # Sjögren for example.” [0073]
It would have been obvious to one of ordinary skill in the art before the effective filing date to include in the method and system of the combined references the ability to use social media posts as taught by Quintero Padron et al. since the claimed invention is merely a combination of old elements and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. Further motivation is provided by Quintero Padron et al. who teaches the advantages of using post written by persons for finding symptom and related therapies. The combined references benefit as they also are interested in finding information about symptoms and therapies from data sources.
Regarding claim 3
The computer-implemented method according to claim 1, further comprising:
building a symptom-mitigation database by extracting symptoms and mitigation strategies from the textual data, wherein the one or more symptom-mitigation clusters are generated using a clustering algorithm, the profile information from the profile database, and the symptom-mitigation database.
Durham et al. teaches:
Fig. 1 with “1st Classification” and ref. 120 (features), 130 (cauterizations), 136 (therapeutic), therefore symptom-mitigation cluster…
PNG
media_image2.png
532
398
media_image2.png
Greyscale
Example of data store (database) with profiles and features…
“In one or more implementations, the computational biology system 102 can be coupled to, or otherwise have access to information stored by a data store 622. The data store 622 can store a plurality of subgroup feature profiles 624. Each of the subgroup feature profiles 624 can indicate a set of features associated with a subgroup of individuals in which a neurodevelopmental condition is present. For example, the subgroup feature profiles 624 can include at least a first subgroup feature profile 626, a second subgroup feature profile 628, and a third subgroup feature profile 630. The set of features included in each respective subgroup feature profile 626, 628, 630 can be different from one another. To illustrate, a first set of features included in the first subgroup feature profile 626 can be different from a second set of features included in the second subgroup feature profile 628. Additionally, a third set of features included in the third subgroup feature profile 630 can be different from the first set of features included in the first subgroup feature profile 626 and the second set of features included in the second subgroup feature profile 628. In one or more illustrative examples, the respective subgroup feature profiles 624 can include at least one of genetic features, observable features, co-occurring conditions, or levels of analytes. In various examples, the features included in the subgroup feature profiles 624 can be determined by the computational biology system 102 for at least a portion of the individuals 602 based on information included in at least one of the genetic data 610, the clinical data 612, or the pathway data 614.” [0138]
Regarding claim 4
The computer-implemented method according to claim 2, wherein the textual data is taken at least in part from the social media posts, which are selected from a social media database by classifying the social media posts by patients from patient groups based on content of the social media posts.
The combined references teach social media. They do not teach database and classifying.
Quintero Padron et al. also in the business of social media teaches:
Social media posts and related to (classifying) disease (content)….
“This social media analytics module displays all public posts in social media related to a specific disease, for example related to “Sjögren's syndrome”, for example on Facebook (registered trademark) and or on Twitter (registered trademark). This social media analytics module is accessible for researchers 41 and clinicians 42 and allows the user to include “free text” in order to refine the list of posts by adding time, location, symptoms, therapies, etc . . . , but only if the public post contains this information. Embedded Posts are a simple way to put public posts, by a page or by a person, into “HarmonicSS” platform. Only public posts are embedded. To access Facebook (registered trademark) post, an application programming interface is used to get data out of, and put data into, Facebook's platform (registered trademark). A low-level HTTP-based application programming interface is used to programmatically query data, post new stories, manage ads, upload photos, and perform other usual tasks that an application programming interface may usually implement. To access to Twitter (registered trademark) post, there is not a dedicated functionality, so this is worked around, for example, by using “HarmonicSS” Twitter (registered trademark) user and get mentions # hashtags, like # Sjögren for example.” [0073]
Social media with database…
“Preferably, said using phase includes one or more social media analytics functions displaying, into social media, public posts made by users of said big data database.” [0048]
Clustering of data…
“The role of the big data mining services 21 is to offer, to the users of the platform, tools and algorithms to analyze the medical data integrated cohorts. A series of tools is offered for analytics, ranging from preprocessing, feature selection and creation, clustering, prediction and association analysis. The requirements provided by the users are thereby covered. There are provided both a learning phase, for example to create a model of prediction or to create a scoring system, but also the possibility to deploy this prediction model. Included algorithms and tools take into account the time dimension, since longitudinal data will be available in the integrated cohorts.” [0068]
It would have been obvious to one of ordinary skill in the art before the effective filing date to include in the method and system of the combined references the ability to use social media database and to classify content as taught by Quintero Padron et al. since the claimed invention is merely a combination of old elements and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. Further motivation is provided by Quintero Padron et al. who teaches the advantages of using social media for symptoms and therapy and accessing social media databases and classifying the content provides useful medical information.
