Automated systems for acquiring, processing, and communicating health data, as well as methods for doing so.

Abstract

Regarding automated health data acquisition, more specifically, it automatically optimizes values ​​associated with the probability of future events. [Solution] Modulated outputs from each of multiple models are integrated to quantify the factors and generate multiple values, each of which lies within a continuous distribution. Each discrete category is associated with some of the values ​​to represent the likelihood of future occurrence. The values ​​are received from the first, second, and third data models. The values ​​are modulated to scale the values ​​that represent the aspects associated with the likelihood of future occurrence. The modulation is based on at least one factor derived from at least one of the first, second, and third data models. The values ​​are then integrated as functional artificial intelligence provided on at least one computing device, and each discrete category associated with the integrated values ​​is selected, with each of the integrated values ​​and the selected category representing the likelihood of future occurrence.

Description

Technical Field

[0001] This application generally relates to automated health data acquisition, and more particularly to automatically optimizing values associated with the probability of future events.

Background Art

[0002] Quantitatively evaluating an individual's health state is complex and requires considering various factors. Some people evaluate health considering various risks and exposures to health risks, but often other factors, such as factors derived from an individual's quality of life and lifestyle patterns, are not considered. Furthermore, although broad definitions of health are known, such definitions are often qualitative or at most semi - quantitative, and in many cases, in evaluations, a holistic approach for quantitatively determining the health state is not used.

[0003] The evaluation of the health state is often done by measuring the absence of a health state or the presence of a disease, usually in a severe state, which can be measured by physical evaluation, objective procedures, and tests. Furthermore, traditional evaluations of health are related to morbidity and the risk of death. In such cases, the measured values can be quantified, but for other important health elements, little is revealed, which may lead to misunderstandings. Therefore, deriving a single numerical value, such as a quantitative score, as a consistent and accurate measure of an individual's health represents a significant challenge.

[0004] A holistic approach for measuring or evaluating a person's health includes at least some information based on self - assessment. Obtaining accurate and up - to - date self - assessment information can be problematic, especially when participants do not provide information regularly and consistently. Unfortunately, participants often stop providing accurate and up - to - date self - evaluated health - related information when the platform seeking such information is not very simple, not attractive enough, and / or not appealing.

[0005] Furthermore, known health assessment inductive models (i.e., models derived directly from data) do not (or cannot) include all possible predictors. This is partly because not all predictors associated with a particular event are known, but also because some known predictors may be difficult or impractical to measure. Moreover, it is recognized herein that known models constructed using data collected from a given population, which can be used to summarize and generalize the morbidity and mortality characteristics of that population, may inaccurately provide risk estimates derived from other populations. Thus, relevant predictors may not be included in existing models and may differ significantly between two different populations. This may lead to adopting predictive models that underestimate or overestimate certain risks. [Overview of the project] [Means for solving the problem]

[0006] Computer-implemented systems and methods for integrating modulated outputs from each of several models to quantify factors for generating multiple values ​​are disclosed herein. Each value lies within a continuous distribution. A discrete category is selected to be associated with each value, representing the likelihood of future occurrence. In one or more implementations, a first data model running on at least one computing device quantifies each of several endpoints contributing to the current state and the likelihood of future occurrence, with each quantified value of each quantified endpoint being computed within a continuous distribution. Furthermore, a second data model running on at least one computing device generates each value representing at least one aspect of the current state that influences the likelihood of future occurrence, with each generated value lies within a continuous distribution. Furthermore, a third data model running on at least one computing device identifies individual factors among several factors associated with a subset of endpoints and / or aspects that can be individually modified. The values ​​are generated within a continuous distribution representing each of the several factors.

[0007] Continuing with one or more implementations of this application, a modulation model running on a computing device can modulate at least one value, quantified by a first data model, generated by a second data model, and / or identified by a third data model, to scale a value representing at least one aspect associated with the likelihood of future occurrence. The modulation can be based on at least one factor derived from at least one of the first, second, and third data models. Furthermore, at least one of artificial intelligence and machine learning may be provided on at least one computing device and used to integrate at least two of the values ​​associated with each of a plurality of continuous distributions. A discrete category can be selected to be associated with each integrated value, and each integrated value and the selected category represents the likelihood of future occurrence.

[0008] In one or more implementations of this application, at least one computing device substitutes at least one value that is not included in the set of inputs used by the first data model.

[0009] In one or more implementations of this application, at least one computing device substitutes at least one other value that is not included in previously substituted values ​​or quantified values ​​associated with a previously quantified endpoint, and at least one other substituted value depends on at least one of the previously substituted values. Furthermore, at least one other value lies within a continuous distribution.

[0010] In one or more implementations of this application, at least one computing device recalibrates at least one of a first data model, a second data model, and a third data model as a function of information received over time or from multiple data sources.

[0011] In one or more implementations of this application, the first data model uses each endpoint as a composite input function in the fitting procedure.

[0012] In one or more implementations of this application, at least one computing device is configured using a software application that provides a graphical user interface on the user computing device, and the user computing device receives at least some of values ​​and aspects from a user operating the user computing device. Furthermore, the at least one computing device receives at least some of values ​​and aspects via a data communication session, and transmits to the user computing device quantified values ​​associated with at least some of the endpoints, quantified values ​​associated with at least some of the aspects, and generated values ​​associated with at least some of the factors from a first data model, a second data model, and a third data model, respectively. Furthermore, the user computing device is further configured by a software application to display the received values ​​received from the at least one computing device.

[0013] Furthermore, in one or more implementations of the present application, at least one computing device is configured using a software application that provides a graphical user interface on the user computing device, the graphical user interface prompts the user periodically and periodically to input values ​​associated with a factor, and further, the graphical user interface automatically provides an interactive display screen if no values ​​associated with a factor are received after previously received values.

[0014] Furthermore, at least one of the first, second, and third data models may include selections of at least two other data models. Additionally, the values ​​can be computed in a continuous distribution as a function of a parametric nonlinear mapping. Furthermore, at least one of the first, second, and third data models may include at least one of artificial intelligence and machine learning.

[0015] These and other embodiments, features, and advantages can be understood from the attached description of specific embodiments of the present invention, as well as from the attached drawings and claims.

[0016] Various features, embodiments, and advantages of the present invention can be understood from the following detailed description and the accompanying drawings. [Brief explanation of the drawing]

[0017] [Figure 1A] This block diagram shows the model, components, and the metric health score generated as a function thereof, relating to an exemplary implementation of this application. [Figure 1B] This block diagram shows components associated with multiple models (risk model and modulator model) and processes associated with modulation, according to the present application. [Figure 1C] This figure shows a graph representing the accuracy of the assignment engine in an exemplary implementation. [Figure 1D] This figure shows the distribution of unequalized scores (plotted in the upper panel) and the corresponding distribution of equalized scores. [Figure 2] This block diagram shows the hardware configuration related to an exemplary implementation of this application. [Figure 3] This figure shows exemplary components and configurations of a computing device that can be used to implement the techniques described herein. [Figure 4]A diagram showing an exemplary screen display associated with a graphical user interface operable to provide the functions described herein. [Figure 5A] A diagram showing an exemplary screen display associated with a graphical user interface operable to provide the functions described herein. [Figure 5B] A diagram showing an exemplary screen display associated with a graphical user interface operable to provide the functions described herein. [Figure 6A] A diagram showing an exemplary screen display associated with a graphical user interface operable to provide the functions described herein. [Figure 6B] A diagram showing an exemplary screen display associated with a graphical user interface operable to provide the functions described herein. [Figure 6C] A diagram showing an exemplary screen display associated with a graphical user interface operable to provide the functions described herein. [Figure 7] A diagram showing an exemplary screen display associated with a graphical user interface operable to provide the functions described herein. [Figure 8] A diagram showing an exemplary screen display associated with a graphical user interface operable to provide the functions described herein. [Figure 9A] A diagram showing an exemplary screen display associated with an exemplary data conversion process in an implementation of the present application. [Figure 9B] A diagram showing an exemplary screen display associated with an exemplary data conversion process in an implementation of the present application.

Embodiments for Carrying Out the Invention

[0018] In summary and introduction, this application includes systems, techniques, and interfaces for data processing, including for collecting, designing, modeling, simulating, and generating information related to an individual's health. Such information may include not only specific factors of an individual's health, but also a more general state of the individual's health at a given point in time, in the past, or in the future. Furthermore, this application includes systems and methods for transforming collected and / or generated information and interface with a computing platform to increase and / or influence the actions associated with such devices as a function of such transformed information.

