A machine learning architecture that determines user responsiveness based on multimodal measurement data.

Abstract

This paper provides a system and method for selecting weight-loss medications using machine learning models to address obesity. [Solution] System 100 receives physiological measurements from the user and applies the measurements to a machine learning model. The machine learning model can be trained using multiple examples, each example including sample physiological measurements from a sample user and a corresponding sample weight-loss drug administered to the sample user. The system also generates metrics indicating expected outcomes associated with the user's weight-loss drug based on the application of one or more physiological measurements to the machine learning model, and sends a message indicating the expected outcomes associated with the weight-loss drug to the user's associated user device. [Effects] It can improve the effectiveness of weight-loss drugs when dealing with obesity.

Description

Technical Field

[0001] Cross - reference to related applications This application claims the priority and benefit of U.S. Provisional Patent Application No. 18 / 962,355, entitled "Machine Learning Architecture for Determining User Responsiveness Based on Multimodal Measurement Data," filed on November 27, 2024, the entire disclosure of which is hereby incorporated by reference.

Background Art

[0002] Obesity refers to the excessive accumulation of fat accompanied by weight gain and health risks. This condition can arise from a variety of underlying factors, such as lifestyle factors, genetic properties, environmental factors, or medical conditions. For example, specific underlying factors contributing to obesity include lifestyle choices (e.g., poor diet, overeating, lack of exercise), genetics (e.g., genetic traits), psychological states (e.g., depression, stress, anxiety, overeating), medical conditions (e.g., hyperthyroidism), certain medications (e.g., steroids or antidepressants), or hormones (e.g., leptin resistance). At the molecular level, in obese individuals, fat cells become larger and increase in number, causing fat cells to store excess energy in the form of fat. Furthermore, irregular hormonal activities such as leptin resistance, which inhibits the brain's effective response to signals related to energy storage, can lead to overeating despite a large energy storage amount.

[0003] Obesity has an adverse impact on the mental health, physical health, and social well - being of the individuals affected. Obese individuals have an increased risk of various physical health problems such as cardiovascular diseases (e.g., hypertension, heart failure), cardiometabolic diseases (e.g., diabetes), respiratory abnormalities (e.g., sleep apnea, asthma), and cancers (e.g., kidney, liver, pancreas). There are also numerous mental health problems that obese individuals may face, such as depression, anxiety, social isolation, and chronic stress. The quality of life of obese individuals is affected not only by health problems but also by accessibility (e.g., mobility impairments) and economic burdens (e.g., medical costs).

[0004] Certain obesity medications can be administered to individuals to promote weight loss. However, even these medications have side effects, making it difficult for users to adhere to individual prescriptions, and as a result, they may discontinue taking the medication. Furthermore, failing to consider an individual's physiological measurements and other relevant information can lead to prescriptions that cause multiple side effects, resulting in users discontinuing the medication and ultimately little improvement in health outcomes. In addition, some individuals may physiologically tolerate certain obesity medications better than others. Administering such individuals with medications that have no equivalent effect may result in little improvement in clinical outcomes. Medication adherence is a crucial factor in successful weight loss. However, if a prescription causes multiple side effects, albeit to varying degrees of severity, individuals may struggle to adhere to the prescription and achieve their weight loss goals.

[0005] One approach to addressing medication adherence is the trial-and-error method. In this method, an individual is prescribed a dose-reducing drug, and then their response to the prescription and any side effects are monitored. However, the trial-and-error approach has several drawbacks. First, this method involves initiating a drug therapy, waiting to collect additional data to determine if it is effective, and then switching to another drug therapy if the first prescribed drug is deemed ineffective or causes side effects. This can result in prolonged periods of untreated status or suboptimal clinical outcomes. Second, each time a new drug is introduced, there is a possibility that the individual taking it may experience side effects or adverse events. Monitoring individuals for side effects and other events is complex and requires frequent checks and data collection from individuals. In addition, dose-reducing drugs have their own unique challenges. Responses to drug reduction vary greatly from person to person due to differences in genetics, metabolism, baseline body composition, and underlying health status. Because of this large variation in responsiveness, some individuals may not adhere to their medication. [Overview of the project]

[0006] To address these and other technical issues related to adherence to obesity medications, digital therapeutic applications can utilize machine learning models, such as those detailed herein, which leverage real-time, multimodal data from individuals to output expected outcomes for various weight-loss medications, including side effects, weight loss, and dropout rates. There are several advantages for users of the digital therapeutic applications detailed herein. Firstly, digital therapeutic applications can aggregate data from a wide variety of modalities from various sources in real time and apply it to machine learning models to personalize weight-loss medication selection. By analyzing data from users using machine learning models, digital therapeutic applications can identify which obesity medication is most likely to yield the highest weight-loss outcome with the fewest side effects for that particular user. Outputs may also include the exact type, dosage, frequency, and administration regimen of the weight-loss medication. Compared to trial-and-error approaches, machine learning models output the most effective treatment and optimal clinical outcomes for a particular user.

[0007] Another advantage is that digital therapeutic applications can use machine learning models to predict potential side effects based on user data. Predicting side effects is valuable for weight loss medications, which are known to have a wide range of potential side effects. By proactively identifying the risk of such side effects, digital therapeutic applications can adjust the administration of weight loss medications (e.g., type of drug, dosage, frequency) to mitigate them before they occur. This proactive approach minimizes user discomfort, improves medication adherence, and ultimately contributes to better outcomes. Furthermore, machine learning models can be leveraged to identify users at particular risk of discontinuing weight loss medications. Using these models, digital therapeutic applications can identify behavioral and physiological factors that are typical causes of discontinuation for such users and offer different medications or administration parameters to avoid discontinuation. By providing personalized support, digital therapeutic applications can offer specific interventions to mitigate both the user's risk of side effects and dropout, thereby increasing the likelihood of medication adherence and achieving better clinical outcomes.

[0008] Furthermore, instances of the machine learning model are created and adapted for specific users to take into account each user's unique characteristics. For example, multiple instances of the machine learning model can be created for different individuals supervised by a physician, with each instance tailored to each individual's specific response profile and preferences. For an individual who has previously experienced a predisposition to side effects from a particular drug, an instance of the machine learning model for this individual would be created to output a higher probability of side effects from that drug. Another instance of the machine learning model would be trained to account for factors known to lead to poor medication adherence (e.g., high dosing frequency) for another individual who has no concerns about the same drug therapy but has difficulty adhering to medication due to work schedules. Because each instance of the machine learning model is tailored to these individuals, it has the ability to produce different outputs even with the same or similar measurement data. This increases the accuracy and relevance of the machine learning model's output across diverse user profiles and characteristics.

[0009] Furthermore, the digital therapeutic application incorporates real-time data from the user regarding their physiological measurements (e.g., blood glucose, heart rate, or blood pressure) and iteratively updates the risk of side effects and weight loss predictions as the user adheres to their medication prescription. By continuously monitoring the user's measurements in real time as the user progresses through the weight loss process, the digital therapeutic application can iteratively and continuously update the user's medication, dosage, administration parameters, and the user's risk of dropping out and side effects. Unlike other approaches that do not incorporate real-world data, the digital therapeutic application uses machine learning models to evaluate real-time data and consider all relevant factors that may influence the user's response to a specific weight loss medication. The use of the digital therapeutic application described herein makes it possible to create highly personalized prescriptions while optimizing the integration of pharmacological and digital therapeutic interventions.

[0010] From a computer perspective, machine learning models can process and analyze large datasets across various modalities with high efficiency. Using machine learning models, digital therapeutic applications can extract complex, high-dimensional data into embedded representations, reducing computation time without compromising predictive accuracy. Compared to trial-and-error approaches involving repeated trials and manual monitoring and tracking, digital therapeutic applications can continuously monitor and provide recommendations for drug selection, optimized dosages, administration frequency, and timing. By reducing reliance on manual evaluation and sequential trials, digital therapeutic applications can deliver data-driven outputs more accurately and quickly.

[0011] Furthermore, digital therapeutic applications can provide numerous visualizations to promote medication adherence. They offer visual motivation to encourage adherence to drug prescriptions. Digital therapeutic applications can generate simulations representing the user's experience when taking medication. For example, based on physiological measurements, simulations can identify causes of dropout, the probability of side effects, or predicted weight loss, and this information can be represented graphically over a period of time. These visualizations serve as compelling tools to motivate users to take their medication and prepare for predicted side effects. The likelihood of side effects and predicted weight loss are continuously updated to reflect real-time data received by the digital therapeutic application, ensuring that the information provided to users remains accurate and effective.

[0012] Therefore, digital therapeutic applications address the lack of accurate, personalized integration of real-world data by providing models that dynamically predict individual responses to weight-loss medications. This application offers a precise, individually tailored approach to weight loss aligned with each patient's physiological profile to promote overall weight loss while minimizing side effects. Such integration of digital and pharmacological solutions significantly improves obesity treatment, leading to superior outcomes and an overall improvement in patient care.

[0013] Aspects of this disclosure relate to systems and methods for selecting weight-loss medications for a user's obesity condition. One or more processors can receive one or more physiological measurements of the user. The one or more processors can apply the one or more physiological measurements to a machine learning model. The machine learning model is trained using a plurality of examples, each example may include one or more sample physiological measurements of a sample user and a corresponding sample weight-loss medication administered to the sample user. Based on the application of the one or more physiological measurements to the machine learning model, the one or more processors can generate metrics indicating expected outcomes associated with the user's weight-loss medication. The one or more processors can provide a user device associated with the user with a selection of a weight-loss medication based on the metrics indicating the expected outcomes, or a message indicating at least one of the expected outcomes associated with the weight-loss medication.

