Granted Patent B2

US 12,651,286 · App. 16/380,771 · Granted Jun 9, 2026

Inventor: Erik M Simpson (Houston, TX)

Assignee: CIRCLESX LLC

G06Q30/0631G01N33/48792G01N33/492G06N3/042G06N3/08G06Q30/0633G16B40/00G16H20/60

Abstract

Claims (96)

1 . A method, comprising:

acquiring biomarker data corresponding to a user, wherein the biomarker data comprises data corresponding to one or more measurement values of one or more biomarkers for the user for one or more time periods;

receiving, by one or more computing systems, nutritional data corresponding to a plurality of food ingredients;

determining, by the one or more computing systems, a plurality of food combinations based on the plurality of food ingredients, wherein a respective food combination comprises two or more food ingredients of the plurality of food ingredients;

training, by the one or more computing systems, one or more machine learning models to determine a plurality of optimized weightings for the respective food combination for the user based on the biomarker data and the nutritional data, wherein a respective optimized weighting corresponds to a serving proportion for a respective food ingredient of the respective food combination, and wherein the plurality of optimized weightings corresponds to probability weightings with iterative feedback based on at least the biomarker data;

determining, by the one or more computing systems, a plurality of optimized food combinations based on the plurality of optimized weightings, wherein the plurality of optimized food combinations is a subset of the plurality of food combinations;

generating, by the one or more computing systems, node rankings of the plurality of optimized food combinations based on a ranking function utilizing the plurality of optimized weightings; and

receiving, via one or more user interfaces of a user device, a selection from the user, wherein the selection corresponds to one or more selected food combinations from the plurality of optimized food combinations.

2 . The method of claim 1 , wherein:

the biomarker data further comprises complete blood count data, red blood cell data, white blood cell data, platelets data, hemoglobin data, hematocrit data, mean corpuscular volume data, blood chemistry tests data, basic metabolic panel data, blood glucose data, calcium data, electrolytes data, kidneys data, blood enzyme test data, troponin data, creatine kinase data, cholesterol data, LDL cholesterol data, HDL cholesterol data, triglyceride data, lipoprotein panel data, coagulation panel data, echocardiogram data, nuclear perfusion study data, magnetic resonance imaging data, positron emission tomography data, or combinations thereof;

the nutritional data comprises fat data, sugar data, caloric data, or combinations thereof;

the biomarker data further corresponds to a plurality of biological samples obtained from the user during the one or more time periods, wherein the plurality of biological samples was obtained before and after the user consumed at least a subset of the plurality of food ingredients during the one or more time periods;

the serving proportion for the respective food ingredient comprises a serving size proportion for the respective food ingredient, a calorie count proportion for the respective food ingredient, or combinations thereof; or

combinations thereof.

3 . The method of claim 1 , wherein acquiring the biomarker data comprises:

obtaining biological samples data corresponding to the user before and after the user consumed at least a subset of the plurality of food ingredients during the one or more time periods; and

determining, by the one or more computing systems, the biomarker data based on the biological samples data.

4 . The method of claim 1 , wherein receiving, via the one or more user interfaces of the user device, the selection from the user comprises receiving, by the one or more computing systems, selection data from the user device.

5 . The method of claim 1 , wherein receiving, via the one or more user interfaces of the user device, the selection from the user comprises:

generating a database search engine based on the plurality of optimized food combinations;

providing a search interface for the database search engine to the user device, wherein the search interface is configured to be displayed by the user device;

receiving search query input data from the user, wherein the search query input data corresponds to the plurality of food ingredients;

generating search results data based on the search query input data, the biomarker data, the nutritional data, and the one or more machine learning models, wherein the search results input data corresponds to at least a subset of the plurality of optimized food combinations;

providing a results interface to the user device based on the search results data, wherein the results interface includes at least the subset of the plurality of optimized food combinations ordered based on the node rankings, and wherein the results interface is configured to be displayed by the user device; and

receiving the selection from the user via the results interface of the user device.

6 . The method of claim 5 , wherein generating the search results data comprises:

determining location data corresponding to the user device using one or more satellite navigation systems, wherein the location data corresponds to a geographic location of the user device; and

generating the search results data based on the location data, the search query input data, the biomarker data, the nutritional data, and the one or more machine learning models.

7 . The method of claim 1 , wherein:

the one or more machine learning models are configured to use linear and non-linear optimization systems, wherein the linear and non-linear optimization systems comprise one or more vector maximization and minimization equations;

the one or more machine learning models comprise one or more neural networks, one or more linear regression models, one or more logistic regression models, one or more linear discriminant analysis models, one or more classification or regression tree models, one or more naïve Bayes models, one or more learning vector quantization models, one or more posterior density function models, one or more independent stochastic regressor models, one or more general stochastic regression models, one or more general non-linear hypothesis models, or combinations thereof; or

combinations thereof.