[0019] This application provides a health assessment according to a multidimensional framework that includes estimates of early mortality and life expectancy, as well as physical, emotional, and cognitive factors to varying degrees. Such factors may be measured and may be based on an individual's perception of their current health status and future health prospects that take into account potential lifestyle changes. Thus, each person's health status is assessed to the extent to which their quality of life may change through modifications of behavior and lifestyle.

[0020] The systems and methods of this application include one or more modules for estimating the respective probabilities associated with an individual suffering one or more serious health events over a given period of time. In one or more implementations of this application, such estimations include probabilities based on a model constructed using observed characteristics of associated subjects within one or more populations. For example, measurement of a variable is performed on a set of subjects at a specific point in life to generate a baseline. Subjects are then tracked and a set of subject metrics is recorded over time. Using this data, and by applying the data to health events or endpoints such as stroke, cancer, or myocardial infarction, a mathematical model can be constructed and used to estimate the probability that an individual will suffer one or more of such events over time. A set of baseline measurements of a variable can be generated that can be used as statistically relevant predictors for various health-related events. Although not all target events are necessarily fatal, a time lag between the time the predictors were measured and the corresponding target health event predicted by such a model can be used to calculate a “survival probability” value.

[0021] This application enables the implementation of one or more validated and maintained mathematical models, including periodic recalibration of model parameters using new data. For example, a model derived from data can be validated using information received from one or more sources, such as survey data received from the National Health and Nutrition Examination ("NHANES"). This application may include the operation of updating one or more models by including additional covariates that are observed to correlate with the recorded frequencies of one or more target health-related events. One simple example is the Framingham risk function, which can be recalibrated using a variety of possible techniques, such as logistic regression or a Cox proportional hazards model applied to another dataset, such as NHANES III.

[0022] Here, referring to drawings where similar reference numbers point to similar elements, Figure 1A is a simple block diagram showing multiple models (metric health model 102, quality of life model 104, and lifestyle model 106), as well as the respective components (e.g., models) used to create and / or form models 102, 104, and 106. Furthermore, as will be described in detail herein, the outputs from models 102, 104, and 106 can be used, among other things, to generate a metric health score 110 that provides a quantification of the user's health, which is highly personal and contextual, but retains personal health-related details associated with a person. Figure 1B is a block diagram showing the components associated with multiple models (risk model, and modulator model) and the processes associated with modulation, as will be shown and described herein.

[0023] More specifically, this application may include a particular mathematical model, commonly referred to herein as Metric Health Model (MHM) 102, which can be used to quantify the extent to which each of a set of measurable health-related parameters, such as blood pressure or total cholesterol, affects an individual's health status. More specifically, MHM 102 can be used to measure the range of measured parameters that affect an individual's risk of developing each of the respective health conditions in the future. In one or more implementations, MHM 102 is derived using data from existing and ongoing studies, as well as existing metric health survival models.

[0024] In one or more implementations, MHM102 can be based on cardiovascular and cerebrovascular endpoint data and predictive risk models, and may further include one or more cancer risk models. MHM102 may include a selection of specific models that can be combined in specific ways to generate general measures of an individual's overall health. Furthermore, risks can be assessed by combining probabilities or by using averaging procedures, thereby enabling appropriate estimation and definition of total risk probabilities. In one or more implementations, models can be combined at two levels: (1) as a simple approximation such as a direct arithmetic mean of the corresponding risks, and (2) more precisely by using hazard ratios or the risk itself as a composite function used in fitting procedures, such as constructing a Cox proportional hazards model. In one or more implementations, MHM102 excludes one or more explicit cancer risk models, but the assessment of risk is supported not only for individuals at risk of various cancers, but also for a variety of other medical conditions that may be excluded, such as gastrointestinal disorders and Alzheimer's disease. Thus, MHM102 of this application can be used to generate accurate predictions of very general aggregated health metrics. Furthermore, the MHM102 of this application can also be used to generate accurate predictions related to an individual's long-term and short-term future medical expenses. In addition, the MHM102 can function as an accurate survival predictor.

[0025] In addition to MHM102, this application may include a specific mathematical model commonly referred to herein as Quality of Life Model (QLM)104. QLM104 can be derived from quantified aspects of an individual's health that are important for determining accurate measures of an individual's health relating to an individual's quality of life, but such factors are difficult to quantify. For example, certain factors associated with an individual being bedridden due to severe chronic pain, or factors associated with debilitating depression, can have a significant impact on an individual's quality of life and their potential risk of contracting that illness. These and other eligible factors have traditionally been considered to have less or secondary relevance in assessing values ​​representing an individual's health, at least in part, because such values ​​are difficult to measure and quantify in a clear manner. The application of QLM104 in this application addresses such concerns and quantifies such factors by including self-assessed factors that can be used to estimate numerical values ​​representing an individual's quality of life, at least in part.

[0026] In one or more implementations, self-assessments can be received from a user computing device as a function of a module operating on such a device that provides prompts for various factors in the form of questions and / or comments in an interactive computing program platform. Such factors can be recognized as being common to both metric and self-assessment health estimates. The prompts and software platform providing such prompts according to this application can be configured to ensure that the received self-assessments are up-to-date and accurate, thereby ensuring a statistically significant correlation between qualitative assessments (e.g., an individual's mood) and accurate and realistic health measurements. Furthermore, it can generate at least semi-quantitative measurements that represent an individual's quality of life. The QLM104 can take into account the various influences of an individual's diverse emotional tendencies, such as mood and emotional state, with respect to cardiovascular and other health-related risks.

[0027] Accordingly, MHM102 and QLM104, as applied pursuant to this application, provide both a quantitative and comprehensive view of an individual's current health status, describing numerous health-related factors, including those determined from genetic, familial data, pre-existing conditions, personification, demographics, inflammation, as well as metabolic data, lifestyle data, and self-assessment data.

[0028] In addition to MHM102 and QLM104, this application may further include a specific mathematical model commonly referred to herein as Lifestyle Model (LSM)106. LSM106 can be used to estimate the potential future impacts on an individual's health, determined from the individual's current lifestyle, which includes a variety of lifestyle-related risk factors. Such negative and / or positive factors may include, for example, alcohol consumption, smoking, drug use, exercise, diet, etc.

[0029] While not a clinical tool itself for identifying and quantifying risk factors as causes of overall health risk, this application works to derive multiple measurements. These measurements are used to calculate relative measures representing an individual's current health status, whether combinatorial or otherwise, as well as the best possible ways to improve that health status. Regardless of the possible direction of causality, the presence or absence of specific measurable factors in an individual that are known to be quantitatively and significantly correlated with health risk is paramount in determining the representation of their current state.

[0030] In one or more implementations, MHM102 can be derived from estimates of cardiovascular, cancer, and other risks associated with measurable parameters such as age, sex, blood pressure, weight, and lipid levels. These factors are associated with the most common vascular, cancer, and other risks. As an example, the predictive accuracy of many existing risk functions (e.g., Framingham, AHA-ASCVD) that use categorical values ​​for one or more binary, risk-modifying diseases such as hypertension as input can be improved by replacing the categorical factors with appropriate probabilities of developing such diseases over a period of time. Thus, instead of assigning a YES / NO (1 / 0) value to type 2 diabetes, for example, one could assign the probability of a person developing diabetes over time from another model that quantifies the risk of developing diabetes. If the person already has diabetes at baseline, this model will produce YES (1), but in all other cases, it will produce a non-zero continuous value.

[0031] In some implementations of this application, the input is transformed (for example, logarithmically) according to the individual data characteristics. In other cases, interaction terms, such as the product of two inputs, are included as additional features. In yet another case, more complex features are constructed, such as from basic inputs (for example, age or blood pressure) or from complex combinations of basic inputs.

[0032] The overall MH score 110 can broadly include factors covering a set of disease endpoints, as well as associated risk factors, including precursor and central risk factors. Precursor risk factors include risks associated with major diseases such as type 2 diabetes and hypertension, while central risk factors include risks associated with major vascular and cancer-related mortality. Furthermore, modulator models can operate to modify (or modularize) the output of the model associated with each factor. For example, a modulator model can operate to scale risk using factors that may not be explicitly included, such as physical activity, psychosocial status, alcohol consumption, or specific aspects of nutrition. Modulators can include predictors derived from QLM104 and LSM106, each of which can include multiple separate models that are combined to produce a single event risk estimate.