Regarding claim 7
The computer-implemented method according to claim 5, wherein the combined ranking is determined by averaging a distance determined for the similarity to the text of the user symptoms and a distance determined for the similarity to the profile information from the user, and wherein different mitigation strategies are checked to determine if they are similar in order to identify the most common symptom-mitigation strategy.
Durham et al. teaches:
Determine a mean (average) of distribution of distances (profile similarity scores) and determine candidate therapeutic (most common symptom-mitigation strategy)…
“… In various examples, the computational biology system 102 can determine a distribution of pairwise distances, such as profile similarity scores, of the therapeutics included in the gene expression profile distance network 1114 to determine a mean of the distribution and a standard deviation for the distribution. The computational biology system 102 can determine one or more candidate therapeutics corresponding to a consensus gene expression profile 1148 based on a threshold number of standard deviations from the mean for a particular pair of therapeutics having nodes included in the gene expression profile distance network 1114 in relation to an inverted consensus gene expression profile 1148. The threshold number of standard deviations can include no greater than about 3 standard deviations, no greater than about 2.5 standard deviations, no greater than about 2 standard deviations, no greater than 1.5 standard deviations, no greater than about 1 standard deviation, or no greater than about 0.5 standard deviations.” [0194]
Regarding claim 8
The computer-implemented method according to claim1, further comprising filtering the textual data based on whether the textual data contains information about how to mitigate a system symptom.
Durham et al. teaches:
Example of determine therapeutics based on standard deviation (therefore filtering textual data)
“… The computational biology system 102 can determine one or more candidate therapeutics corresponding to a consensus gene expression profile 1148 based on a threshold number of standard deviations from the mean for a particular pair of therapeutics having nodes included in the gene expression profile distance network 1114 in relation to an inverted consensus gene expression profile 1148. The threshold number of standard deviations can include no greater than about 3 standard deviations, no greater than about 2.5 standard deviations, no greater than about 2 standard deviations, no greater than 1.5 standard deviations, no greater than about 1 standard deviation, or no greater than about 0.5 standard deviations.” [0194]
Regarding claim 9
The computer-implemented method according to claim1, further comprising filtering out, modifying or marking unsafe information in the textual data.
Durham et al. teaches:
Filter genes according to confidence levels….
“The computational biology system 102 can also determine genes that have at least a threshold probability of being associated with one or more neurodevelopmental conditions by using one or more software tools. In these situations, the one or more software tools can analyze information associated with at least one gene and one or more neurodevelopmental disorders to determine a level of confidence that the at least one gene is predictive of an individual in which one or more neurodevelopmental conditions are present. In various examples, the computational biology system 102 can filter genes with respect to the one or more neurodevelopmental disorders according to the confidence levels determined by the one or more software tools for the respective genes.” [0067]
Regarding claim 10
The computer-implemented method according to claim1, wherein the clusters of profiles are generated based on prototypes.
Durham et al. teaches:
Determine distribution (cluster) based on consensus (prototype) profile…
“…In various examples, the computational biology system 102 can determine a distribution of pairwise distances, such as profile similarity scores, of the therapeutics included in the gene expression profile distance network 1114 to determine a mean of the distribution and a standard deviation for the distribution. The computational biology system 102 can determine one or more candidate therapeutics corresponding to a consensus gene expression profile 1148 based on a threshold number of standard deviations from the mean for a particular pair of therapeutics having nodes included in the gene expression profile distance network 1114 in relation to an inverted consensus gene expression profile 1148. The threshold number of standard deviations can include no greater than about 3 standard deviations, no greater than about 2.5 standard deviations, no greater than about 2 standard deviations, no greater than 1.5 standard deviations, no greater than about 1 standard deviation, or no greater than about 0.5 standard deviations.” [0194]
Claim 5 is rejected under 35 U.S.C. 103 as being unpatentable over the combined references in section (4) above in further view of Pub. No. US 2017/0308792 to Liang et al.