[0014] In various implementations, the one or more processors select a weight-loss drug from a plurality of weight-loss drugs based on a plurality of expected outcomes associated with the plurality of weight-loss drugs. In various implementations, the one or more physiological measurements include at least one of the following: body mass index, body weight, blood pressure, heart rate, smoking status, glucose excretion, total metabolic panel, complete blood count, lipase levels, thyroid panel, magnesium levels, HgA1c, fasting glucose, energy expenditure, physical activity, hormone levels, body weight, body fat percentage, genetic markers, assessment of gut microbiota, or energy intake. The machine learning model includes one or more corresponding weights for generating one or more values, the one or more corresponding weights including at least one of binary weights or continuous weights. The one or more processors can then generate the metric based on the one or more values. The weight-loss drugs can be selected from GLP-1 receptor agonists or GIP receptor agonists. The GLP-1 receptor agonist is selected from one or more of semaglutide, liraglutide, exenatide, and dulaglutide, and the GIP receptor agonist includes tylzepatide.

[0015] In various realizations, the expected outcome further includes at least one administration parameter of the dose-reducing drug, the administration parameter further includes at least one of the dose of the dose-reducing drug, the timing of administration of the dose-reducing drug, the frequency of administration, the route of administration, the dose-increase protocol, the circumstances of administration, or any combination thereof. In various realizations, the expected outcome further includes identifying at least one of the following: discontinuation of treatment, adverse events, or therapeutic efficacy. In various realizations, the adverse events are selected from nausea, vomiting, diarrhea, early satiety, loss of appetite, anorexia, dizziness, increased heart rate, indigestion, headache, hypoglycemia, kidney or ureteral stones, pancreatitis, diabetic retinopathy, depression, suicidal ideation or attempt, abdominal pain, acute kidney injury, muscle weakness and atrophy, constipation, or any combination thereof. To generate the metrics, one or more processors may generate multiple metrics for multiple expected outcome parameters based on the application of one or more physiological measurements to the machine learning model.

[0016] In various realizations, the expected outcome further includes at least one expected outcome parameter, the expected outcome parameter further includes at least one of the timing of discontinuation, the cause of discontinuation, the probability of discontinuation, the mitigation of discontinuation, or any combination thereof. In various realizations, the expected outcome further includes at least one expected outcome parameter, the expected outcome parameter includes at least one of the timing of the onset of the adverse event, the duration of the adverse event, the probability of the adverse event, the mitigation of the adverse event, or any combination thereof. In various realizations, the expected outcome further includes at least one expected outcome parameter, the expected outcome parameter includes at least one of the timing of the onset of the adverse event, the duration of the adverse event, the probability of the adverse event, the mitigation of the adverse event, or any combination thereof. In various implementations, the expected outcome further includes at least one expected outcome parameter, the expected outcome parameter including at least one of weight loss, fat loss, fasting blood glucose, cholesterol levels, hormone levels, duration of fat loss, risk of weight regrowth, change in body mass index, or any combination thereof.

[0017] In various implementations, the one or more processors generate a simulation that identifies a plurality of expected outcomes across a corresponding plurality of points in time. The simulation may include a representation of the plurality of expected outcomes across the corresponding plurality of points in time, the representation including at least one of a timeline, graph, video, audio, or avatar. In various implementations, the simulation identifies that the plurality of expected outcomes include at least one expected outcome parameter, the expected outcome parameter further including at least one of the following: timing of termination, cause of termination, probability of termination, mitigation of termination (also known as reduction, decrease, minimization, or prevention), or any combination thereof. The simulation identifies that the plurality of expected outcomes include at least one expected outcome parameter, the expected outcome parameter including at least one of the following: timing of adverse event occurrence, duration of adverse event, probability of adverse event, mitigation of adverse event, or any combination thereof. The simulation identifies that the plurality of expected outcomes include expected outcome parameters, the expected outcome parameters include at least one of weight loss, fat loss, fasting blood glucose, cholesterol levels, hormone levels, duration of fat loss, risk of weight regain, change in body mass index, or any combination thereof.

[0018] In various implementations, the user has an obesity index greater than 25, a body fat percentage greater than 20%, type 1 diabetes, type 2 diabetes, or non-alcoholic steatohepatitis (NASH). The one or more processors can deliver the message to the user device or the user's clinician. The one or more processors can acquire the one or more physiological measurements of the user by at least one of the user device or an instrumentation device attached to the user. To generate the metric indicating the expected outcome, the one or more processors can identify the metric based on at least one of (i) the mean of the plurality of examples, (ii) a weighted combination of the plurality of examples, or (iii) a comparison with a dataset consisting of the plurality of examples. In various implementations, to receive the one or more physiological measurements, the one or more processors can receive the one or more physiological measurements from at least one of the user device or an instrumentation device over a period of time. To apply the machine learning model, the one or more processors can apply the one or more physiological measurements to the machine learning model in response to the passage of the period of time. To generate the metric, one or more processors may generate the metric that shows the expected outcome associated with the weight-loss drug over a subsequent period.

[0019] In various implementations, one or more processors receive one or more subsequent physiological measurements of the user. In response to the receipt of the one or more subsequent physiological measurements, the one or more processors can apply the one or more subsequent physiological measurements to the machine learning model. Based on the application of the one or more subsequent physiological measurements to the machine learning model, the one or more processors can generate subsequent metrics indicating the expected subsequent outcome associated with the user's subsequent weight-loss medication. Within a defined period of time in response to the receipt of the one or more subsequent physiological measurements, the one or more processors can provide a subsequent message to the user device, the subsequent message indicating at least one of the following: (a) a selection of a subsequent weight-loss medication based on the subsequent metric indicating the weight-loss medication or the expected subsequent outcome, or (b) the expected subsequent outcome associated with the subsequent weight-loss medication. The defined period can range from one second to one hour. [Brief explanation of the drawing]

[0020] The purposes, other purposes, aspects, features, and benefits of this disclosure described above will be better understood by referring to the following description along with the accompanying drawings. [Figure 1] Figure 1 shows a block diagram of a system for selecting a drug for a user according to an exemplary embodiment. [Figure 2] Figure 2 shows a block diagram of a process for receiving measurements from a user according to an exemplary embodiment. [Figure 3] Figure 3 shows a block diagram of the process of generating metrics using a machine learning (ML) model according to an exemplary embodiment. [Figure 4] Figure 4 shows a block diagram of the process for presenting a simulation and selected drug according to an exemplary embodiment. [Figure 4] Figures 5A and 5B show screenshots of several exemplary user interfaces for selecting a drug, according to one exemplary embodiment. [Figure 6] Figures 6A and 6B show screenshots of exemplary sets of user interfaces for selecting drugs according to an exemplary embodiment. [Figure 7] Figure 7 shows a screenshot of an exemplary user interface for selecting a drug according to an exemplary embodiment. [Figure 8] Figure 8 shows a screenshot of an exemplary user interface for notifying side effects of a drug according to an exemplary embodiment. [Figure 9] Figure 9 shows a screenshot of an exemplary user interface for notifying side effects of a drug according to an exemplary embodiment. [Figure 10] Figure 10 shows a screenshot of an exemplary user interface for showing expected outcomes according to an exemplary embodiment. [Figure 11] Figure 11 shows a screenshot of an exemplary set of user interfaces for selecting a drug according to an exemplary embodiment. [Figure 12] Figure 12 shows a flowchart of a method for selecting a drug for a user to address obesity according to an exemplary embodiment. [Figure 13] Figure 13 is a block diagram of a server system and a client computer system according to an exemplary embodiment. **DETAILED DESCRIPTION OF THE INVENTION**

[0021] In reading the following descriptions of the various embodiments, the descriptions and their respective contents listed following each section of the specification should be helpful.

[0022] Section A describes a system and method for selecting a drug to address a user's obesity condition.

[0023] Section B describes network and computing environments that may be useful for carrying out the embodiments described herein.

[0024] A. Describe a system and method for selecting medications to address the user's obesity condition. This specification presents a system and method for selecting weight-loss medications for users with obesity. The digital therapeutic application described herein can receive physiological measurements from the user. These physiological measurements may include values ​​entered by the user, test results, or data from wearable technology. Upon receiving the physiological measurements, the application can apply the measurements to a machine learning model to select a weight-loss medication for the user based on expected outcomes. Expected outcomes include the effects, such as results, effects, and timing of side effects, if the user is administered a particular weight-loss medication.

[0025] Machine learning models can be trained with training data derived from real-world data, including the effects of specific weight-loss drugs and prescriptions on individuals with various physiological measurements from a sample of subjects. As a result of this training, the machine learning model can assign weights to each physiological measurement. These weights can be embedded with patterns (or latent features) revealed in the training data. Furthermore, applications can generate simulations based on physiological measurements to visualize the effects of weight-loss drugs on users, such as side effects and weight loss. As applications receive more data, for example from wearable technology, predicted weight loss and other such factors can be updated to reflect real-time data.

[0026] By using machine learning models, the application can generate metrics that show the expected outcomes associated with a weight-loss drug selected for the user. To select a weight-loss drug, the application can generate a set of expected outcomes and select the drug that has the fewest side effects and delivers the most desirable healthy outcome. Expected outcomes may include, for example, the timing of side effect onset, fat loss, duration of fat loss, or cholesterol levels, blood glucose, insulin levels, etc. The application can consider physiological measurements and generate metrics based on the predicted results of the selected weight-loss drug and its impact on the user. The application can then provide the user with a message indicating the selection of a weight-loss drug based on the metrics showing the expected outcomes, or at least one of the expected outcomes associated with that weight-loss drug. The message may also include visualizations of the expected outcomes, such as graphs, avatars, timelines, videos, or audio.

[0027] In this way, the application can leverage real-world data to identify the optimal weight-loss medication based on the user's physiological measurements and facilitate their weight loss. By considering a wide variety of physiological measurements, the application can select a weight-loss medication and generate expected outcomes tailored to the user as a result of taking the medication. As the application receives subsequent physiological measurements, it can update the weight-loss medication and associated expected outcomes. For example, as the user continues to lose weight, the application may adjust the dosage and frequency of the weight-loss medication, as well as the expected outcomes. The application can integrate pharmacotherapy and digital therapy solutions to facilitate the user's weight loss while avoiding trial-and-error prescriptions.