[0033] More specifically, exemplary risk factors included in MHM102 as precursor risks may include indicators of diabetes, hypertension, chronic kidney disease, and metabolic dysfunction. Exemplary modulators used in MHM102 may include alcohol and coffee consumption, physical activity, nutrition, resting heart rate, heart rate recovery, smoking cessation, and emotional state. Exemplary diseases and endpoints in MHM102 may include cardiovascular risk (including common cardiovascular diseases, coronary heart disease, congestive heart failure, and myocardial infarction), cerebrovascular risk, and broad-spectrum cancer risk.

[0034] MHM102 can be constructed to guarantee a maximum prediction period of a nearly constant duration, such as 10 or 15 years, and can include, among other things, various lifestyle factors such as physical activity, smoking, and nutrition. Furthermore, this application uses machine learning techniques to optimize one or more models. Many of the individual models, such as the various models used to estimate the risk of diabetes, fit data using certain conventional techniques, such as Cox proportional hazards models or one of several types of logistic regression models. In one or more implementations of this application, the training / test cycle is carried out in the same way as in known machine learning techniques. More specifically, fitting is performed using a randomly selected percentage of records (e.g., 70%), and testing is performed using the remaining 30% of records. Multiple individual models can be used in combination with multiple models to form a single predictor.

[0035] Furthermore, prior risks may include modulators as modifiers. Input data for a modulator model may come from several sources, including data received in response to prompts provided by one or more computing devices, which represent family, demographic, and metabolic values. In addition, inputs may include parameters derived from models using these inputs, data derived from QLM104, and information provided in substantially real-time from techniques for tracking endogenous medical and / or exogenous behavioral activities.

[0036] Referring to Figure 1B, as an example, Model ("A") estimates the risk of cardiovascular death. Model A is provided with a set of inputs ("S"). A modulator risk model ("A'") has a set of inputs ("S'") that is not a subset of S. In practice, S'-S ≠ Φ. A' modifies the output of Model A. As a further example, S (i.e., the input set of Model A) does not include resting heart rate, even though heart rate is a significant risk factor in the prediction provided by Model A. Continuing this example, Model A' could be a simple model that outputs a prediction of cardiovascular death based solely on resting heart rate. By modulating Model A with Model A', the present application improves the predictive capability of Model A. Technically, the modulation is preferably performed at the level of the hazard function. Thus, the output of Model A is modulated to include the additional inputs. As shown in Figure 1B, the combination of Model A, Model A', and modulation functions functionally as a new risk model.

[0037] MHM102 operates effectively and accurately by using user-generated input, which is generally easy and inexpensive to obtain. Recognizing the inherent limitations of component models that would otherwise be available, this application automatically substitutes various values ​​associated with risk factors that are difficult or expensive to obtain. This allows the user to obtain an MH score 110 by supplying only a minimal set of values ​​(e.g., age, sex, height, and weight), with other values ​​associated with one or more risk factors automatically substituted. To improve the accuracy of MHM102 (and consequently the MH score 110), an interactive process and mechanism is provided to encourage the user to provide additional data that makes each score more accurate and meaningful.

[0038] More specifically, in one or more implementations, this patent application incorporates an imposition engine that operates to hierarchically and in a specific order to substitute missing values. For example, an imposition stream is provided at a given level where all data is available, including data previously substituted to substitute an input or one or more other missing values. Of course, one of the usual skills is to recognize that the overall accuracy of a given substituted value may depend on the accuracy of one or more previously substituted values. Thus, the order in which values ​​are imported is quite relevant. Given a set of inputs such as age, sex, height, and weight, additional inputs can be substituted using a population model that includes one or more additional inputs that approximate the given set of inputs. The order of imposition can be important because the order may determine the overall accuracy of the imposition procedure. For example, substituting total cholesterol and then fasting blood glucose may not necessarily yield the same result as substituting these variables in the reverse order. The optimal order for a particular dataset can be obtained using explicit calculations and standard optimization procedures.

[0039] A quantitative determination of the accuracy of the imputation engine in this application can be obtained by comparing an unoptimized implementation of MHM102 for a substantially complete set of variables (hereinafter referred to as "complete") with the corresponding results on the same dataset, but with the input limited to age, sex, height, and weight as required by the engine (indicated as "4p" in the table below and the graph shown in Figure 1C). The comparison represents the case of a substantial input set with only the missing variables substituted, and the case where all input variables except age, sex, height, and weight are substituted by the imputation engine. For the purpose of this comparison, ROC curves are shown to contrast the accuracy (sensitivity) in identifying true positive cases with the accuracy (specificity) in cases where survivors are misidentified. In the graph shown in Figure 1C, the thin diagonal line corresponds to the results of a random classifier, and points on this diagonal line indicate, for example, a classifier that is better than a random classifier. A common measure of classifier quality is the area under the corresponding curve (AUC or C statistic). More specifically, the AUC values ​​corresponding to the predictions of all-cause mortality in the two cases of the National Health and Nutrition Examination Survey (NHANES) III study, without additional optimization, are shown in the table below.

[0040] [Table 1]

[0041] In the example graph shown in Figure 1C, the ROC curves compare the predictive power of all-cause mortality associated with the imputation engine using the complete available input set or only the four minimum inputs (age, sex, height, and weight). In the example above, the global impact of the maximum number of missing values ​​is approximately a 4% decrease in AUC, which represents a loss of predictive power. In one or more implementations, the derived value representing the total risk of all-cause mortality is converted to a numerical score in the range of [0, 1000] (e.g., MH score 110), where zero means complete absence of health status and 1000 means perfect (unattainable) health. Furthermore, another score closely related to the MH score 110 is presented to the user with or without the MH score 110, and this is a relatively simple and intuitive score. The simplified score is based on the user's estimate of all-cause mortality risk and represents relative life expectancy, which further represents the impact of various lifestyle factors available to the engine. This value can be expressed as the number of "Quality Years Gained". Therefore, this application further protects confidential and private user health information by concealing the confidential health-related information received (and / or substituted) while simultaneously displaying or providing a simplified score, with or without the MH score 110.

[0042] This specification recognizes that age is a significant risk factor for almost all diseases. Unmodified health score values ​​derived from general risk models are not accurate enough to be used as a universal metric for representing relative health across all age groups. Therefore, age can be used as a modifier to influence the output produced by one or more models. Similarly, sex is used here as a modifier for one or more health risk models to eliminate inaccuracies. For example, an unmodified health model based solely on risk is modified as a function of sex to treat men and women differently. Thus, the derived MH score 110 of this application is substantially independent of both age and sex as a function of an equalization mechanism that includes modifying a base score calculated using MHM 102 and equalizing that score using a model derived from a large-scale population survey.

[0043] The graph shown in Figure 1D illustrates the distribution of unequalized scores (plotted in the upper panel) and the corresponding equalized scores (plotted in the lower panel). As indicated in the legend, the distribution represents age in chronological order. Unequalized scores are strongly influenced by age, while the equalized versions are shown to be largely age-independent where necessary. However, for a given gender and age range, the variation in actual scores for a given gender and age group still measures the relative health status of users within that gender and age range.

[0044] Referring to the graph shown in Figure 1D, the distribution of unequalized scores is shown in the upper panel, and the corresponding distribution of equalized scores is shown in the lower panel. As shown in the legend, the gradient in the distribution represents age in chronological order. It is recognized herein that the determination of whether MHM102 provides an accurate general model of relative health, and in particular whether it is modified according to the teachings herein, can be confirmed through validation. Accordingly, two types of retrospective validation methods are implemented herein: i) comparison of MH score 110 with established, prospectively validated general and / or specific health risk models, such as the Framingham Cardiac Associated Model, or ii) direct comparison of MH score 110 with linked mortality data from one or more large studies.

[0045] The validation results show that (1) the MH score 110 is consistent with well-known and previously validated cardiovascular risk models, (2) when functioning as a classifier of disease and mortality, the MH score 110 is very accurate, considerably more accurate than the score produced by the single model used to construct it, and further, (3) the MH score 110 classifies the mortality risk of various cancers very accurately, even without directly implementing a cancer risk model in MHM102. Thus, MHM102 can function as a general health model assessment tool more than it would be estimated in other ways in light of its constituent model. Perhaps more importantly, MHM102 strongly suggests that the variance in the set of risk factors included in MHM102, many (perhaps most) of which are modifiable by lifestyle changes, indicates a general decline in health beyond the risk of death attributable to cancer and / or vascular disease.