Regarding claim 5
The computer-implemented method according to claim1, further comprising:
computing, for each text entry of the matching profile cluster, similarity to text of the user symptoms;
Durham et al. teaches:
Similarity between features (text entry of symptoms) related to classification (matching profile cluster)…
“Additionally, the computational biology system 102 can include one or more characterization systems 128. The one or more characterization systems 128 can characterize individuals by determining respective classifications in which to place the individuals. The one or more characterization systems 128 can determine features associated with a number of individuals and determine respective classifications for the plurality of individuals based on the features associated with each classification and the features of the individuals. In various implementations, the one or more characterization systems 128 can determine an amount of similarity between features associated with an individual and features related to one or more of the classifications determined by the one or more classification systems 118. The one or more characterization systems 128 can determine a pairing between a set of features of an individual and a set of features of a classification that has a highest level of similarity and characterize the individual as being included in the classification. The one or more characterization systems 128 can determine amounts of similarity between individuals and classifications based on at least one of commonalities between genetic features, morphological features, levels of analytes, co-occurring conditions, electrical activity features, features included in images, or clinical observations.” [0047]
computing, for each profile in the matching profile cluster, similarity to the profile information from the user;
Example of similarity between genetic information (profile) and feature classification…
“Implementation 7. The method of any one of implementations 1-6, comprising: obtaining, by the computing system, third data of third individuals having at least a threshold decrease in one or more symptoms of an additional condition of the one or more additional conditions in response to treatment of the third individuals with the therapeutic or third data of third individuals where a subset of those have at least a threshold decrease in one or more symptoms of a neurodevelopmental condition in response to treatment of the third individuals with the therapeutic, the third data including third genetic information of the third individuals and third health information of the third individuals; and determining, by the computing system, a measure of similarity between at least one of a portion of the third genetic information or a portion of the third health information with respect to the set of features of the classification; and wherein determining the probability that the candidate treatment of the neurodevelopmental condition for the second group of individuals includes the therapeutic is based at least partly on the measure of similarity.” [0349]
determining a combined ranking by combining the similarity to the text of the user symptoms and the similarity to the profile information from the user; and
Assess how close (ranking) features (text of users symptoms) and individuals (profile from the user), therefore combined, are to the centroid (similarity by combined ranking)…
“The reliability of the groups of features 312, 314, 316 and the subgroups of individuals in which a neurodevelopmental condition is present that are identified according to the groups of features 312, 314, 316 can be evaluated by the computational biology system 102. In various implementations, for an example subgrouping of individuals, the overall similarity within subgroups compared to the similarity within other subgroups can be measured using one or more validity indices, such as Dunn Index, Davies-Boulding index and/or Silhouette index. For example, the groups of features 312, 314, 316 and/or the subgroups of individuals generated using the groups of features 312, 314, 316 can be evaluated using a Davies-Bouldin index, assessing how close the members in a cluster are to the centroid of the cluster and how distant the centroids of different clusters are. In various examples, the features included in the group of features 312, 314, 316 can be identified based on Davies-Bouldin index values for the respective groups of features 312, 314, 316 being at most threshold Davies-Bouldin index values. Additionally, the groups of features 312, 314, 316 and/or the sub-groups of individuals generated using the groups of features 312, 314, 316 can be evaluated using a Silhouette index, assessing how close the members of the same cluster are to each other and how close they are to the members of other clusters…” [0095]
Where groups are genetic feature (profile information) and conditions (symptoms, (text of symptoms)…
“… Each group 312, 314, 316 can include at least one of one or more genetic features, one or more observable features, one or more co-occurring conditions, or one or more analytes. For example, the first group of features 312 can include a genetic feature, C, an analyte, A, and a co-occurring condition, E…” [0092]
identifying a most common symptom-mitigation strategy based on the combined ranking.