8 . The method of claim 1 , wherein:

the plurality of optimized food combinations comprises a subset of the plurality of food ingredients; and

training, by the one or more computing systems, the one or more machine learning models to determine the plurality of optimized weightings for the respective food combination for the user comprises:

mapping, from one or more databases, linear and non-linear systems of the biomarker data to generate a first matrix;

mapping, from the one or more databases, linear and non-linear systems of the nutritional data to generate a second matrix; and

performing one or more matrix multiplications of at least the first and the second matrices to generate a first optimized weighting of the plurality of optimized weightings, wherein the first optimized weighting corresponds to one or more food ingredients of the plurality of food ingredients.

9 . The method of claim 1 , further comprising determining, by the one or more computing systems, one or more health care costs based on the biomarker data, the selection, and the one or more machine learning models.

10 . The method of claim 1 , wherein training, by the one or more computing systems, the one or more machine learning models to determine the plurality of optimized weightings for the respective food combination for the user comprises:

determining a plurality of return values of the plurality of food ingredients for the user based on the biomarker data, wherein a respective return value of the respective food ingredient corresponds to an increase or a decrease of the one or more measurement levels towards one or more target values after the respective food ingredient has been consumed by the user;

determining a plurality of expected return values of the plurality of food ingredients for the user based on the plurality of return values; and

determining a plurality of standard deviation values of the plurality of food ingredients for the user based on the plurality of return values and the plurality of expected return values.

11 . The method of claim 10 , wherein training, by the one or more computing systems, the one or more machine learning models to determine the plurality of optimized weightings for the respective food combination for the user further comprises:

determining a plurality of candidate weight value sets for the respective food combination;

determining a plurality of combined expected return values for the respective food combination for the user based on the plurality of candidate weight value sets and the plurality of expected return values;

determining a plurality of covariance values for the plurality of food combinations based on the plurality of return values and the plurality of expected return values, wherein a respective covariance value corresponds to the respective food combination;

determining a plurality of combined standard deviation values for the respective food combination based on the plurality of candidate weight value sets, the plurality of standard deviation values of the plurality of food ingredients, and the respective covariance value; and

determining the plurality of optimized weightings for the respective food combination based on the plurality of combined expected return values, the plurality of combined standard deviation values, and the respective covariance value.

12 . The method of claim 11 , wherein determining the plurality of optimized weightings for the respective food combination based on the plurality of combined expected return values, the plurality of combined standard deviation values, and the respective covariance value comprises:

determining a plurality of ratio values for the respective food combination, wherein a respective ratio value corresponds to a ratio of a respective combined expected return value to a respective combined standard deviation value for a respective candidate weight value set;

determining a maximum ratio value of the plurality of ratio values; and

determining the plurality of optimized weightings for the respective food combination based on the maximum ratio value, the plurality of expected return values, the plurality of standard deviation values of the plurality of food ingredients, and the respective covariance value.

13 . The method of claim 12 , wherein generating, by the one or more computing systems, the node rankings of the plurality of optimized food combinations comprises:

generating the node rankings based on respective ratio values for the plurality of optimized food combinations.

14 . A computing system, comprising:

one or more processors; and

at least one memory, comprising program instructions which, when executed by the one or more processors, cause the one or more processors to:

acquire biomarker data corresponding to a user, wherein the biomarker data comprises data corresponding to one or more measurement values of one or more biomarkers for the user for one or more time periods;

receive nutritional data corresponding to a plurality of food ingredients;

determine a plurality of food combinations based on the plurality of food ingredients, wherein a respective food combination comprises two or more food ingredients of the plurality of food ingredients;

generate one or more machine learning models to determine a plurality of optimized weightings for the respective food combination for the user based on the biomarker data and the nutritional data, wherein a respective optimized weighting corresponds to a serving proportion for a respective food ingredient of the respective food combination, and wherein the plurality of optimized weightings corresponds to probability weightings with iterative feedback based on at least the biomarker data;

determine a plurality of optimized food combinations based on the plurality of optimized weightings, wherein the plurality of optimized food combinations is a subset of the plurality of food combinations;

generate node rankings of the plurality of optimized food combinations based on a ranking function utilizing the plurality of optimized weightings; and

receive, via one or more user interfaces of a user device, a selection from the user, wherein the selection corresponds to one or more selected food combinations from the plurality of optimized food combinations.