[0046] As described herein, one output of MHM102 is a score scaled to the range [0,1000], where the upper end represents perfect (unachievable) health. In one or more implementations of this application, the MH score 110 can be derived according to a two-step process. First, the overall raw score is obtained by combining survival probabilities with other scores generated by individual risk models. The combination of survival probabilities can be done using one of several procedures such as the direct arithmetic mean, the Euclidean mean, or by using the individual probabilities themselves as composite features for use in a quadratic fitting procedure, such as building a Cox proportional hazards model using these features. The calculated raw score represents the resulting survival probability, which is a number in the range [0,1]. The calculated raw score is then converted to a value in the range [0,1000] using a parametric nonlinear mapping function. The parameters of the mapping function are tuned so that the score is linear and has a relatively high slope in the region of a typical polling score. Furthermore, the scores asymptotically slope at the lower and upper bounds of the score distribution. Therefore, the mapping function is designed to respond strongly to changes in the typical score range.

[0047] Returning to the Quality of Life Model ("QLM") 104, one or more implementations of this application used data from a sample of over 4,000 surveys conducted by international organizations for model construction. Of the original sample, nearly 3,250 complete surveys were used to construct the QOL model. Data from the remaining incomplete surveys were excluded from model construction but were still used for basic statistical quality metrics. Unlike conventional questionnaire-based methods that can assign ad-hoc weights to questions based on researchers' beliefs about the relevance associated with each question, the QLM 104 of this application is derived directly from the data used from surveys as a causal model, including factor loadings. By deriving QLM 104 as a causal model, it becomes possible to derive key risk factors, including the levels of depression and stress experienced by individuals. The derived risk factors can then be used as input to other models. Furthermore, unlike conventional quality of life models that produce a single static value, the results of QLM 104 can be dynamically and periodically updated as a function of input from a single user. The updated QLM104 results may reflect changes in individual perceptions and general health perceptions over time.

[0048] In one or more embodiments of this application, the questionnaire associated with QLM104 comprises 25 questions designed to quantify the major emotional states of depression, hope, health, anxiety, and psychological stress. A set of relatively small sample questions represents what is included in one or more larger sets of questions and can be used to develop a relatively short instrument that can consistently quantify the major emotional states.

[0049] More specifically, QLM104 can be constructed as a causal model using inductive clustering and factor analysis. In one or more implementations, the final factor model can be constructed using a 10-fold cross-validation method with approximately 70% of one or more surveys in the training set and the remaining 30% in the test set. In one or more implementations, a set of approximately 500 surveys can be used to validate the final model. The optimal outcome set of four natural clusters or factors with self-consistent interpretation and good statistical properties includes C1=depression (α=0.81), C2=hope (α=0.89), C3=health (α=0.84), and C4=stress (α=0.80). The final factor-specific loadings were estimated as coefficients for the first major Carunenlebe component.

[0050] Unlike MHM102 and QLM104, which collectively address an individual's current health, LSM106 addresses lifestyle-related risks for estimating an individual's future health, which are significantly modifiable and actionable to improve the individual's health. This specification recognizes that several lifestyle factors can have a significant impact on a user's health. Three non-exclusive examples of such factors include smoking, lack of exercise, and malnutrition, which represent significant, modifiable health risks.

[0051] In one or more implementations, the LSM106 comprises several components, each configured to generate its own score, typically directly based on its respective risk model. Examples of components that may be included in the LSM106 include physical movement, nutrition, weight management, smoking cessation, stress reduction, sleep quality, and sleep duration. In one or more implementations, inputs associated with such components that may be included in the LSM106 are provided either through manual user input or via various electronic sensors integrated into the system, such as the Internet of Things ("IoT"). Exemplary IoT devices and / or components include, but are not limited to, smart rings, configured eyeglasses, configured contact lenses, hearing devices, clothing accessories (e.g., smart shoes, configured gloves), and other wearable devices.

[0052] For example, one or more biosensors can be used to collect health information about a user and transmit it to one or more computing devices. The biosensors can be in contact with or placed inside the user's body to measure the user's vital signs or other health-related information. For example, a biosensor may be a pulse meter, heart rate monitor, electrocardiogram, pedometer, blood glucose monitor, or other suitable device or system that can be worn in contact with the user's body to sense the user's pulse. The biosensors according to this application may include a communication module (e.g., a communication subsystem) so that the biosensor can transmit the sensed data either wired or wirelessly. The use of biosensors provides a degree of reliability by eliminating user errors associated with manually entered and / or self-reported data. Furthermore, wearable smart IoT devices, such as fitness and sleep tracking devices, are supported according to this application. Even more devices, such as embedded chips, voice-based interfaces, ultra-thin (e.g., tattoo-style) biosensors, or other interfaces, are supported and can be used to provide medical data usable for modeling and associated generated information as shown and described herein.

[0053] Furthermore, one or more computing devices, including servers, smartphones, laptops, tablets, or other computing devices, can send and receive electronic content to and from a health band worn by a user. The content may include, for example, numerical, text, graphics, images, audio, and video material. Such communication may occur directly and / or indirectly, such as between a server and the band via a mobile computing device like a smartphone, tablet computer, or other device. Alternatively, such communication may occur between a server and the health band without the use of computing devices. Therefore, in one or more implementations, the band may use hardware and software modules to collect and / or receive information, process the information, and transmit the information between the band and the server and / or between the health band and mobile devices.

[0054] The output of each component is designed to maximize user motivation and therefore increase the likelihood of improvement in one or more significant health risks. The LSM score can include a combination of various individual scores, such as physical activity, sleep quality and duration, nutritional quality, and stress reduction, among others. LSM106 can operate to generate a score for each of the components, such as weight management, sleep, and nutrition. Furthermore, LSM106 can provide an aggregate score. Component scores can be based on a double buffer mechanism with a time decay function. The scoring algorithm can use two energy repositories: a score repository and a buffer repository, where the score is calculated. When a user exercises, the energy generated is divided into these two repositories at a fixed ratio, with the majority of the energy going to the buffer repository. To simulate the realistic medical value of exercise, the energy stored in both repositories must decay over time (i.e., the positive effect of exercise in a day has a finite duration). However, if, for example, the user does not exercise the next day, the energy level in the score repository will only decrease when the buffer repository is empty. This mechanism ensures that scores do not fluctuate significantly, allowing for rest days without score penalties. This represents a medically realistic time delay between a sequence of specific activities, such as physical training, and the time it takes for such a sequence to yield measurable health benefits to an individual. Furthermore, the model associated with LSM106 can include a motivating game structure that can be directly connected to an outcomes-based framework.

[0055] This specification recognizes that exercise and other physical activities, when performed consistently and only in a delayed manner, can directly impact and potentially impact an individual's health. The various corresponding components of the LSM106 are based on models that implement these qualities in a medically realistic way. Furthermore, the components are designed to provide users with immediate positive results to initiate lifestyle changes in order to maintain motivation.

[0056] In one or more implementations, one or more components, such as those related to physical activity, have corresponding models that include two reservoirs with a fixed maximum size. For example, one of the two principal components ("H" component) directly contributes to the overall MH score of 110. The second component ("B" component) is a health buffer that gives an elastic component to the H component. For example, if a user tracks physical activity, a small portion of the score corresponding to that activity is added directly to H, and the majority of the score is added to B. However, the validity period of the score is limited and decays to zero over time. The decay rate associated with the score is nonlinear, and its value depends on the current magnitudes of both H and B, and is applied directly to B as long as B is greater than zero. When this decay reduces B to 0, the decay is applied to H, directly affecting the component's overall score. The relatively rapid decay of the health buffer provides a strong incentive for users to maintain the value of the health buffer so as not to lower their overall health score.

[0057] The mechanism associated with LSM106 provides a medically accurate measurement of user effort related to maintaining or improving good health. If not maintained through consistent activity, the value will decrease to zero. The nonlinearity of the decay function adapts to various levels of corresponding activity, being most demanding for the most active lifestyles and most lenient for less active ones. Conversely, low levels of activity will not result in a high overall score, thus rewarding higher and more sustained levels of activity.