[0028] Referring here to Figure 1, a block diagram of system 100 for presenting interactive sessions to address user obesity is shown. In summary, system 100 may include at least one session management service 105 connected to communicate with each other via at least one network 115, a set of user devices 110A-N (hereinafter generally referred to as user devices 110), and an instrumentation device 135. At least one user device 110 (e.g., the first user device 110A shown) may include at least one application 125. Application 125 may include or provide at least one user interface 130. Session management service 105 may include at least one data collector 140, a model applyer 145, a metric evaluator 150, a simulation handler 155, an output generator 160, and at least one machine learning (ML) model 165. Session management service 105 may include or be able to access at least one database 170. The database 170 may store, maintain, or otherwise include at least one user profile 175A-N (hereinafter generally referred to as user profile 175) and a training dataset 180. The functionality of application 125 may be partially executed on the session management service 105. Conversely, the functionality of application 125 may incorporate operations performed on the session management service 105. The user device 110 and the session management service 105 may collectively be part of a computing system that provides application 125.

[0029] More specifically, the session management service 105 (sometimes referred to as the service in general terms) is a computing device comprising one or more processors coupled with memory and software, and may be any computing device capable of performing the various processes and tasks described herein. The session management service 105 can communicate with one or more user devices 110 and a database 170 via a network 115. The session management service 105 may be located in, deployed to, or otherwise associated with at least one server group. This server group may correspond to a data center, branch office, or site where one or more servers corresponding to the session management service 105 are located. The session management service 105 may be located in, deployed to, or otherwise associated with one or more user devices 110. Some components of the session management service 105 may be located within a server group, and some may be located within client devices. For example, the session management service 105 may operate on or be deployed on a user device 110, and the ML model 165 may operate on or be deployed on a server group.

[0030] Within the session management service 105, the data collector 140 can present data for input to the user on each user device 110 via the application 125. The data collector 140 can then collect the data provided by the user, and the model applyer 145 can receive this data. The model applyer 145 can analyze the data by applying the user data to a machine learning model (e.g., ML model 165) to determine the metrics for weight loss medication. Based on the data, the simulation handler 155 can simulate various outcomes of weight loss medication for the user and provide the user with visual output. The output generator 160 can select a drug based on the metrics and provide the user with a message along with the drug and visual output.

[0031] Using ML model 165 (sometimes referred to herein as a machine learning (ML) model), a set of expected outcomes can be generated for a set of weight-loss drugs, and one of the weight-loss drugs can be selected based on this set of expected outcomes. The architecture of the machine learning model may include, for example, deep learning neural networks (e.g., convolutional neural network architecture, residual network, or transformer-based architecture), regression models (e.g., linear regression models or logistic regression models), random forests, gradient boosting, K-nearest neighbors classifiers and / or regressors, support vector machines (SVMs), clustering algorithms (e.g., k-nearest neighbors), or naive Bayes models, and may be supervised, unsupervised, or self-supervised.

[0032] Generally, the ML model 165 can have at least one input and output. The input and output may be related through a set of weights. The input may be data from a user, while the output may include at least one drug option, expected outcome, and visual output. Visual output may include a timeline, graph, video, audio, or avatar. The ML model 165 can be trained using a training dataset 180. The training dataset 180 may include a set of examples representing subjects treated with weight-loss therapy. Each example may include an input (e.g., weight-loss drug, physiological measurements, selected drug) and an outcome (e.g., weight loss, side effects, discontinuation). In some embodiments, the session management service 105 may maintain a set of ML models 165, each model relating to a specific user or the clinician treating that user.

[0033] The user device 110 (sometimes referred to herein as an end-user computing device or client device) may be any computing device that includes one or more processors coupled with memory and software and is capable of performing the various processes and tasks described herein. In some embodiments, the user device 110 may be associated with a user taking weight-loss medication. In some embodiments, the user device 110 may be associated with any individual clinician (e.g., for administering weight-loss medication to that individual). The user device 110 can communicate with the session management service 105, the instrumentation device 135, and the database 170 via the network 115. The user device 110 may be a smartphone, another mobile phone, a tablet computer, a wearable device (e.g., a smartwatch, glasses), or a laptop computer. The application 125 can be accessed using the user device 110. In some embodiments, the application 125 can be downloaded (e.g., via a digital distribution platform) and installed on the user device 110. In some embodiments, the application 125 may be a web application with resources accessible via the network 115.

[0034] The instrumentation device 135 (which may also be referred to herein as wearable technology, wearable device, or device) can be any computing device capable of collecting measurements from a user. The measurement device 135 may include a fitness tracker, smartwatch, heart monitor, pedometer, or glucose monitor. The instrumentation device 135 can be worn by the user (e.g., attached to the user) to collect various data. The measurement data can be continuously provided to the user device 110 (e.g., application 125 on the user device 110) or the data collector 140. The instrumentation device 135 can communicate with the session management service 105, the user device 110, and the database 170 via the network 115.

[0035] An application 125 running on a user device 110 can be a digital therapeutic application and can provide sessions (sometimes referred to herein as therapeutic sessions) to address obesity and diseases associated with obesity. Users of application 125 can be individuals diagnosed with or at risk of having diseases or conditions associated with obesity. For example, a user may have diabetes, hypertension, hypercholesterolemia, fatigue, excess body fat, psychological problems, snoring, shortness of breath, or physical disability. The diseases or conditions can include any number of conditions that contribute to the user's obesity. A user may have a Body Mass Index (BMI) of 25 or higher, 30 or higher, 32 or higher, or 35 or higher. A user may have a body fat percentage of 20% or higher, 25% or higher, or 30% or higher. A user may have type 1 diabetes, type 2 diabetes, non-alcoholic steatohepatitis (NASH), or non-alcoholic fatty liver disease (NAFLD). Causes of obesity include genetic, behavioral, environmental, physiological, and psychological factors. For example, certain genetic traits can influence weight gain, such as slowing metabolism and / or regulating appetite. A family history of obesity can also increase the likelihood of developing it, based on shared lifestyle and genetics. In another example, metabolic conditions (e.g., diabetes) can lead to weight gain due to increased appetite and calorie intake resulting from elevated blood glucose levels.

[0036] Obesity has physical, mental, social, and economic consequences, affecting the children of obese individuals. Physical health consequences include, for example, cardiovascular diseases (e.g., hypertension, heart disease, or stroke) or metabolic disorders (e.g., diabetes or dyslipidemia). These conditions can impair or hinder social skills, causing individuals to face discrimination and feel ashamed in various social settings, such as the workplace and healthcare facilities.

[0037] Application 125 allows users to select a medication and receive the expected outcomes associated with that medication. This medication can be presented as a result of the user inputting data (e.g., physiological measurements), selecting a medication, or generating metrics (also referred to herein as scales, indicators, parameters, or evaluation or judgment criteria, etc.) that indicate the expected outcomes. This medication may be presented with visual outputs of the expected outcomes, such as the timing of side effect onset or a timeline showing weight loss over time. By providing users with digital treatments (e.g., interventions) through Application 125, the adverse effects of obesity can be addressed.

[0038] Physiological measurements may include at least one of the following: user's body mass index, weight, blood pressure, heart rate, smoking status, glucose excretion, total metabolic panel, complete blood count, lipase levels, thyroid panel, magnesium levels, HgA1c, fasting glucose, energy expenditure, physical activity, hormone levels, weight, body fat percentage, genetic markers, assessment of gut microbiota, or energy intake. Expected outcomes may include at least one of the following: treatment discontinuation (e.g., dropout, discontinuation of medication), side effects, or therapeutic efficacy (e.g., drug efficacy, efficacy, or success rate, etc., as specified herein).

[0039] A user may be receiving treatment to address a medical condition or side effects of a medical condition at least partially concurrently with the drug prescription presented by Application 125. For example, a user may be receiving treatment for diabetes. A user may be receiving treatment at least partially concurrently with a first-line drug option, a second-line drug option, a third-line drug option, or any combination thereof. This treatment may include taking medication. Such medications may be administered at least orally, intravenously, or topically. For example, in the case of a metabolic disorder (e.g., diabetes or hypothyroidism), a user may be taking diabetes medication (e.g., insulin, biguanides, sulfonylurea, or meglitinide) or a beta-blocker (e.g., propranolol or atenolol), etc. Treatment may include blood tests, bariatric surgery, metabolic surgery, nutritional counseling, psychological counseling, or weight management programs.

[0040] To address obesity, drugs such as glucagon-like peptide-1 receptor agonists (GLP-1 RAs) and / or gastrointestinal suppressant peptides (GIPs) can be administered to users to promote weight loss. GLP-1 RAs mimic the action of GLP-1, a naturally occurring hormone that regulates blood glucose levels, insulin secretion, and appetite. GIP therapeutics are analogs of the human GIP hormone, which stimulates insulin secretion from the pancreas. Therefore, GLP-1 RAs and GIPs can increase insulin secretion, decrease glucagon secretion, slow gastric emptying to suppress appetite, lower blood glucose levels, and promote weight loss. Examples of GLP-1 RAs include semaglutide, liraglutide, exenatide, and dulaglutide. An example of a GIP is tylzepatide. In some embodiments, the drug may include any analogue of a GLP-1 RA or GIP. The drug may include functionally equivalent peptides or therapeutics. Furthermore, the drugs may include any drugs that demonstrate therapeutic efficacy in treating obesity and obesity-related conditions. For example, the drugs may include drugs that enable GLP-1 or GIP agonism (receptor activation), or drugs that can induce fat loss in the user and / or suppress appetite.