15 . The computing system of claim 14 , wherein the biomarker data further corresponds to a plurality of biological samples obtained from the user during the one or more time periods, wherein the plurality of biological samples were obtained before and after the user consumed at least a subset of the plurality of food ingredients during the one or more time periods.

16 . The computing system of claim 14 , wherein the program instructions which, when executed by the one or more processors, cause the one or more processors to receive, via the one or more user interfaces of the user device, the selection from the user further cause the one or more processors to:

generate a database search engine based on the plurality of optimized food combinations;

provide a search interface for the database search engine to the user device, wherein the search interface is configured to be displayed by the user device;

receive search query input data from the user, wherein the search query input data corresponds to the plurality of food ingredients;

generate search results data based on the search query input data, the biomarker data, the nutritional data, and the one or more machine learning models, wherein the search results input data corresponds to at least a subset of the plurality of optimized food combinations;

provide a results interface to the user device based on the search results data, wherein the results interface includes at least the subset of the plurality of optimized food combinations ordered based on the node rankings, and wherein the results interface is configured to be displayed by the user device; and

receive the selection from the user via the results interface of the user device.

17 . The computing system of claim 14 , wherein:

the one or more machine learning models are configured to use linear and non-linear optimization systems, wherein the linear and non-linear optimization systems comprise one or more vector maximization and minimization equations;

the one or more machine learning models comprise one or more neural networks, one or more linear regression models, one or more logistic regression models, one or more linear discriminant analysis models, one or more classification or regression tree models, one or more naïve Bayes models, one or more learning vector quantization models, one or more posterior density function models, one or more independent stochastic regressor models, one or more general stochastic regression models, one or more general non-linear hypothesis models, or combinations thereof; or

combinations thereof.

18 . A non-transitory computer-readable medium having stored thereon a plurality of computer-executable instructions which, when executed by a computer, cause the computer to:

acquire biomarker data corresponding to a user, wherein the biomarker data comprises data corresponding to one or more measurement values of one or more biomarkers for the user for one or more time periods;

receive nutritional data corresponding to a plurality of food ingredients;

determine a plurality of food combinations based on the plurality of food ingredients, wherein a respective food combination comprises two or more food ingredients of the plurality of food ingredients;

train one or more machine learning models to determine a plurality of optimized weightings for the respective food combination for the user based on the biomarker data and the nutritional data, wherein a respective optimized weighting corresponds to a serving proportion for a respective food ingredient of the respective food combination, and wherein the plurality of optimized weightings corresponds to probability weightings with iterative feedback based on at least the biomarker data;

determine a plurality of optimized food combinations based on the plurality of optimized weightings, wherein the plurality of optimized food combinations is a subset of the plurality of food combinations;

generate node rankings of the plurality of optimized food combinations based on a ranking function utilizing the plurality of optimized weightings; and

receive, via one or more user interfaces of a user device, a selection from the user, wherein the selection corresponds to one or more selected food combinations from the plurality of optimized food combinations.

19 . The non-transitory computer-readable medium of claim 18 , wherein the plurality of computer-executable instructions which, when executed by the computer, cause the computer to receive, via the one or more user interfaces of the user device, the selection from the user further cause the computer to:

generate a database search engine based on the plurality of optimized food combinations;

provide a search interface for the database search engine to the user device, wherein the search interface is configured to be displayed by the user device;

receive search query input data from the user, wherein the search query input data corresponds to the plurality of food ingredients;

generate search results data based on the search query input data, the biomarker data, the nutritional data, and the one or more machine learning models, wherein the search results input data corresponds to at least a subset of the plurality of optimized food combinations;

provide a results interface to the user device based on the search results data, wherein the results interface includes at least the subset of the plurality of optimized food combinations ordered based on the node rankings, and wherein the results interface is configured to be displayed by the user device; and

receive the selection from the user via the results interface of the user device.

20 . The non-transitory computer-readable medium of claim 18 , wherein:

the one or more machine learning models are configured to use linear and non-linear optimization systems, wherein the linear and non-linear optimization systems comprise one or more vector maximization and minimization equations;

the one or more machine learning models comprise one or more neural networks, one or more linear regression models, one or more logistic regression models, one or more linear discriminant analysis models, one or more classification or regression tree models, one or more naïve Bayes models, one or more learning vector quantization models, one or more posterior density function models, one or more independent stochastic regressor models, one or more general stochastic regression models, one or more general non-linear hypothesis models, or combinations thereof; or

combinations thereof.

Assignments (3)

Continuity (2)

References Cited (400)