[0058] The MHM102, QLM104, and LSM106, commonly referred to herein as the "three pillars," do not need to operate independently of each other; the output from each of the models is reflected in the MH score 110, which also reflects the interactions between the models. Maintaining and / or improving an individual's future health, as described herein, is a key feature of this application, and the MH score 110 can have individually varying sensitivity to modifiable lifestyle behaviors.

[0059] In one or more implementations, the scores associated with each of the three pillars are combined into an overall MH score 110, which can be static and / or dynamic in various implementations. The following table shows examples of the contribution rates to the MH score 110 associated with each of MHM102, QLM104, and LSM106.

[0060] [Table 2]

[0061] In addition to the effectiveness of the MH score 110 at the individual level in relation to improving and / or maintaining an individual's health, this application further includes an interactive data processing platform that includes data modeling and machine learning related to each risk factor. Commonly referred to herein as the “risk engine,” this application provides a system and method associated with health risk-based decision-making. In particular, risk factors can be quantified with highly precise specificity in a continuous distribution, and one or more cutoff values ​​associated with classification and stratification can then be defined and / or selected as a function thereof. The associated output values ​​can be used to influence third-party applications, such as in the case of insurance.

[0062] Despite the conceptual similarities between the output of the risk engine and conventional actuarial tools used, for example, in life insurance, this application has distinctive features and capabilities that are not supported or included in actuarial applications. Perhaps most notably, the output of this application, including the MH score 110, or the respective scores output as functions of MHM102, QLM104, and LSM106, and their respective associated components, are continuous measurements, as opposed to individual table-based risk estimates.

[0063] As described herein, the output is computed in relation to MHM102, QLM104, and LSM106, as well as their component models that consider a continuous probability space between 0 and 1, which is multiplied by 1000 in one or more implementation forms. Such computations offer a fundamental, industry-based paradigm shift by providing an output generated in a manner different from conventional actuarial practices, which determines discretely classified values ​​as a function of one or more decision trees, which are effectively classifiablely distributed values. This application provides indexing on a continuous scale for determining precise health conditions.

[0064] Furthermore, as described herein, this application is operable to run on partial data and implements an imposition engine to generate missing values. Obtaining each of the multiple values ​​associated with a health assessment can be burdensome, costly, inconvenient, and / or impractical. The imposition engine of this application overcomes such drawbacks associated with various health platforms, such as those in the insurance industry, by automatically generating such missing values. The advantages of the process of this application associated with assessing a user's health status are greatly improved by reducing the number of questions required of the participant and by speeding up the entire process.

[0065] Other benefits provided in accordance with this application include achieving performance benchmarks, including predicting all-cause mortality, cardiovascular mortality, and cancer mortality over 10–14 years. Other conditions such as diabetes and hypertension may be less, such as up to 5 years. Furthermore, this application provides highly personalized output values, including considering more than 100 health-related inputs representing well-being from various domains, including blood biochemistry, anthropometric measurements, well-being, stress, nutrition, medical history, sleep, and physical activity. In addition, the risk models provided in accordance with this application can be derived and validated from observations, including from studies conducted by groups of researchers internationally. Furthermore, this application supports straight-through processing ("STP"), which enables underwriting to be performed substantially in real time, for example. Thus, this application enables users to provide liquid-free underwriting, which uses an imputation engine to substantially automatically calculate risk assessments even with limited amounts of data. Furthermore, the features and operations described herein support individual-specific calibrations and modulations related to pay-as-you-live ("PAYL") and premium calculations.

[0066] This specification recognizes that long observation periods associated with one or more studies often do not yield more accurate predictors. This is due to a variety of factors, at least in part, that may or may not be taken into account by a given model, and that change over time. Unraveling and understanding the effects of such changes is extremely difficult, or even impossible. For example, unpredictable lifestyle and / or environmental factors change over time, as do systems and methods for treating diseases. Furthermore, environmental conditions such as noise, light, air pollution, and adverse weather conditions affect human health.

[0067] The risk engine of this application improves predictive accuracy by defining boundaries such as 10-15 years, thereby optimizing the balance between accuracy and inevitable changes in many parameters that are not always considered or omitted in known model-building processes. Some insurance companies use long-term risk estimates to generate products, but it is recognized that such estimates can be inaccurate because, for reasons identified herein, as well as because certain risk factors (which change over time) are unknown at baseline.

[0068] In one or more implementations, the risk engine of this application estimates a number of increasing risks. For example, some non-limiting lists of risks include type 2 diabetes, hypertension, metabolic syndrome, indicators of metabolic dysfunction, chronic kidney disease, congestive heart failure, stroke, myocardial infarction, coronary heart disease, cancer, chronic obstructive pulmonary disease, neurological dysfunction, and dementia. The accuracy of estimates associated with such risks depends, at least in part, on the completeness of the input data. The risk engine of this application has a clear advantage over known systems by utilizing values ​​generated by an imputation engine, thereby enabling the risk engine to function accurately and generate reasonable estimates of risk using minimal input data, including date of birth, sex, height, and weight.

[0069] Therefore, the risk engine of this application generates risk estimates and scores based on data and models from a large set of cohorts. This makes it possible to generate risks and scores that are generally valid for everyone worldwide. The risk estimates generated in accordance with this application can be optimized for a given population.

[0070] The risk engine of this application can be used to influence third-party computing platforms, such as computing platforms associated with health insurance, life insurance, and reinsurance providers. For example, risk estimation in existing insurance products, including robust actuarial and other methodologies, cannot achieve the precision and accuracy available according to the teachings herein. Nevertheless, this application is not intended to replace any methodologies or business practices currently practiced for the insurance industry. Instead, this application, including the risk engine, provides a value-added service to the underwriting process by generating more accurate risk estimates in specific key areas at a lower cost and with greater convenience. This is particularly true for conventional methods that rely on life tables at the heart of risk estimation.

[0071] In one or more implementations, a population-specific optimal model can be generated in a dataset split into two subsets: a modeling dataset and a prediction dataset. The modeling dataset can be a subset of the optimization dataset containing the target field data. Furthermore, during the model building process, this dataset is split into a training set from which a candidate model is derived and a test set from which the model derived in the training set is tested. Robust model building can proceed with the N-fold cross-validation process, in which multiple training / test set pairs are randomly selected from the modeling dataset to build a model that maximizes accuracy while maximizing generalizability and minimizing the possibility of overtraining. Typically, the training and test sets contain about 70%–75% of the records from the complete dataset created by the insurance company. Finally, a third set, the prediction set (often also called the evaluation set), consisting of the remaining modeling dataset records, is set aside to evaluate the properties of the final model in an unbiased manner. Both splitting into training and test sets, and the use of N-fold cross-validation, are data science techniques that can be used alternatively or in combination.

[0072] Furthermore, the prediction dataset, sometimes called the blind set, contains a subset of records from the full dataset with the target field removed and set aside. Once the optimal model is generated using the data in the modeling dataset, the prediction dataset is scored using the optimal model. The score, representing the probability of a positive target value, can be used to assess the accuracy of the modeling procedure, for example, including all the precision measures reported. In addition, population-specific optimizations are supported, including endpoint probabilities at a fixed time.

[0073] Referring to Figure 2, this is a diagram of an exemplary hardware configuration that operates to provide a system and method disclosed herein and generally designated as a health platform 200. The health platform 200 preferably comprises one or more information processing devices 202 coupled to one or more user computing devices 204 via a communication network 206. User computing devices may include, for example, mobile computing devices such as tablet computing devices, smartphones, and personal digital assistants. Furthermore, it may include a plurality of sensing devices that transmit various health-related information to the computing devices, either directly or indirectly included in the information processing devices 202.

[0074] The information processing device 202 preferably includes all databases necessary for this application, including image files, metadata, and other information. However, the information processing device 202 is intended to be able to access any necessary databases via the communication network 206 or any other communication network accessible to the information processing device 202. The information processing device 202 can communicate with the device containing the database using any known communication method, including direct serial, parallel, and USB interfaces, or via a local or wide-area network.

[0075] As shown in Figure 3, the functional elements of each information processing device 202 or workstation 204 preferably include one or more central processing units (CPUs) 302 used to execute software code to control the operation of the information processing device 202, read-only memory (ROM) 304, random access memory (RAM) 306, one or more network interfaces 308 for sending and receiving data to and from other computing devices via a communication network, storage devices 310 such as hard disk drives, flash memory, CD-ROMs or DVD drives, databases and application code, one or more input devices 312 such as keyboards, mice, or trackballs, and a display 314.