[0041] Database 170 can store and maintain various resources and data associated with the session management service 105 and the application 125. Database 170 may include a database management system (DBMS) for organizing and structuring the data maintained on the database. Database 170 can communicate with the session management service 105 and one or more user devices 110 via the network 115. While performing various operations, the session management service 105 and the application 125 can access database 170 and retrieve identified data from it. The session management service 105 and the application 125 can also write data to database 170 from the execution of such operations.

[0042] Such actions may include maintaining a user profile 175 (sometimes referred to herein as a subject profile). The user profile 175 may include information related to the user's medical condition, as described herein. For example, the user profile 175 may include information related to the severity of the condition, the onset of the condition (such as the onset of symptoms associated with the condition that affect the user's cognitive function), medications or treatments the user is taking for the condition, and / or the duration of the condition. The user profile 175 may be updated in particular in response to a schedule (e.g., daily, weekly), in response to changes in user information (e.g., entered by the user via the user interface 130 or learned from the user device 110), or in response to a clinician (e.g., a doctor or nurse) addressing the user's medical condition.

[0043] The user profile 175 can store and maintain information associated with a user of application 125 via the user device 110. Each user profile 175 can be associated with or correspond to each user of application 125. This directional approach can reduce the need for numerous communications with the user, thereby reducing bandwidth and increasing the benefits of user-computer interaction. In some embodiments, the user profile 175 may identify or include information about the treatment regimen the user is practicing, such as the type of treatment (e.g., therapy, pharmacotherapy, or psychotherapy), duration (e.g., days, weeks, or years), and frequency (e.g., daily, weekly, quarterly, or yearly). The user profile 175 can be stored and maintained in a database 170 using one or more files (e.g., Extended Markup Language (XML), comma-separated format (CSV) delimited text file, or Structural Query Language (SQL) file). The user profile 175 may be iteratively updated as the user provides responses, makes choices, and performs actions in relation to a session, data collector 140, or output generator 160. For example, the user profile 175 can be updated using data collected by the data collector 140.

[0044] Referring here to Figure 2, a block diagram of process 200 in system 100 for collecting measurements from user 210 in order to select a drug is shown. Process 200 includes, or can include, actions performed by system 100 to receive data from the user and process it. In process 200, a data collector 140 running on session management service 105 can communicate with user devices 110 or instrumentation devices 135 associated with at least one user 210. User 210 may have a body mass index (BMI) greater than 25, a body fat percentage greater than 20%, type 1 diabetes, type 2 diabetes, or NASH, etc. User 210 may have an obesity condition associated with diabetes, hypertension, or high cholesterol, etc.

[0045] The data acquisition device 140 can create, write, or otherwise generate one or more instructions 205 (hereinafter generally referred to as instructions 205) for at least one user 210. Instructions 205 may request measurements from user 210 (or application 125 or instrumentation device 135). Instructions 205 can identify one or more measurement fields to be acquired via user device 110. The one or more measurement fields may be related to physiological measurements of user 210. Measurements may span multiple modalities, such as numerical data, time-series data, images, Boolean values, or free text. One or more measurement fields may include at least one of the following: Body Mass Index (BMI) (e.g., calculated from the user's height and weight measurements), body weight (e.g., the total mass of user 210), blood pressure (e.g., the force exerted by circulating blood on the arterial walls), heart rate (e.g., the number of times the heart beats in 60 seconds), smoking status (e.g., an indication of whether user 210 smokes), glucose excretion, total metabolic rate panel, complete blood cell count, lipase levels, thyroid panel, magnesium levels, HgA1c, fasting glucose, energy expenditure, physical activity, hormone levels, body weight, body fat percentage, genetic markers, assessment of gut microbiota, or energy intake.

[0046] In some embodiments, one or more measurement fields may include at least one questionnaire. The questionnaire may be for user 210 to indicate at least one of the following: trauma (e.g., a physical trauma affecting the user's motor skills; the user recently experiencing dizziness which has affected their ability to tolerate nausea), or side effects (e.g., a negative response to a particular weight-loss drug or other non-weight-loss drug), or preferences regarding side effects (e.g., a preference to limit nausea as a side effect because the user is a pilot and cannot afford to feel nauseous while at work). For example, the questionnaire may ask the user about (i) common side effects since the start of drug treatment, (ii) specific symptoms (e.g., nausea, vomiting, diarrhea, early satiety, loss of appetite, anorexia, dizziness, increased heart rate, indigestion, headache, hypoglycemia, kidney or ureteral stones, pancreatitis, diabetic retinopathy, depression, suicidal thoughts or attempts, abdominal pain, acute kidney injury, muscle weakness, or muscle atrophy), (iii) the date the side effects began, (iv) the frequency of the side effects, or (v) their impact on daily life. As another example, the measurement field may include questions for the user 210 to answer regarding side effects that the user 210 can tolerate. The measurement field may also ask the user 210 to enter information about recent injuries, such as knee or ankle injuries that affect the user 210's ability to move. The questionnaire may be generated based on questions identified in the user profile 175.

[0047] In generating instruction 205, the data collector 140 may select or identify one or more physiological measurement fields based on at least the user profile 175. For example, the physiological measurements identified in instruction 205 may depend on the user's medical condition. The user profile 175 may include the user's past and current medical conditions (e.g., type 1 diabetes) available for drug selection, and the data collector 140 may identify blood glucose levels in addition to other measurements in instruction 205. In some embodiments, instruction 205 may identify one or more measurement fields associated with a period of time. For example, instruction 205 may define a period during which user 210 inputs one or more measurement fields. This period may range from one day to six months.

[0048] Once the data collector 140 generates an instruction 205, it can send, provide, or otherwise transmit it to a user device 110 or an instrumentation device 135. The transmission of the instruction 205 may follow a schedule (e.g., a period from 5 minutes to 2 weeks). In some embodiments, the user data collector 140 may send the instruction 205 to the user device 110 for at least a portion of the measurements. The data collector 140 may send the instruction 205 to the instrumentation device 135 for another portion of the measurements. The instruction 205 can take various forms associated with the application 125. In some embodiments, the instruction 205 can be displayed, rendered, or otherwise presented to the user 210 via the user interface 130 of the application 125. The instruction 205 can take the form of a list, table, database, graph, or chart displaying one or more measurement fields for the user 210 to input. In some embodiments, the instruction 205 may include a short message service (SMS) (e.g., text message) or a multimedia message service (MMS) (e.g., audio message, video message). For example, instruction 205 may include a link to open application 125, which instructs the user to enter one or more measurement fields via the user interface 130 of application 125.

[0049] An application 125 on the user device 110 can retrieve, identify, or otherwise receive instruction 205 from the session management service 105. Upon receiving instruction 205, application 125 can parse it to identify the measurement fields to be provided to the session management service 105. In some embodiments, application 125 can use instruction 205 to display, render, or otherwise present the user interface 130. Application 125 can then prompt or instruct user 210 to provide measurement values ​​215A-N (for example, one or more measurement values ​​referred to herein as measurement value 215) for each of one or more measurement fields. Measurement values ​​215 may be text strings, strings, or numbers entered by user 210 on the user interface 130. In some embodiments, upon receiving instruction 205, application 125 can retrieve, fetch, or otherwise identify the measurement values ​​215 as specified by instruction 205. The application 125 on the user device 110 may have previously generated and stored data associated with the measurement 215 from the user 210 (or the instrumentation device 135 communicating with the user device 110). In response to the instruction 205, the application 125 can fetch the measurement 215 stored in the data. Based on the above identification, the application 125 can return, provide, or otherwise transmit the measurement 215 to the session management service 105.

[0050] In some embodiments, the instrumentation device 135 can retrieve, identify, or otherwise receive an instruction 205 from the session management service 105. Upon receiving the instruction 135, the application 125 can parse the instruction 205 to identify the measurement fields to be provided to the session management service 105. The instrumentation device 135 can retrieve, fetch, or otherwise identify the measurement 215 as specified by the instruction 205. The instrumentation device 135 may have previously generated and stored data associated with the measurement 215 from the user 210. With the above identification, the instrumentation device 135 can return, provide, or otherwise transmit the measurement 215 to the session management service 105. In some embodiments, the instrumentation device 135 can acquire measurement 215 of one or more measurement fields (e.g., heart rate). The instrumentation device 135 may include a wearable continuous glucose monitor, a smartwatch capable of monitoring the user's heart rate and biometrics, a scale, or other technology that can acquire data associated with the user's physiological health. The instrumentation device 135 can provide the measured values ​​215 to the data collector 140 in real time (e.g., continuously) via the application 125.

[0051] The data collector 140 can then retrieve, identify, or otherwise receive the measured values ​​215 from the user 210 (or application 125 or instrumentation device 135). Upon receiving the measured values ​​215, the data collector 140 can correlate or associate them with the respective measurement fields. For example, instruction 205 may include a weight measurement field, and the data collector 140 receives the user 210's weight value. The data collector 140 can store the measured values ​​215 in the database 170 and associate them with the user profile 175. By storing them, the data collector 140 can add or update the user profile 175 to include the measured values ​​215. The user 210's measured values ​​215 can be obtained by at least one of the user device 110 or the instrumentation device 135 attached to the user 210. The data collector 140 can receive measured values ​​215 from at least one of the user device 110 or the instrumentation device 135 over a period of time. In various embodiments, the data collector 140 receives one or more subsequent physiological measurements from the user 210. For example, the data collector 140 may receive a measurement 215 after providing the measurement to the model applyer 145. In various embodiments, the measurement 215 is continuously updated based on the instrumentation device 135 provided to the user 210.

[0052] Referring here to Figure 3, a block diagram of process 300 in system 100 is shown, which applies measurements from user 210 to ML model 165 in order to select a drug. Process 300 includes, or can correspond to, actions performed by system 100 to receive data provided by the user and process it. With respect to process 300, ML model 165 can be initialized, trained, and configured using training dataset 180 on database 170 (for example, by model applyer 145). Training dataset 180 may identify or include a set of examples. In training dataset 180, each example may include one or more sample physiological measurements of a sample user and a corresponding sample weight-loss drug administered to the sample user. Each example may, in particular, include a sample measurement 215'AN (hereinafter referred to as sample measurement 215') and a sample metric 305'AN (hereinafter referred to as sample metric 305').