A distribution of distances of the therapeutics and determine candidate (most common) therapeutics…
“In one or more implementations, the computational biology system 102 can determine one or more nodes of the gene expression profile distance network 1114 having a minimum distance with respect to an inverted consensus gene expression profile 1148 or having less than a threshold distance with respect to an inverted one of the consensus gene expression profiles 1148. The nodes having less than the threshold distance can correspond to a candidate therapeutic for individuals included in the subgroup corresponding to one of the consensus gene expression profiles 1148. In various examples, the computational biology system 102 can determine a distribution of pairwise distances, such as profile similarity scores, of the therapeutics included in the gene expression profile distance network 1114 to determine a mean of the distribution and a standard deviation for the distribution. The computational biology system 102 can determine one or more candidate therapeutics corresponding to a consensus gene expression profile 1148 based on a threshold number of standard deviations from the mean for a particular pair of therapeutics having nodes included in the gene expression profile distance network 1114 in relation to an inverted consensus gene expression profile 1148. The threshold number of standard deviations can include no greater than about 3 standard deviations, no greater than about 2.5 standard deviations, no greater than about 2 standard deviations, no greater than 1.5 standard deviations, no greater than about 1 standard deviation, or no greater than about 0.5 standard deviations.” [0194]
The combined references teach centroid. They do not teach rankings.
Liang et al. also in the business of centroid teaches:
Identify elements most relevant with ratings (ranking)…
“At block 1712, one or more of the knowledge elements associated with the matching clusters are provided as one or more recommendations to the target user. For example, the knowledge automation system may look up the key-value pair using the centroids of the matching clusters to retrieve the list of knowledge element identifiers associated with the centroids to identify the knowledge elements to recommend to the target user. In some embodiments, the knowledge elements can be filtered to identify knowledge elements that are most relevant or useful to the target user. For example, knowledge elements that the target user has consumed, knowledge elements that are stale, and/or knowledge elements with low ratings can be filtered out when providing the one or more recommendations to the target user.” [0137]
It would have been obvious to one of ordinary skill in the art before the effective filing date to include in the method and system of the combined references the ability to rank as taught by Liang et al. since the claimed invention is merely a combination of old elements and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. Further motivation is provided by Liang et al. who teaches the advantages of rating (ranking) recommendations. The combined references benefit as they are also using clustering information and centroids and would benefit providing recommendations such as symptom-mitigation that have been ranked as the most relevant.
Claim 6 is rejected under 35 U.S.C. 103 as being unpatentable over the combined references in section (4) above in further view of Pub. No. US 2013/0204876 to Szues et al.
Regarding claim 6
The computer-implemented method according to claim 5, further comprising outputting to the user the most common symptom-mitigation strategy along with the textual data that links to the most common symptom-mitigation strategy and profile matches.
The combined references teach symptoms mitigation. They do not teach links.