[0076] The various components of the information processing device 202 do not need to be physically contained within the same chassis, nor do they need to be located in a single location. For example, as described above with respect to a database that can reside on a storage device 310, the storage device 310 may be located separately from the rest of the information processing device 202 and may even be connected to the CPU 302 via a communication network 206 through a network interface 308.

[0077] The functional elements shown in Figure 3 (indicated by reference numerals 302-314) are preferably of the same category as the functional elements that are preferably present in the user computing device 204. However, not all elements are required to be present, and for example, the capacity of storage devices and various elements is adjusted to meet the expected user demands. For example, the CPU 302 in the user computing device 204 may have a smaller capacity than the CPU 302 present in the information processing device 202. Similarly, the information processing device 202 is likely to include a storage device 310 with a much larger capacity than the storage device 310 present in the workstation 204. Of course, those skilled in the art will understand that the capacity of functional elements can be adjusted as needed.

[0078] Figures 4–8 show exemplary screen displays associated with a graphical user interface capable of providing the functions described herein. The screen displays shown in Figures 4–8 are associated with a mobile computing device such as a smartphone and include, for example, inputting and reviewing data submitted by the user via a graphical user interface through screen controls. As shown in the display screen 400 shown in Figure 4, graphical screen controls are selectable for the currently logged-in user. Such controls include options for reviewing the user's profile information, the current point status, and results.

[0079] Figures 5A and 5B show an exemplary data entry display screen 500 that includes options associated with viewing progress related to the user's MH score 110, as well as options for setting parameters related to nutrition, addiction, activity, mindfulness, and sleep, all of which contribute to the MHM 102 as shown and described herein. When any of the options are selected, most of the values ​​associated with the parameters are self-assessed, aiding in the automatic generation of the MH score 110. Furthermore, prompts are included for the user to provide responses to more detailed prompts related to health and lifestyle.

[0080] Figures 6A–6C show a series of exemplary data input display screens 602, 604, 606, 608, 610, 612, and 614 (collectively referred to herein as “cards”) that are provided to the user in response to the user’s selection, depending on the user’s selection of each option (in this case, nutrition) from the display screen 500 (Figure 5). Figures 6A–6C represent exemplary implementations of the present application, and exemplary cards 602–614 provide interactive prompts for user responses regarding details of the user’s food and eating habits in order to obtain information on LHM106 and to suggest simple (e.g., 3-day) miniature (“mini”) goals, such as adjusting the user’s food intake in an organized plan and improving the user’s MH score 110. In one or more implementations, the selection and display of cards are based on complex artificial intelligence and machine learning algorithms designed to maintain and / or improve the user’s MH score 110.

[0081] As shown in Figures 6A to 6C, the user is asked to identify whether they primarily cook with olive oil (card 602), the type of meat they eat, red or white (card 604), the type of fruit they eat (card 606), how often they eat more than two servings of fruit per day (card 608), the number of soft drinks they drink per day (card 610), and how many servings of nuts they eat per week (card 612). In this way, a fun, simple, and convenient interface is provided for the user to submit health-related (in this case, case, food consumption) data. It is recognized herein that providing other information related to such food and habits may be unpleasant for the user, and that this is mitigated or eliminated by the convenient interface provided in accordance with this application. Furthermore, in accordance with this application, improvements to food intake habits can be suggested, including setting easily achievable and moderate goals, such as a three-day mini-goal to reduce sweets consumption and / or increase healthy food consumption. For example, upon completion of cards 602-612 and / or one or more goals, the user is given positive reinforcement, such as unlocking a “nutrition badge” that rewards the user. In addition to virtual rewards, as shown in card 114, monetary or other reimbursable rewards may be provided in accordance with this application for the user's completion of various tasks, such as providing details of the user’s nutrition, indulgence, activity, mindfulness, and sleep habits, as well as detailed health information. Furthermore, such rewards may be provided when improvements occur in the user’s health-related habits and are identified in accordance with this application. Those skilled in the art will recognize that the reward systems provided herein are powerful tools for motivating users and increasing long-term user engagement.

[0082] Exemplary cards 602–614 are illustrative, and the present application supports many such interactive navigation and self-prompting cards for receiving information in each category for modeling and generating information, as shown and described herein. For example, navigation modules and associated cards are supported for addiction (e.g., smoking, drinking, etc.), activity (e.g., running, sports activities, etc.), quality of life (e.g., well-being), and sleep. Furthermore, cards are automatically provided to the user in each of each category to generate manageable tasks and encourage behavioral changes such as reducing addiction, increasing activity, getting better quality sleep, and improving the user's mental health.

[0083] In addition to behavioral information, this application provides cards for configuring the automated collection and verification of health-related information. Figure 7 shows an exemplary data entry card 700 provided to a user who has selected options to provide additional, more detailed information related to the user's health. In the example shown in Figure 7, the user is prompted to select whether or not they have been diagnosed with each cardiac condition, such as cardiac hypertrophy, hypertension, arrhythmia, and heart failure. As a function of artificial intelligence, the user is prompted to provide a large amount of information related to their health through a series of engaging cards. As described herein, this application includes features to ensure that the user's health-related and behavioral information is kept strictly confidential, and privacy is ensured as a function of security measures configured in the computing software and hardware modules. Furthermore, raw data elements received and / or substituted in accordance with this application are maintained in a background process, and only anonymous values ​​generated as a function of modeling as shown and described herein are viewable. Behavioral and medical values ​​are converted into masked composite numerical values ​​(e.g., MH score 110), and the values ​​themselves are kept confidential and not made public. Therefore, the underlying medical and behavioral parameters are protected, and third parties cannot reverse the calculations to determine those parameters based on values ​​generated according to the model. In one or more implementation forms, each user and / or authorized person with sufficient access control privileges can access the underlying raw data elements.

[0084] Figure 8 shows an exemplary display screen 800, commonly referred to herein as the “Wheel of Life,” which is provided to the user for reviewing their progress and includes the user’s current health score and the target value the user is striving to reach. The Wheel of Life shown on the exemplary display screen 800 includes the categories described herein, each relating to MHM102, QLM104, or LHM106, and MH score 110, including details on activity, nutrition, addiction, mindfulness, sleep, and medical care. In one or more implementations, the categories displayed on the display screen 800 are configured as selectable options. During use, the user selects a category from which one or more cards can be displayed for viewing and / or entering information. Furthermore, based on artificial intelligence, one or more of the respective categories can be automatically highlighted and / or selected to review information relevant to the user’s health and / or to prompt the user to enter partially entered, missing, or updated information. Thus, the Wheel of Life serves not only as an informational tool but also as a hub for the user to review and process information.

[0085] Therefore, as shown in the sample data input display screens in Figures 4 to 8, users can submit information related to health and well-being through a convenient, engaging, and enjoyable interface. Furthermore, users can monitor their progress to maintain and / or improve their health, including as a function of a health score. In addition, data received according to the user interface can be transmitted and used in accordance with this application in relation to modeling and output, as shown and described herein.

[0086] In one or more implementation forms, the interactive platform provides a simple stepping stone toward a comprehensive view of human health and coaching for a conscientious and healthy lifestyle. The interactive platform can be implemented as a smartphone app integrated with the teachings described herein. This platform motivates people to be physically active, including increasing their average daily step count, especially for the average individual (who is not an athlete).

[0087] In one or more implementations, a calculated score (referred to as the “step score”) is generated, which is a risk-based function that uses information associated with physical activity and body mass index as key values. This value can be used, for example, to generate the probability of a 10-year risk of death from all causes. Just as additional physical activity and lower obesity reduce the risk of early all-cause mortality, the step score of this application increases under such conditions.

[0088] Preferably, the physical activity component of the step score is based on the daily energy expenditure from taking steps. The input is calculated from the number of steps, step frequency, terrain grade, and stride length, while simultaneously providing the user with immediate satisfaction by assigning more weight to the present time as opposed to the past through an exponential moving average. Furthermore, since the input is generated as weekly MET hours by the exponential moving average, local variations in the daily energy input are smoothed out, thereby generating a medically meaningful estimate of energy expenditure.