[0053] In some embodiments, at least one example of the training dataset 180 may include a drug identifier (ID) 310'AN (hereinafter referred to as drug identifier 310') for each sample user. Sample users may be real-world patients. Sample users may be 18 years of age or older at the time they first begin taking the weight-loss drug. Sample users may have a BMI greater than 25 within 60 days of first beginning to take the weight-loss drug. Sample measurements 215' may also include information about each sample user, such as patient identification information, year of birth, or diagnosis (e.g., obesity, type 1 diabetes, type 2 diabetes, or NASH). Sample users included in the training dataset 180 may have at least one BMI or height and weight measurement before starting weight-loss drug treatment and at least one BMI or height and weight measurement after starting weight-loss drug treatment.

[0054] Sample measurements 215' may include corresponding physiological measurements 215, which may include: body mass index, weight, blood pressure, heart rate, smoking status, glucose excretion (measured as blood glucose or urinary glucose), overall metabolic panel (e.g., renal and hepatic function, blood protein levels, blood electrolyte concentrations, etc.), complete blood cell count (e.g., red blood cell count and white blood cell count), lipase levels, thyroid panel (measurements of thyroid-stimulating hormone, thyroxine, triiodothyronine, and thyroid antibodies), magnesium levels, HgA1c This may include at least one of the following: (average blood glucose level at a specified timer time), fasting glucose, energy expenditure (measured as calorie expenditure), physical activity, hormone levels (e.g., testosterone, estrogen, progesterone), body weight, body fat percentage, genetic markers (e.g., genetic markers for obesity, diabetes, genetic markers associated with metabolic conditions), assessment of gut microbiota (assessed by a fecal panel), or energy intake (calorie, fat, protein, and carbohydrate intake). Each sample measurement can be associated with the sample user over a given period.

[0055] Sample metric 305' may include at least one of the following associated with a sample user: treatment discontinuation, adverse event, or treatment efficacy. For example, sample metric 305' may include at least 20 adverse events (e.g., vomiting, diarrhea, headache), (e.g., complete blood count, total metabolic panel), total cholesterol, lipase, thyroid panel, magnesium, HgA1c, dropout time (e.g., discontinuation of weight-loss medication), and diagnostic codes from the International Classification of Diseases, 10th Revision (ICD-10) for available life measures (e.g., height, weight, smoking status). Sample metric 305' may be associated with a sample user. The sample metric 305' may include a set of administration parameters for the dose-reducing drug, such as the dose of the dose-reducing drug, the timing of administration (e.g., morning or night), the frequency of administration (e.g., daily, weekly, or monthly), the route of administration (intravenous, subcutaneous, or intramuscular), the dosage increase protocol (increase or decrease in dose over time), and the circumstances of administration (e.g., in combination with other drugs or treatments).

[0056] Sample metric 305' may include information about drug side effects, such as the timing of onset, duration of side effects, probability of side effects, or mitigation of side effects. Sample metric 305' may include information related to the likelihood of discontinuation of treatment, such as the timing of discontinuation (length of drug treatment period until discontinuation), cause of discontinuation, probability of discontinuation (likelihood of a user discontinuing treatment within a given period), or discontinuation mitigation (e.g., psychotherapy, drug side effect mitigation measures). Sample metric 305' may include information about the effectiveness of treatment, such as weight loss, fat loss, fasting blood glucose, cholesterol levels, hormone levels, duration of fat loss, change in body mass index, or risk of weight regrowth.

[0057] Each of the sample user's sample measurement 215' and sample metric 305' can be associated with a drug identifier 310'. The drug identifier 310' can identify which weight-loss medication the sample user is taking and / or has taken. The drug identifier 310' may include at least one identifier from dulaglutide, exenatide, semaglutide, liraglutide, lixisenatide, or tilzepatide. In some embodiments, the drug identifier 310' may further indicate at least one of the following: the dose of the weight-loss medication, the timing of administration, the frequency of administration, the route of administration, the dosage increase protocol, the administration status, or any combination thereof.

[0058] For initialization, the model applyer 145 can set the values ​​of a set of weights of the ML model 165 to starting values ​​(e.g., random or defined values). For training, the model applyer 145 can input, supply, or otherwise apply sample measurements 215' and drug identifiers 310' and compare the output metric of the ML model 165 with the sample metric 305' for each sample user. Based on the comparison, the model applyer 145 may determine a loss metric according to a loss function (e.g., mean squared error, cross-entropy loss, hinge loss, or Huber loss). Using the loss metric, the model applyer 145 can update one or more weights of the ML model 165. The weight update can follow a backpropagation and optimization function (sometimes referred to herein as the objective function) having one or more parameters (e.g., learning rate, momentum, weight decay, and number of iterations). The optimization function can define one or more parameters for which the weights of the ML model 165 are updated. The optimization function can follow stochastic gradient descent and may include, for example, adaptive moment estimation (Adam), implicit update (ISGD), or adaptive gradient algorithm (AdaGrad). ML model 165 may be repeatedly updated until convergence occurs.

[0059] The establishment of the ML model 165 allows the model applyer 145 to supply, provide, or otherwise apply the measured values ​​215 to the ML model 165. During application, the model applyer 145 can process the measured values ​​215 according to a set of weights of the ML model 165. In some embodiments, a set of weights of the ML model 165 corresponds to a measured value 215. For example, for each of the measured values ​​215, the ML model 165 includes a set of corresponding weights. In some embodiments, this set of corresponding weights may include at least one of binary weights (e.g., 0 or 1) or continuous weights (e.g., numerical values ​​in the range of -100 to 100). For example, user 210's body fat percentage may be weighted higher than user 210's weight. Based on the set of corresponding weights, the ML model 165 generates one or more values. These one or more values ​​may be a function of the measured values ​​215, each having their corresponding weights.

[0060] In some embodiments, the ML model 165 may be specific to a particular user of the application 125 (e.g., user 210). The ML model 165 may have a bias on a set of weights according to the machine learning architecture so as to provide higher or lower values ​​depending on user 210's measurements. The bias may be performed using reweighting (e.g., manual setting of weight values), data augmentation, regularization, sampling, or other techniques. The weights of the ML model 165 can be assigned or set according to a user profile 175 of a particular user 210. For example, the user profile 175 may indicate that user 210 has experienced a particular type of side effect (e.g., dizziness or nausea) with a particular drug, while user 210 has had no problems with another drug. The weights of the ML model 165 can be updated or assigned so as to have a bias that produces an output indicating a higher value for user 175's expected outcome for one drug compared to another. The weights of the ML model 165 can be assigned or set by a clinician (e.g., a physician examining user 210) using a user interface provided by the session management service 105. For example, suppose a physician has two patients. One of them may have a prior experience of side effects that predisposes them to a particular drug. The other may not have concerns about the same drug, but may have difficulty adhering to other frequently administered medications due to a busy work schedule. One instance of the ML model 165 for the first individual can be set with a bias that outputs a higher probability for the factors related to that drug identified by that individual. Another instance of the ML model 165 for the second individual can have weights set higher for side effects known to lead to low adherence (e.g., higher administration frequency). As a result, the ML model 165 for these two individuals may produce different expected outcomes for these two individuals, even if the same or similar measurement data are used.

[0061] In some embodiments, the model applyer 145 may apply the measurement 215 in response to the passage of a period defined with respect to the instruction 205. The model applyer 145 may wait to apply the measurement 215 received over that period by the user device 110 or instrumentation device 135 defined in the instruction 205. Once that period has elapsed, the model applyer 145 may apply the measurement 215 to the ML model 165. In some embodiments, the model applyer 145 may apply one or more subsequent physiological measurements to the ML model 165 when one or more subsequent physiological measurements are received by the data collector 140.

[0062] Based on the application of measurement 215 to ML model 165, ML model 165 can calculate, determine, or otherwise generate at least one metric 305A-N (hereinafter generally referred to as metric 305). Metric 305 can identify or indicate the expected outcome associated with the user 210's weight-loss medication. Metric 305 may be for a period subsequent to the acquisition of measurement 215. The period of metric 305 may be in the future relative to the period of measurement 215 and can have any range between 1 day and 6 months. The expected outcome for metric 305 may identify or include at least one dosing parameter for the weight-loss medication. The dosing parameter may include at least one of the following: the dose of the weight-loss medication, the timing of administration of the weight-loss medication, the frequency of administration, the route of administration, the dose-increase protocol, the circumstances of administration, or any combination thereof.

[0063] In some embodiments, the ML model 165 can determine metric 305 based on at least one of the following: the mean of a set of examples, a weighted combination of a set of examples, or a comparison with a training dataset 180 containing a set of examples. For example, the ML model 165 can determine metric 305 based on the mean of sample measurements 215' for each of the drug identifiers 310'. As another example, the ML model 165 can determine metric 305 based on comparing information about user 210 along with the measurements 215' for each of the sample users. For example, the ML model 165 can identify a sample user with the highest similarity to user 210 (e.g., weight, height, side effect preferences, weight loss goals, drug dosage) and generate metric 305 based on the sample metric 305' of the sample user with the highest similarity. The ML model 165 can also determine metric 305 based on user 210's responses to a side effect questionnaire. Based on the questionnaire, ML model 165 may be adjusted to metric 305 in response to side effects that users 210 wish to avoid, such as nausea.