Szues also in the business of symptoms teaches:
Hypertext (link) from symptoms in patients history (profile) to corpus of medical knowledge…
“Applicants' believe that various exemplary embodiments represent a significant advance in the practical state of the art and science of topic and category identification and have a variety of potential applications. Some example applications may include, e.g., but not limited to: Automatically indexing, categorizing, linking, and/or tagging content for, e.g., but not limited to, subsequent browsing, searching, and/or retrieval as part of a general-purpose information retrieval and/or knowledge management system. According to an exemplary embodiment, such an exemplary system, method, and/or computer program product can range from Internet-scale Web search engines, directories, and/or social media platforms to enterprise content management systems and/or personal and/or social "notebook" applications and/or applets for portable devices. Automatically filtering and/or routing dynamic content (for example, but not limited to, from messaging, syndication, and/or social media technologies and/or services such as, e.g., but not limited to, email, IM, chat, SMS, MMS, RSS/Atom, Twitter, Facebook, Tumblr, Myspace, and/or Google+). One specific example of such an application may include automatically identifying new items of dynamic content that may be of potential interest to a user, and/or organizing such identified items by, e.g., topic. Improving Internet search by allowing search disambiguation or refinement by topic (e.g., "Did you mean Chicago Bears or bears?") and/or grouping search results by topics (e.g., "6 results about Chicago Bears, 4 results about bears"). Improving topic-based Internet dictionaries, such as opendirectory.org. Mapping specific content of an input document to more general topics in a hypertext corpus. A domain-specific example of this in health care may include, e.g., but not limited to, diagnostic support by mapping symptoms documented in a patient's history to potential conditions described in a rich hypertext corpus of medical knowledge.” [0156] – [0160]
“Mapping specific content of an input document to more general topics in a hypertext corpus. A domain-specific example of this in health care may include, e.g., but not limited to, diagnostic support by mapping symptoms documented in a patient's history to potential conditions described in a rich hypertext corpus of medical knowledge.” [0160]
It would have been obvious to one of ordinary skill in the art before the effective filing date to include in the method and system of the combined references the ability to use hypertext (links) as taught by Szues et al. since the claimed invention is merely a combination of old elements and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. Further motivation is provided by Szues et al. who teaches the advantages of hypertext to access information. The combined references benefit as they are also clustering information and providing access via link to support information could benefit users by providing additional information.
Claim 11 is rejected under 35 U.S.C. 103 as being unpatentable over the combined references in section (4) above in further view of Pub. No. US 2024/0403597 to Mancuso.
Regarding claim 11
The computer-implemented method according to claim1, wherein the one or more symptom-mitigation clusters are created using disease or symptom labels, wherein textual descriptions of symptoms are encoded as vectors using a large language model (LLM), and wherein the encoded vectors are used to train a clustering algorithm.
Durham et al. teaches:
Example of labels for variants of pathogens (disease labels)…
“Additionally, the pathogenicity of a variant can be estimated by generating a model based on pathogenicity of additional variants as indicated in the number of data sources 204. In these scenarios, the variants can be labeled as pathogenic or non-pathogenic and can be used to train the model to identify variants that are pathogenic. The individual variants can also be labeled according to a number of features and the variants having one or more specified features can be used to train the model. In various examples, the one or more features can include position of the variant in the genome, overlap of the variant with one or more exons, frequency of the variant in a population in which the neurodevelopmental condition is not present, conservation scores, pathogenicity scores, scores indicating association between the variant and risk genes for the neurodevelopmental condition, overlap of the variant with regions of the genome that correspond to the neurodevelopmental condition, or one or more combinations thereof. In this way, the model can be implemented to determine the first score after the model is trained and validated.” [0072]
Vectors with neural network (artificial intelligence)…
“In one or more illustrative implementations, a number of output vectors can be generated that include a number of weights. The weights can be generated by one or neural network models based on input vectors that include ones and zeros corresponding to each enumeration of one or more enumerations of the window throughout the content. In various examples, a one can indicate that a word is present in the center of a window and a zero can indicate that a word is not present in the center of the window. In one or more examples, the one or more neural networks can include a two-layer neural network that implements a normalization function, such as a normalized exponential function.” [0083]
The combined references teach machine learning, vectors, and clustering. They do not teach large machine learning including language models, training, clustering, and vector.