[0089] In one or more implementations, the risk model of this application may include a step score. For example, central obesity, using Body Mass Index (BMI) as a proxy, plays two roles in the step score: firstly, through the all-cause mortality risk model described below, which adjusts the physical activity portion of the score; and secondly, as an energy amplifier to encourage users with some degree of obesity. This booster function produces the greatest amplification at low energy expenditure and decreases exponentially as users with higher BMI increase their daily energy expenditure.

[0090] Furthermore, relative life expectancy estimates are used in conjunction with step scores as an additional incentive for users. The impact of both physical activity and obesity on longevity is considered. Estimates associated with relative years obtained from increased physical activity and relative years lost due to obesity are further used and combined. For example, the resulting relative years can be estimated from physical activity based on inactivity-based mortality estimates, and the years lost due to obesity can be estimated by comparing them to a broad ideal BMI centered around 23.5 kg / m2.

[0091] Therefore, this application is designed and constructed to induce lasting behavioral changes in users. Various incentives, such as reduced insurance premiums and / or built-in stores, provide external motivation for users to walk more. Features associated with step scores can be self-determined and function as a reward system. For example, in relation to a game implementation, value can be earned through playing the game and activities (e.g., walking). Thus, in one or more implementations, this application is configured as a location-based game that users choose to play. This is more likely to bring about behavioral changes because it offers interesting choices rather than prescriptions or obligations.

[0092] Gamification can provide a platform around a level system from level 1 to 15. To level up, players need to earn experience points, or XP for short. Since 1 XP is earned at each step, players can progress simply by walking. As levels increase, additional features are unlocked, including step scores (level 3), quests (level 2), teams (level 5), and abilities and rewards (levels 8, 10, and 15). This transforms the level system into a powerful, progress-dynamic system where every step is important. Players earn XP by walking, completing quests, participating in team battles, etc., which increases their level and unlocks new features. In the later stages of the game, players can earn coins that can be used in the rewards shop and other places. Furthermore, step scores can be used to offer discounts on insurance premiums.

[0093] In one or more implementations, the step score is unlocked at level 3. The step score includes two strong motivation subsystems: a boost and a reservoir for overweight players. Players fill the reservoir when they walk, which prevents the step score from decreasing if the player does not walk as much as usual for a day or two. Alternatively, it provides a strong incentive to walk more after the step score has decreased and the reservoir has been depleted, by encouraging further walking.

[0094] For those who aren't particularly interested in games, the use of step scores and the associated premium reductions is considered a major attraction. For users who enjoy playing, it's offered in the form of casual games and a rewards shop.

[0095] In one or more implementations, the quest system is unlocked at level 2. Players can walk to quests displayed on the map in places of interest, such as shops and public spaces. Quests ask them to complete tasks, such as visiting a set of nearby locations. These particular quests allow players to rediscover their neighborhood in new ways while accumulating XP by walking. Completing quests grants additional XP, allowing players to level up more quickly.

[0096] At level 5, players can unlock teams and join specific teams (for example, red or blue). Once a team is selected, players can participate in daily battles for dominance by collecting gems displayed on the map and accumulating steps. During gameplay, players walk around to pick up gems. When a battle ends, all active players on the winning team receive additional XP. Thus, the game is social, emphasizing relevance and collaboration, meaning actions that have meaning beyond the individual.

[0097] Furthermore, the ability system is designed to give players specific boosts that they must discover. In one or more implementations, once a player reaches level 15, gaining XP becomes meaningless. At this point, the reward system can be fully unlocked. Gaining XP from quests and team victories is replaced by gaining coins, which are a virtual currency. Players can use the coins accumulated over time to purchase rewards from the built-in store. The rewards provide a modest, long-term incentive.

[0098] Figures 9A and 9B illustrate a series of operations in an exemplary user interface 900 according to the implementation of this application. The sequences shown in Figures 9A and 9B illustrate the data input, editing, processing, and conversion processes in the implementation of this application. It is recognized herein that the automatic conversion of data from one data format to another is important, including supporting foreign data platforms and transmitting information across continents.

[0099] The data values ​​shown in Figures 9A and 9B can be provided to Interface 900 by the user via graphical screen controls included in Interface 900. In the examples shown in Figures 9A and 9B, user credentials such as tokens are issued by their respective key providers in Section 902. A forecast time frame (e.g., 10 years) is provided, as shown in exemplary block 904, and data field names, field attributes (e.g., whether a value is required, data description, minimum and maximum allowed values), and data values ​​are provided to Interface 900, as shown in exemplary blocks 906 and 908. Inputs can be associated with MHM102, QLM104 (not shown), and / or LSM106 (not shown). Blocks 904, 906, and 908 may include graphical screen controls, such as slider controls and drop-down lists, for the user to input or adjust values.

[0100] In one or more implementations, after information is input to interface 900, a data communication session is established and the data can be provided to information processing device 202. JavaScript can be used to generate REST parameters and instantiate a JSON object (block 910), which is passed to information processing device 202 for processing. For example, information processing device 202 parses values ​​from the JSON object to compute outputs related to MHM102, QLM104, and LSM106, as well as their component models considering a continuous probability space between 0 and 1. In another example, information processing device 202 does not perform processing or computation directly but accesses a risk engine API, i.e., another computing device (including another information processing device).

[0101] Figure 9B shows the sequence and continuation on interface 900 after the values ​​calculated by the information processing device 202 are returned and displayed on interface 900. For example, blocks 912 and 914 show the output representing the 10-year risk and other values ​​from MHM102, such as the input values ​​and / or values ​​that were not input but were substituted.

[0102] It should be understood that the specific combinations of applications shown in Figures 9A and 9B, as well as the associated web servers, are just one of several technical ways to access the Risk Engine API. For example, spreadsheets, data management systems (e.g., insurance policy management applications, or health management systems) can be connected to the Risk Engine API. Such connections can be provided in automated ways, such as automated workflow management and rule-based processing. Furthermore, information can be provided from one or more remote sources, such as devices that provide CSV or other well-formatted data files. In one or more implementations, a device can connect to the Risk Engine API and then pass requests based on information that is converted, for example, to JSON and passed to the Risk Engine API.

[0103] Therefore, as shown in the exemplary sequences in Figures 9A and 9B, client-side editing is supported (for example, via blocks 904, 906, and 908), which can occur with or without an active data communication session with the information processing device 202 or other suitable device. The data associated with the editing is passed to the information processing device 202 via parameter data contained in a JSON object. The information processing device sends the results of MHM102, QLM104, and LSM106 back to interface 900, where the information is converted for display by one or more computing devices, for example. Furthermore, the information can be downloaded in a file format such as CSV.

[0104] Therefore, this application engages and empowers users for sustained behavioral changes that support long-term health improvement. Automated displays related to the user's well-being and health are provided in virtually real-time. Furthermore, personal goals are defined to improve the user's health, and information is automatically generated and provided to enable long-term health improvement and a healthy lifestyle. This application includes modules that interface with numerous IoT and computing devices, which support tracking, verification, and aggregation of information from various sources for modeling and provisioning to a single value. Using artificial intelligence and machine learning, the user's health is enhanced through an interactive computing platform that motivates the user to engage in a healthy lifestyle to improve or maintain their overall health.

[0105] The use of ordinal numbers such as "first," "second," and "third" in the claims to modify claim elements does not, in itself, imply priority, order, or temporal order in which the actions of one claim element are performed over another, but rather is used solely as a label to distinguish one claim element with a particular name from another element with the same name (however, because ordinal numbers are used).

[0106] Furthermore, the expressions and terms used herein are for illustrative purposes only and should not be considered limiting. The use of “including,” “comprising,” “having,” “containing,” “involving,” and variations thereof herein means that they encompass the items described therein and their equivalents, as well as any additional items.