[0064] In some embodiments, the ML model 165 can calculate, determine, or otherwise generate a set of metrics 305 for a corresponding set of weight-loss drugs. The set of weight-loss drugs can be selected from GLP-1 receptor agonists or GIP receptor agonists. GLP-1 receptor agonists can be selected from one or more of semaglutide, liraglutide, exenatide, and dulaglutide, while GIP receptor agonists include tylzepatide. In some embodiments, the drugs can include any analogue of GLP-1 RA or GIP. The drugs may include functionally equivalent peptides or therapeutic agents. Furthermore, the drugs may include any drugs that demonstrate therapeutic efficacy for the treatment of obesity and obesity-related conditions. For example, the drugs may include drugs that enable GLP-1 or GIP agonism (receptor activation), or drugs that can induce fat loss in the user and / or suppress appetite. Each metric 305 can identify or indicate the expected outcome associated with the weight-loss drug for the user 210. The expected outcome may identify or include at least one administration parameter for the weight-loss drug.

[0065] In some embodiments, the expected outcome of metric 305 may include identifying at least one of the following: discontinuation of treatment, adverse reactions, or therapeutic efficacy. Adverse reactions may be selected from nausea, vomiting, diarrhea, early satiety, loss of appetite, anorexia, dizziness, increased heart rate, indigestion, headache, hypoglycemia, kidney or ureteral stones, pancreatitis, diabetic retinopathy, depression, suicidal ideation or attempt, abdominal pain, acute kidney injury, muscle weakness and atrophy, constipation, or any combination thereof. The expected outcome may also include identifying adverse reactions associated with at least one of the dose-reducing drugs and administration parameters for user 210.

[0066] In some embodiments, the model applyer 145 can generate a set of metrics 305 for a set of expected outcome parameters based on applying the measured values ​​215 to the ML model 165. The expected outcome may include at least one expected outcome parameter. The expected outcome parameter may include at least one of the following: timing of discontinuation, cause of discontinuation, probability of discontinuation, mitigation of discontinuation, or any combination thereof. For example, the expected outcome may include an expected outcome parameter associated with discontinuation of treatment. The expected outcome parameter may also include at least one of the following: timing of adverse event onset, duration of adverse event, probability of adverse event, mitigation of adverse event, or a combination thereof.

[0067] In some embodiments, the expected outcome of metric 305 may include parameters associated with side effects. The expected outcome parameters may also include at least one of the following: weight loss, fat loss, fasting blood glucose, cholesterol levels, hormone levels, duration of fat loss, risk of weight regain, change in body mass index, or any combination thereof. The expected outcome may include parameters associated with the therapeutic effectiveness of the weight-loss drug. For example, model applyer 145 may generate metric 305 for each weight-loss drug regarding the timing of the onset of side effects and the duration of side effects. In another example, model applyer 145 may generate metric 305 that identifies a predicted change in body mass index (e.g., increase or decrease) below a certain threshold (e.g., 25-30).

[0068] Using metric 305, the metric evaluator 150 can identify or select at least one drug identifier 310 from a set of drug identifiers 310. The set of drug identifiers 310 may include weight-loss drugs. In some embodiments, the metric evaluator 150 can select a weight-loss drug for user 210 from a set of weight-loss drugs. The drug identifier 310 may indicate the drug selected for user 210 based on metric 305. In some embodiments, the drug identifier 310 may include administration parameters such as route of administration and frequency. In some embodiments, the metric evaluator 150 can select at least one weight-loss drug based on a comparison of each weight-loss drug's metric 305 with at least one threshold. The threshold can be defined, specified, or otherwise identified by defining the value of metric 305 for selecting the corresponding weight-loss drug. If metric 305 meets the threshold (e.g., is greater than or equal to the threshold), the metric evaluator 150 can select the corresponding weight-loss drug for user 210. On the other hand, if metric 305 does not meet the threshold (e.g., below the threshold), the metric evaluator 150 can exclude the corresponding weight-loss drug with respect to user 210.

[0069] In some embodiments, the metric evaluator 150 can select a weight-loss drug for user 210 from a set of weight-loss drugs as a function of metric 305. For example, the metric evaluator 150 can select and generate drug identifiers 310 based on the weight-loss drug with the lowest probability of discontinuation, the lowest probability of side effects, and the highest amount of fat loss. In some embodiments, the set of metrics 305 includes a set of dosing parameters related to the weight-loss drug. In some embodiments, the metric evaluator 150 can select a weight-loss drug based on side effects expected to occur in user 210. In response to expected side effects corresponding to selections in a side effect questionnaire of side effects that user 210 wants to avoid, the metric evaluator 150 selects different weight-loss drugs. In some embodiments, the metric evaluator 150 selects one or more weight-loss drugs for user 210 and / or user 210's clinician to choose from. Thus, the metric evaluator 150 can generate multiple drug identifiers 310 along with multiple expected outcomes associated with each of these drug identifiers 310.

[0070] In some embodiments, the metric evaluator 150 can also use metric 305 and / or drug identifier 310 to generate interventions to mitigate and / or address side effects indicated by the expected outcome. For example, in response to metric 305 indicating that nausea occurs during administration of a selected drug, the metric evaluator 150 may provide interventions such as drinking more water or avoiding fatty foods. The metric evaluator 150 can also map interventions based on the timing, duration, and probability of the onset of side effects associated with the drug identifier 310.

[0071] Referring to Figure 4, a block diagram is shown of a process 400 in system 100 that uses metric 305 and drug identifier 310 to generate message 405. Process 400 includes, or can correspond to, actions performed by system 100 to receive data provided by the user and process it. In process 400, a simulation handler 155 running on session management service 105 can generate at least one simulation 410. Simulation 410 can identify a set of expected outcomes over a corresponding set of time points. Simulation 410 can identify expected outcome parameters over a period including this set of time points. For example, simulation 410 can illustrate the probability of side effects occurring over the duration of a treatment regimen (e.g., from one week to six months). To generate simulation 410, simulation handler 155 can identify a set of expected outcomes and expected outcome parameters over a period by applying measurement values ​​215 to ML model 165 over multiple periods. These periods can be progressively further away from the present, such as from one day to six months from the present. The simulation handler 155 may generate a simulation 410 for each of a set of expected outcomes.

[0072] Using the expected outcomes, the simulation handler 155 can generate a display of a set of expected outcomes. The simulation 410 may include this display of the set of expected outcomes. The display may include at least one of the following: a timeline (e.g., showing metrics of expected outcomes over time), a graph (e.g., expected outcomes by type), a video (e.g., showing changes in expected outcomes over time in numerical form), audio (e.g., narration of expected outcomes), or an avatar (e.g., animation). For example, the simulation 410 may include an avatar showing the weight loss of user 210 over a set of points in time. The size of the avatar may vary based on the expected outcomes identified in the simulation 410.

[0073] In conjunction with this, the output generator 160 can create, produce, or otherwise generate at least one message 405. The generation of message 405 is based on the metric 305 and the drug identifier 310. The generation and provision of message 405 may follow a schedule (e.g., at intervals of 5 minutes to 2 weeks). Message 405 may identify or indicate the expected outcome associated with the weight loss metric. In some embodiments, the output generator 160 can generate message 405 to include the metric 305 and the drug identifier 310, etc. The drug identifier 310 may be selected from a set of drug identifiers 310 (corresponding to weight loss drugs) using the metric 305.

[0074] In some embodiments, the output generator 160 can generate a message 405 that includes information derived or generated from a metric 305, a drug identifier 310, or a simulation 410, etc. The output generator 160 can generate a message 405 that includes information according to a template. The template may include a set of predefined content (e.g., text, images, videos, or audio) having one or more placeholders for including a metric 305, a drug identifier 310, or a simulation 410, etc. The message 405 may indicate the expected outcome associated with the weight loss drug. For example, the output generator 160 can extract information about the expected outcome to include in the message 405. In this case, the output generator 160 may include expected outcome parameters in the message 405, such as the duration of fat loss associated with the drug identifier 310. The output generator 160 can generate the message 405 by inserting the expected outcome parameters (e.g., duration of fat loss and drug treatment) into the template. Message 405 may include the expected outcome associated with the drug identifier 310. In some embodiments, message 405 may include interventions to mitigate and / or address side effects.

[0075] In some embodiments, the output generator 160 can generate a message 405 that includes a simulation 410 (e.g., a display of a set of expected outcomes). In some embodiments, the message 405 may include the simulation 410. Following the generation of the simulation 410, the simulation handler 155 may provide the simulation 410 to the output generator 160 for inclusion in the message 405. In this case, the message 405 includes both the drug identifier 310 (e.g., a selected weight-loss drug) and the simulation 410. Both the drug identifier 310 and the simulation 410 can be provided to the user device 110 and / or the clinician of user 210. The drug identifier 310 and the simulation 410 may be displayed together or sequentially on the user device 110. By generating the message 405, the output generator 160 can send, transmit, or otherwise provide the message 405 to the user device 110. In some embodiments, the output generator 160 may also provide the message 405 to a computing device associated with the clinician of user 210.

[0076] Upon receiving it, the application 125 on the user device 110 can display, render, or otherwise present information based on the message 405 via the user interface 130. In some embodiments, a computing device associated with a clinician can receive the message 405 and present information based on the message 405. This information may include or identify metrics 305 and drug identifiers 310. For example, the application 125 may display a set of metrics 305 related to the type of weight-loss drug, potential side effects, and likelihood of compliance. In some embodiments, the application 125 may display information identifying an intervention for the user 210. In some embodiments, the application 125 can render, display, or otherwise present a simulation 410 via the user interface 130. An example of a message 405 presented to the user device 110 is shown in Figure 5A-11.

[0077] Using the information presented via the user interface 130, user 210 (or the clinician examining user 210) can decide whether to take or be administered at least one drug 415 corresponding to drug identifier 310. The drug 415 may, for example, be provided by user 210's clinician. The drug 415 may include, for example, a GLP-1 receptor agonist or a GIP receptor agonist (for example, as detailed herein). User 210 and / or user 210's clinician may administer the drug 415 according to the administration parameters indicated in message 405. User 210 may consume or take the drug 415 corresponding to drug identifier 310.