Mancuso also in the business of machine learning and vectors teaches:
Train machine learning model to provides intelligent cmmunication…
“Furthermore, as used herein, the term “machine learning model” refers to a computer representation that can be tuned (e.g., trained) based on inputs to approximate unknown functions. Indeed, a machine learning model can refer to a computer representation that can be tuned (e.g., trained) based on inputs to generate electronic communications (or other digital content, such as documents). Additionally, a machine learning model can refer to a computer representation that can be tuned (e.g., trained) based on inputs to analyze text and/or images. In one or more implementations, parameters of a machine learning model can be adjusted or trained to create a communication generation neural network that intelligently electronic communications from prompts (e.g., text within an electronic communication thread) that also emulate one or more composition styles from a user account and leverages content items corresponding to the user account.” [0038]
Machine learning model with large language, training, clustering models and vector learning…
“For instance, a machine learning model can include, but is not limited to, one or more convolutional neural networks, recurrent neural networks, generative adversarial neural networks), residual neural networks, diffusion models, or a combination thereof. Additionally, a machine learning model can also include, but is not limited to one or more large language models, differentiable function approximators, contrastive language-image pre-training models, clustering models, convolution neural network-based image classifiers, recurrent neural network-based image classifiers, Term Frequency Inverse Document Frequency (TF-IDF) encoders, Word2Vecs, matrix factorization vector learning approaches, local context window vector learning approaches, Global Vectors for Word Representation (GloVe), Bidirectional Encoder Representations from Transformers, natural language processing approaches (e.g., spaCy), and/or generative pre-trained transformer models.” [0039]
It would have been obvious to one of ordinary skill in the art before the effective filing date to include in the method and system of the combined references the ability to use large language models with vectors as taught by Mancuso since the claimed invention is merely a combination of old elements and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. Further motivation is provided by Mancuso who teaches the advantages of using various existing technology such as large language models with vectors for intelligent communication. The combined references benefit as they also use machine learning and using existing large language technology provides intelligent results.
Claims 12 and 13 are rejected under 35 U.S.C. 103 as being unpatentable over the combined references in section (4) above in further view of Pub. No. US 2024/0147025 to Graciarena et al.
Regarding claim 12
The computer-implemented method according to claim1, further comprising concatenating the profile information for each of the authors of the textual data into a feature vector, which in each case is assigned to a mitigation strategy extracted from the textual data, and training a clustering model for each of the one or more symptom-mitigation clusters to generate the plurality of profile clusters in the one or more symptom-mitigation clusters.
The combined references teach cluster. They do not teach concatenating.
Graciarena et al. also in the business of vector cluster:
Train clustering…
“In some examples, the computing system 300 may, for any one or more of neural network 314, scoring module 318, or training module 324, apply one or more of nearest neighbor, naïve Bayes, decision trees, linear regression, support vector machines, neural networks, k-Means clustering, Q-learning, temporal difference, deep adversarial networks, or other supervised, unsupervised, semi-supervised, or reinforcement learning algorithms to train the machine learning models.” [0046]
Concatenated feature vector with source profile…
“… In some examples, statistical model 320 may map a concatenated modality feature vector that includes a fused version of each modality feature vector for modality feature of multimodal content 352. Statistical model 320 may map the concatenated modality feature vector to one or more feature characteristics associated with a concatenated version of a source model included in source profile 344.” [0047]
It would have been obvious to one of ordinary skill in the art before the effective filing date to include in the method and system of the combined references the ability to concatenate data as taught by Graciarena et al. since the claimed invention is merely a combination of old elements and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. Further motivation is provided by Graciarena et al. who teaches the advantages of concatenated feature vectors. The combined references benefit as they also use analysis with source information to provide results and using vectors provides a way to associate source profile with combined information.
Regarding claim 13
The computer-implemented method according to claim1, further comprising generating a new feature vector based on a determination that a respective feature vector of the profile information of one of the authors has a distance to a centroid of one of the profile clusters that is greater than a threshold.
Durham et al. teaches:
Cluster with distance to centroid…
“The reliability of the groups of features 312, 314, 316 and the subgroups of individuals in which a neurodevelopmental condition is present that are identified according to the groups of features 312, 314, 316 can be evaluated by the computational biology system 102. In various implementations, for an example subgrouping of individuals, the overall similarity within subgroups compared to the similarity within other subgroups can be measured using one or more validity indices, such as Dunn Index, Davies-Boulding index and/or Silhouette index. For example, the groups of features 312, 314, 316 and/or the subgroups of individuals generated using the groups of features 312, 314, 316 can be evaluated using a Davies-Bouldin index, assessing how close the members in a cluster are to the centroid of the cluster and how distant the centroids of different clusters are. In various examples, the features included in the group of features 312, 314, 316 can be identified based on Davies-Bouldin index values for the respective groups of features 312, 314, 316 being at most threshold Davies-Bouldin index values…” [0095]
The combined references teach cluster. They do not teach feature vector.