[0107] This specification describes specific embodiments of the subject matter described herein. Other embodiments are within the scope of the following claims. For example, the actions described in the claims may be performed in a different order and still achieve the desired results. As an example, the process shown in the accompanying drawings does not necessarily require the specific order or sequence shown to achieve the desired results. In certain implementations, multitasking and parallel processing may be advantageous. [Explanation of Symbols]

[0108] 102 Metric Health Model (MHM) 104 Quality of Life Model (QLM) 106 Lifestyle Models (LSM) 110 Metric Health (MH) Score 200 Health Platforms 202 Information Processing Equipment 204 User Computing Devices 204 Workstations 206 Communication Networks 302 Central Processing Unit (CPU) 304 Read-only memory (ROM) 306 Random Access Memory (RAM) 308 Network Interfaces 310 Storage Devices 312 Input Devices 314 displays 400 display screens 500 Data Entry Display Screen 602, 604, 606, 608, 610, 612, 614 Data Input Display Screens 602-614 Cards 700 Data Entry Cards 800 display screens 900 User Interface Section 902

Claims

1. A computer implementation method for integrating modulated outputs from each of several models in order to quantify the factors that generate multiple values, each of which lie within a continuous distribution, wherein each discrete category associated with each of the values ​​represents the likelihood of future occurrence, and the method A step of quantifying each of a plurality of endpoints that contribute to the current state and the likelihood of the future occurrence by a first data model running on at least one computing device, wherein each quantified value of each quantified endpoint is calculated within a continuous distribution, A step of generating values ​​by a second data model running on at least one computing device, each representing a mode of at least one current state that influences the likelihood of the future occurrence, wherein each of the generated values ​​lies within a continuous distribution. A third data model running on at least one computing device identifies individual factors among a subset of the endpoints and / or the individually modifiable aspects associated with them, and generates values ​​in a continuous distribution representing each of the factors. A step of modulating at least one value, which is quantified by the first data model, generated by the second data model, and / or identified by the third data model, in order to scale a value representing at least one aspect associated with the possibility of the future occurrence, by a modulation model performed on the at least one computing device, wherein the modulation step is based on at least one factor derived from at least one of the first data model, the second data model, and the third data model, A step of using at least one of artificial intelligence and machine learning provided on the at least one computing device to integrate at least two of the values ​​associated with each of the multiple continuous distributions, and selecting each discrete category associated with the integrated value, wherein each of the integrated values ​​and the selected category represents the likelihood of future occurrences. A computer implementation method comprising the following features.

2. The method according to claim 1, further comprising the step of substituting at least one value not included in the set of inputs used by the first data model using the at least one computing device.

3. A step of substituting a previously substituted value or at least one other value not included in the quantified value associated with a previously quantified endpoint by the at least one computing device, the step further comprising: the at least one other substituted value depending on at least one of the previously substituted values; The method according to claim 2, wherein at least one of the substituted other values ​​lies within a continuous distribution.

4. The method according to claim 1, further comprising the step of recalibrating at least one of the first data model, the second data model, and the third data model as a function of information received over time or information received from multiple data sources.

5. The method according to claim 1, wherein the first data model uses each of the endpoints as an input function in the fitting procedure.

6. A step of configuring a user computing device using a software application that provides a graphical user interface on the user computing device, wherein the graphical user interface receives at least some of the values ​​and embodiments from a user operating the user computing device. The steps include: receiving at least one of the values ​​and embodiments from the user computing device via a data communication session using the at least one computing device; The steps include transmitting to the user computing device the quantified values ​​associated with at least some of the endpoints, the quantified values ​​associated with at least some of the embodiments, and the generated values ​​associated with at least some of the factors from the first data model, the second data model, and the third data model, respectively. Furthermore, The user computing device is controlled by the software application, The method according to claim 1, further configured to display the received value received from the at least one computing device.

7. The method according to claim 1, comprising the steps of configuring a user computing device using a software application that provides a graphical user interface on the user computing device, the graphical user interface prompting the user to periodically and periodically input a value associated with the factor, and further, the graphical user interface automatically providing an interactive display screen if no value associated with the factor is received after a previously received value.

8. The method according to claim 1, wherein at least one of the first data model, the second data model, and the third data model comprises selections of at least two other data models.

9. The method according to claim 1, wherein the value is calculated in the continuous distribution as a function of parametric nonlinear mapping.

10. A computer implementation system for integrating modulated outputs from each of several models in order to quantify the factors that generate multiple values, each of which lie within a continuous distribution, wherein each discrete category associated with each of the values ​​is selected to represent the likelihood of future occurrence, and the system A first data model, which runs on a computing device, that quantifies each of a plurality of endpoints that contribute to the current state and the likelihood of the future occurrence, wherein each quantified value of each quantified endpoint is calculated within a continuous distribution. A second data model, which runs on at least one computing device, that generates values ​​representing at least one aspect of the current state that influences the likelihood of the future occurrence, wherein each of the generated values ​​lies within a continuous distribution; A third data model, which runs on the at least one computing device, that identifies individual factors among a subset of the endpoints and / or the individually modifiable aspects thereof, and which generates values ​​within a continuous distribution representing each of the factors. A modulation model performed on the at least one computing device modulates at least one value, which is quantified by the first data model, generated by the second data model, and / or identified by the third data model, in order to scale a value representing at least one aspect associated with the possibility of the future occurrence, wherein the modulation is based on at least one factor derived from at least one of the first data model, the second data model, and the third data model, At least one artificial intelligence and machine learning system provided in the at least one computing device integrates at least two of the values ​​associated with each of the multiple continuous distributions and selects each discrete category associated with the integrated value, wherein each of the integrated values ​​and the selected category represents the likelihood of future occurrences. A computer-implemented system equipped with the following features.

11. The system according to claim 10, further comprising at least one computing device configured to substitute at least one value not included in the set of inputs used by the first data model.

12. The at least one computing device is further configured to substitute at least one other value that is not included in the previously substituted value or the quantified value associated with the previously substituted endpoint, and the at least one other substituted value depends on at least one of the previously substituted values. Furthermore, the system according to claim 11, wherein the other substituted endpoints are within a continuous distribution.

13. The system according to claim 10, further comprising at least one computing device configured to recalibrate at least one of the first data model, the second data model, and the third data model as a function of information received over time or from multiple data sources.

14. The system according to claim 10, wherein the first data model uses each of the endpoints as an input function in the fitting procedure.

15. When executed on a user computing device, the computing device will To provide a graphical user interface that receives at least one of the values ​​and embodiments from a user operating the user computing device, Receiving the quantified values ​​associated with at least some of the endpoints, the quantified values ​​associated with at least some of the embodiments, and the generated values ​​associated with at least some of the factors from the first data model, the second data model, and the third data model, respectively. Display the received value mentioned above. The system according to claim 10, further comprising a software application for causing the system to perform the following.

16. When executed on a user computing device, the computing device will The system according to claim 10, further comprising a graphical user interface that prompts a user to periodically and periodically input a value associated with the factor, the graphical user interface further comprising a software application that causes the graphical user interface to automatically provide or cause to provide an interactive display screen if no value associated with the factor is received after a previously received value.

17. The system according to claim 10, wherein at least one of the first data model, the second data model, and the third data model comprises selections of at least two other data models.

18. A computer implementation method for integrating modulated outputs from each of several models in order to quantify the factors that generate multiple values, each of which lie within a continuous distribution, wherein the method is A first data model executed on a computing device quantifies each of several endpoints that contribute to the current state and the likelihood of future occurrences, wherein each quantified value of each quantified endpoint is calculated within a continuous distribution. A step of generating, by a second data model running on at least one computing device, each value representing at least one current state that influences the likelihood of the future occurrence, wherein each of the generated values ​​lies within a continuous distribution; A third data model running on at least one computing device identifies individual factors among a subset of the endpoints and / or the individually modifiable aspects associated with them, and generates values ​​in a continuous distribution representing each of the factors. A step of modulating at least one value, which is quantified by the first data model, generated by the second data model, and / or identified by the third data model, in order to scale a value representing at least one aspect associated with the possibility of the future occurrence, by a modulation model performed on the at least one computing device, wherein the modulation step is based on at least one factor derived from at least one of the first data model, the second data model, and the third data model, A computer implementation method comprising the following features.

19. The method of claim 18, further comprising the step of selecting each individual category, wherein the selected category represents the possibility of future occurrence.

20. The method according to claim 18, further comprising the steps of: integrating at least two of the values ​​associated with each of the plurality of continuous distributions; and selecting each discrete category associated with the integrated values, wherein each of the integrated values ​​and the selected category represents the likelihood of future occurrence.

21. The method according to claim 20, wherein at least one of the steps of integrating at least two of the aforementioned values ​​and selecting each of the respective discrete categories is performed using at least one of artificial intelligence and machine learning.

22. The method according to claim 18, wherein at least one of the first data model, the second data model, and the third data model comprises at least one of artificial intelligence and machine learning.

Images