[0078] The process described above can be repeated any number of times. In some embodiments, the data collector 140 can retrieve or receive subsequent measurements over a subsequent period. The model applyer 145 can generate another metric 305 for the subsequent period by applying the subsequent measurements to the ML model 165. The metric generated after the period may differ from the metric 305 generated before the period. In some embodiments, the model applyer 145 generates a subsequent metric (for example, after generating metric 305) that shows the subsequent expected outcome associated with the subsequent weight-loss medication for user 210. The model applyer 145 can generate the subsequent metric after the period in which user 210 is taking the weight-loss medication. In this case, the model applyer 145 may use the ML model 165 to update metric 305 based on the changes represented by the subsequent physiological measurements of user 210. Based on the subsequent physiological measurements, the subsequent expected outcome may differ from the expected outcome, such as a decrease in the frequency of administration of the subsequent weight-loss medication. Subsequent weight-loss medications may be different from the first weight-loss medication.

[0079] In some embodiments, the output generator 160 provides a subsequent message within a defined period in response to the reception of one or more subsequent physiological measurements. The output generator 160 can provide a subsequent message in response to the reception of a subsequent metric. Thus, the subsequent message may indicate (a) a selection of a subsequent weight-loss drug based on a subsequent metric indicating a subsequent expected outcome, or (b) at least one of the subsequent expected outcomes associated with the subsequent weight-loss drug. The defined period can range from one second to one hour. In this case, the metric evaluator 150 can also generate a subsequent drug identifier. Thus, given a subsequent physiological measurement, the ML model 165 can generate at least one subsequent metric, the metric evaluator 150 can generate a subsequent drug ID, and the output generator 160 can generate a subsequent message within a defined period based on the subsequent drug ID and the subsequent metric. In some embodiments, the simulation handler 155 can generate subsequent simulations included in subsequent messages.

[0080] In this way, the session management service 105 aggregates physiological measurements from the user 210 and uses the ML model 165 to determine the optimal weight-loss medication that aligns with the user's unique physiological profile to improve weight-loss outcomes. By leveraging diverse multimodal data in the form of physiological parameters, the session management service 105 can algorithmically select target weight-loss medications and dynamically generate personalized predicted outcomes. As more physiological measurements are received over time, the expected outcomes and, consequently, the recommended weight-loss medications can be dynamically adjusted by the session management service 105. For example, as the user's weight loss progresses, the session management service 105 can automatically adjust the drug dosage and frequency based on updated physiological data, thereby further improving the expected outcomes. This session management service 105 integrates pharmacotherapy and digital treatment to create a comprehensive, data-driven weight-loss solution, thereby reducing reliance on trial-and-error drug prescriptions and providing a personalized, adaptive approach to addressing the user 210's obesity condition.

[0081] Figures 5A and 5B show screenshots of a set of 500 exemplary user interfaces for selecting a drug, according to an exemplary embodiment. The set of 500 user interfaces may be part of application 125 and may be presented to a user 210 being administered a weight-loss drug via user interface 130. User interface 505 may prompt the user to enter values ​​(e.g., one or more measurements) corresponding to displayed measurement fields. The user can then enter numbers, text strings, or strings, etc., into the measurement fields. User interfaces 510 and 515 may present the user with recommendations regarding the selected weight-loss drug. Recommendations may include expected outcomes, such as the amount of weight loss and side effects associated with the weight-loss drug.

[0082] Figures 6A and 6B show screenshots of a set of 600 exemplary user interfaces for selecting a drug, according to an exemplary embodiment. The set of 600 user interfaces may be part of an application and presented via the user interface of a computing device associated with a clinician examining the user. User interface 605 may prompt the user to enter values ​​(e.g., one or more measurements) corresponding to displayed measurement fields. The user can then enter numbers, text strings, or strings, etc., into the measurement fields. User interfaces 610 and 615 may display multiple recommended weight-loss drugs to the user or the user's physician and prompt the user or the user's physician to select a weight-loss drug. User interfaces 610 and 615 may also display potential side effects and the possibility of discontinuation (e.g., withdrawal risk).

[0083] Figure 7 shows a screenshot of an exemplary set of 700 user interfaces for selecting a drug, according to an exemplary embodiment. The set of 700 user interfaces can be part of application 125 and can be presented via user interface 130. User interface 705 can be presented to a user clinician (e.g., a physician). User interface 705 allows the clinician to select recommendations (e.g., specific drugs and dosage parameters) to present to the user. For example, the selected recommendation may be the administration of semaglutide twice daily for one week to a subject. In some embodiments, user interface 705 can be displayed to the user, allowing the user to select a recommendation.

[0084] Figure 8 shows a screenshot of a set of 800 exemplary user interfaces for selecting a drug, according to an exemplary embodiment. The set of 800 user interfaces can be part of application 125 and can be presented via user interface 130. User interface 805 can display notifications to the user indicating the occurrence of side effects associated with the weight-loss drug. The notifications can be generated based on the user's wearable technology. The notifications can identify actions to prevent or reduce side effects.

[0085] Figure 9 shows a screenshot of a set of 900 exemplary user interfaces for selecting a drug, according to an exemplary embodiment. The set of 900 user interfaces can be part of application 125 and can be presented via user interface 130. User interface 905 can be presented to a user clinician and can provide recommendations to mitigate withdrawal, in addition to an estimated time for discontinuing the user's dose-reducing medication.

[0086] Figure 10 shows a screenshot of an exemplary user interface 1000 for selecting a drug according to an exemplary embodiment. A set of user interfaces 1000 can be part of application 125 and can be presented via user interface 130. User interface 1005 displays a simulation related to weight loss drug recommendations. For example, the simulation includes a graph that displays both weight loss and side effects over time. The graph can compare recommendations for multiple weight loss drugs.

[0087] Figure 11 shows a screenshot of a set of exemplary user interfaces 1100 for selecting a drug, according to one exemplary embodiment. The set of user interfaces 1100 may be part of application 125 and may be presented via user interface 130. User interface 1105 may prompt the user to enter values ​​(e.g., one or more measurements) corresponding to displayed measurement fields. The user can then enter numbers, text strings, or strings, etc., into the measurement fields. User interfaces 1110 and 1115 together display recommendations for weight loss drugs, as well as expected side effects and weight loss over time. User interfaces 1110 and 1115 also enable the user to select a drug based on the information provided by user interfaces 1110 and 1115.

[0088] Figure 12 shows a flowchart of Method 1200 for a user to select a drug to address obesity, according to an exemplary embodiment. Method 1200 can be performed by any component of System 100, such as a session management service 105, a user device 110, or a user 210. In Method 1200, one or more processors can provide instructions (1202). The instructions may include one or more measurement fields for the user to input. One or more processors can receive the user's measurements (1204). These measurements may be input by the user and / or a measuring device worn by the user. One or more processors can then apply the measurements to a machine learning model (1206). The machine learning model may be trained on a dataset including sample measurements, metrics, and drug IDs. The machine learning model can generate metrics (1208). The metrics may indicate the expected outcome of a weight-loss drug. The metrics may be a set of metrics for each of a set of weight-loss drugs. One or more processors can select one of the drugs (1210). One or more processors can select a drug based on a set of metrics for a set of weight-loss medications. One or more processors can provide a message (1212). The message can be based on the drug and the metrics.

[0089] B. Network and computing environment The various operations described herein can be performed on a computer system. Figure 13 shows a simplified block diagram of a typical server system 1300, a client computing system 1314, and a network 1326 that can be used to implement a particular embodiment of this disclosure. In various embodiments, the server system 1300 or a similar system can implement the services or servers or parts thereof described herein. The client computing system 1314 or a similar system can implement the clients described herein. System 100 described herein may be similar to the server system 1300. The server system 1300 may have a modular design incorporating many modules 1302 (e.g., blades in an embodiment of a blade server), two modules 1302 are shown, but any number can be provided. Each module 1302 may include one or more processing units 1304 and local storage devices 1306.

[0090] One or more processing units 1304 may include a single processor having one or more cores, or multiple processors. In some embodiments, one or more processing units 1304 may include a general-purpose primary processor in addition to one or more dedicated coprocessors, such as a graphics processor or a digital signal processor. In some embodiments, some or all of one or more processing units 1304 may be implemented using customized circuits such as application-specific integrated circuits (ASICs) or user-writable grid arrays (FPGAs). In some embodiments, such integrated circuits execute instructions stored in the circuit itself. In other embodiments, one or more processing units 1304 may execute instructions stored in local memory 1306. Any combination of any type of processor may be included in one or more processing units 1304.

[0091] The local storage device 1306 may include volatile storage media (e.g., DRAM, SRAM, SDRAM, etc.) or non-volatile storage media (e.g., magnetic disks or optical disks, flash memory, etc.). The storage media incorporated into the local storage device 1306 may be fixed, removable, or upgradeable, as desired. The local storage device 1306 may be physically or logically divided into various subunits such as system memory, read-only memory (ROM), and permanent storage. The system memory may be a read-write memory device or a volatile read-write memory such as dynamic random-access memory. The system memory may store some or all of the instructions and data required by one or more processing units 1304 at runtime. The ROM may store static data and instructions required by one or more processing units 1304. The permanent storage may be a non-volatile read-write memory device that can store instructions and data even when the power to module 1302 is turned off. As used herein, the term “storage medium” includes any medium capable of storing data indefinitely (without overwriting, electrical interference, power loss, etc.), but does not include carrier waves or transient electronic signals propagated wirelessly or via wired connections.

[0092] In some embodiments, the local storage device 1306 can store one or more software programs executed by one or more processing units 1304, such as an operating system, or programs that implement various server functions, such as functions of system 1300 or any other system described herein, or any other one or more servers associated with system 1300 or any other system described herein.