Graciarena et al. also in the business of cluster:
Train clustering…
“In some examples, the computing system 300 may, for any one or more of neural network 314, scoring module 318, or training module 324, apply one or more of nearest neighbor, naïve Bayes, decision trees, linear regression, support vector machines, neural networks, k-Means clustering, Q-learning, temporal difference, deep adversarial networks, or other supervised, unsupervised, semi-supervised, or reinforcement learning algorithms to train the machine learning models.” [0046]
Concatenated feature vector with source profile…
“… In some examples, statistical model 320 may map a concatenated modality feature vector that includes a fused version of each modality feature vector for modality feature of multimodal content 352. Statistical model 320 may map the concatenated modality feature vector to one or more feature characteristics associated with a concatenated version of a source model included in source profile 344.” [0047]
It would have been obvious to one of ordinary skill in the art before the effective filing date to include in the method and system of the combined references the ability to concatenate data as taught by Graciarena et al. since the claimed invention is merely a combination of old elements and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. Further motivation is provided by Graciarena et al. who teaches the advantages of concatenated feature vectors. The combined references benefit as they also use analysis with source information to provide results and using vectors provides a way to associate source profile with combined information.
Claim 16 is rejected under 35 U.S.C. 103 as being unpatentable over the combined references in section (4) above in further view of Pub. No. US 2020/0367807 to Lassoued et al.
Regarding claim 16
The computer-implemented method of claim 1, wherein the symptom- mitigation strategy is used to support decision making, and wherein the method is applied to optimize a healthcare machine learning or artificial intelligence system for a medical patient.
Durham et al. teaches:
Machine learning and determine features to include in classification of individuals…
“…In one or more implementations, the one or more classification systems 118 can implement at least one of one or more machine learning techniques or one or more statistical techniques to determine amounts of correlation between features of individuals in which a neurodevelopmental condition is present. To illustrate, the one or more classification systems 118 can determine features related to classification of individuals in which a neurodevelopmental condition is present based on the features having at least a threshold amount of correlation with respect to one another. The threshold amount of correlation can be based on a number of the individuals in which a neurodevelopmental condition is present that exhibit the features in relation to a broader group of individuals in which the neurodevelopmental condition is present. In various examples, the one or more classification systems 118 can implement one or more combinations of multiple machine learning techniques and/or one or more combinations of statistical techniques to determine features to include in each of a number of classifications of individuals in which a neurodevelopmental condition is present.”
The combined references teach machine learning. They do not teach optimize.
Lassoued et al. also in the business of machine learning teaches:
Optimize health state based on reinforcement (machine) learning…
“In an additional aspect, the well-being controller 512 may identify and/or learn an optimized/best policy to maximize a user's health state over a time period/horizon (e.g., one working day/week), based on a Markov decision processes (“MDP”) and reinforcement learning operations.” [0102]
It would have been obvious to one of ordinary skill in the art before the effective filing date to include in the method and system of the combined references the ability to optimize a model as taught by Lassoued et al. since the claimed invention is merely a combination of old elements and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable. Further motivation is provided by Lassoued et al. who teaches the advantages of optimizing a model to maximize a user’s health. The combined references benefit as they also are improving user’s health.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
The following prior art teaches at least clustering and centroid or social media:
CN-110534190-A; US-20200279622-A1; US-20140172864-A1; US-20240135291-A1; US-20210319887-A1; US-20220406405-A1; US-20180253650-A9; US-20240266009-A1; US-20190131004-A1; US-20220020504-A1; US-20140052475-A1; US-12431223-B1; US-10769241-B1; US-10946311-B1; US-11894117-B1
Any inquiry concerning this communication or earlier communications from the examiner should be directed to KENNETH BARTLEY whose telephone number is (571)272-5230. The examiner can normally be reached Mon-Fri: 7:30 - 4:00 EST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, 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.
/KENNETH BARTLEY/Primary Examiner, Art Unit 3684
Prosecution Insights