[0093] "Software" generally refers to a sequence of instructions that, when executed by one or more processing units 1304, cause the server system 1300 (or a part thereof) to perform various operations, and thus defines one or more specific machine embodiments that execute and perform the operations of a software program. Instructions can be stored as firmware residing in read-only memory for execution by one or more processing units 1304, or as program code stored in a non-volatile storage medium that can be loaded into volatile working memory. Software can be implemented as a single program, or as a collection of separate programs or program modules that interact as desired. To perform the various operations described above, one or more processing units 1306 can retrieve program instructions to execute and data to process from local storage devices 1304 (or non-local storage devices described later).

[0094] In some server systems 1300, multiple modules 1302 can be interconnected via a bus or other interconnect 1308 to form a local area network that supports communication between the modules 1302 and other components of the server system 1300. The interconnect 1308 can be implemented using various technologies, including server racks, hubs, routers, etc.

[0095] The wide area network (WAN) interface 1310 can enable data communication between the local area network (e.g., via interconnect 1308) and a network 1326 such as the Internet. The server system can be connected to the network 1326 in a communicative manner using other technologies, including wired technology (e.g., Ethernet, IEEE 802.3 standard) or wireless technology (e.g., Wi-Fi, IEEE 802.11 standard).

[0096] In some embodiments, the local storage device 1306 is intended to provide working memory to one or more processing units 1304, resulting in fast access to the programs or data being processed while reducing traffic on the interconnect 1308. Storage for larger amounts of data can be provided on the local area network by one or more mass storage subsystems 1308 that can be connected to the interconnect 1312. The mass storage subsystem 1312 can be based on magnetic, optical, semiconductor, or other data storage media. Direct-attached storage, storage area networks, network-attached storage, etc., can be used. Any data storage mechanism or other aggregate of data described herein as being generated, consumed, or maintained by a service or server can be stored in the mass storage subsystem 1312. In some embodiments, additional data storage resources are accessible via the WAN interface 1310 (although this may increase latency).

[0097] The server system 1300 can operate in response to requests received via the WAN interface 1310. For example, one of several modules 1302 may implement a management function and, in response to a received request, assign individual tasks to other modules 902. Task assignment techniques can be used. Once a request is processed, the result can be returned to the requesting party via the WAN interface 1310. Generally, such operations can be automated. Furthermore, in some embodiments, the WAN interface 1310 can connect multiple server systems 1300 to each other, providing a scalable system capable of managing a large volume of activity. Other techniques can be used to manage server systems and server farms (collections of server systems cooperating with each other), including dynamic resource allocation and reallocation.

[0098] The server system 1300 can interact with various user-owned or user-operated devices via a wide-area network such as the internet. An example of a user-operated device is shown in Figure 13 as a client computing system 1314. The client computing system 1314 can be implemented as a consumer device such as a smartphone, other mobile phone, tablet computer, wearable computing device (e.g., smartwatch, glasses), desktop computer, or laptop computer.

[0099] For example, the client computing system 1314 can communicate via the WAN interface 1310. The client computing system 1314 may include computer components such as one or more processing units 1316, a storage device 1318, a network interface 1320, a user input device 1322, and a user output device 1324. The client computing system 1314 may be a computing device implemented in various form factors, such as a desktop computer, a laptop computer, a tablet computer, a smartphone, other mobile computing devices, or a wearable computing device.

[0100] The processing unit 1316 and the storage device 1318 can be the same as the one or more processing units 1304 and local storage device 1306 described above. The appropriate devices can be selected based on the requirements imposed on the client computing system 1314. For example, the client computing system 1314 can be implemented as a "thin" client with limited processing power, or as a high-performance computing device. The client computing system 1314 may have program code executable by one or more processing units 1316 to enable various interactions with the server system 1300.

[0101] The network interface 1320 can provide a connection to a network 1326, such as a wide area network (e.g., the Internet), to which the WAN interface 1310 of the server system 1300 is also connected. In various embodiments, the network interface 1320 may include a wired interface (e.g., Ethernet) or a wireless interface that implements various wireless data communication standards such as Wi-Fi, Bluetooth, or cellular data network standards (e.g., 3G, 4G, LTE, etc.).

[0102] The user input device 1322 may include any device (or more devices) on which the user can send signals to the client computing system 1314, which can interpret the signals as indicating a specific user request or information. In various embodiments, the user input device 1322 may include any or all of the following: a keyboard, touchpad, touchscreen, mouse or other pointing device, scroll wheel, click wheel, dial, button, switch, keypad, microphone, etc.

[0103] The user output device 1324 may include any device on which the client computing system 1314 can provide information to the user. For example, the user output device 1324 may include a display-to-display shared image generated by or transmitted to the client computing system 1314. The display may incorporate various image generation technologies, such as liquid crystal displays (LCDs), light-emitting diode (LED) displays including organic light-emitting diodes (OLEDs), projection systems, and cathode ray tubes (CRTs), along with auxiliary electronic equipment (e.g., digital-to-analog or analog-to-digital converters, signal processors, etc.). Some embodiments may include a device such as a touchscreen that functions as both an input and an output device. In some embodiments, other user output devices 1324 may be provided in addition to, or instead of, the display. Examples include indicator lights, speakers, haptic "display" devices, and printers.

[0104] Some embodiments include electronic components such as microprocessors, memory devices, and other devices that store computer program instructions on a computer-readable storage medium. Many of the features described herein can be implemented as processes designated as a set of program instructions encoded on a computer-readable storage medium. When these program instructions are executed by one or more processing units, they cause one or more processing units to perform the various operations indicated by the program instructions. Examples of program instructions or computer code include machine code generated by a compiler and files containing higher-level code that is executed by a computer, electronic components, or microprocessor using an interpreter. With appropriate programming, one or more processing units 1304 and 1316 can provide a variety of functions, including any of the functions described herein as being executed by a server or a client, or other functions, to the server system 1300 and the client computing system 1314.

[0105] It will be understood that the server system 1300 and client computing system 1314 are illustrative and are subject to modification and alteration. Computer systems used in connection with embodiments of this disclosure may have other functions not specifically described herein. Furthermore, while the server system 1300 and client computing system 1314 are described with reference to specific blocks, it should be understood that these blocks are defined for illustrative purposes only and are not intended to imply a specific physical arrangement of components. For example, different blocks may, but are not required to, be located in the same facility, in the same server rack, or on the same motherboard. Moreover, these blocks do not need to correspond to physically separate components. Blocks can be configured to perform various operations, for example, by programming a processor or by providing appropriate control circuits, and the various blocks may be reconfigurable or non-reconfigurable depending on how the initial configuration is obtained. Embodiments of this disclosure can be realized in a variety of devices, including electronic devices implemented using any combination of circuitry and software.

[0106] While this disclosure has described specific embodiments, those skilled in the art will recognize that numerous modifications are possible. Embodiments of this disclosure can be implemented using a variety of computer systems and communication technologies, including but not limited to the specific examples described herein. Embodiments of this disclosure can be implemented using dedicated components or any combination of programmable processors or other programmable devices. The various operations described herein can be performed with the same processor or any combination of different processors. Where a component is described as being configured to perform a particular operation, such configuration can be achieved, for example, by designing an electronic circuit to perform that operation, by programming a programmable electronic circuit (such as a microprocessor) to perform that operation, or by any combination thereof. Furthermore, while the embodiments described above may refer to specific hardware and software components, those skilled in the art will understand that different combinations of hardware or software components may also be used, that a particular operation described as being implemented in hardware may also be implemented in software, and vice versa.

[0107] Computer programs incorporating various features of this disclosure can be encoded and stored on various computer-readable storage media. Suitable media include optical storage media such as magnetic disks or tapes, compact discs (CDs) or digital versatile discs (DVDs), flash memory, and other non-transient media. The computer-readable media encoded with the program code may be packaged together with compatible electronic devices, or the program code may be provided separately from the electronic devices (for example, via internet download or as separately packaged computer-readable storage media).

[0108] Thus, although this disclosure has described specific embodiments, it will be understood that this disclosure is intended to cover all modifications and equivalents within the scope of the following claims.

Claims

1. A method for selecting a weight-loss drug for a user with obesity: The steps include: receiving one or more physiological measurements of the user by one or more processors; The step of applying the one or more physiological measurements to a machine learning model using one or more processors, wherein the machine learning model is trained using multiple examples, each example including one or more sample physiological measurements of a sample user and a corresponding sample weight-loss drug administered to the sample user; The steps include: generating a metric representing the expected outcome associated with the user's weight loss medication based on the application of the one or more physiological measurements to the machine learning model using one or more processors; A method comprising the step of providing a user device associated with the user with a selection of a weight-loss drug based on the metric indicating the expected outcome, or a message indicating at least one of the expected outcomes associated with the weight-loss drug, using one or more processors.

2. The method according to claim 1, further comprising the step of selecting a weight-loss drug from a plurality of weight-loss drugs based on a plurality of expected outcomes associated with the plurality of weight-loss drugs using one or more processors.

3. The method according to claim 1, wherein the one or more physiological measurements include at least one of the following: obesity index, body weight, blood pressure, heart rate, smoking status, glucose excretion, total metabolic panel, complete blood count, lipase level, thyroid panel, magnesium level, HgA1c, fasting glucose, energy expenditure, physical activity, hormone levels, body weight, body fat percentage, genetic labeling, assessment of gut microbiota, or energy intake.

4. The machine learning model includes one or more corresponding weights for generating one or more values, and each of these one or more corresponding weights includes at least one of binary weights or continuous weights; The method according to claim 1, wherein the step of generating the metric further includes the step of generating the metric based on the one or more values ​​by the one or more processors.

5. The method according to claim 1, wherein the weight-loss agent is selected from a GLP-1 receptor agonist or a GIP receptor agonist.

6. The method according to claim 5, wherein the GLP-1 receptor agonist is selected from one or more of semaglutide, liraglutide, exenatide, and dulaglutide, and the GIP receptor agonist comprises tylzepatide.

7. The method according to claim 1, wherein the expected outcome further comprises at least one administration parameter of the dose-reducing drug, the administration parameter further comprises at least one of the following: the dose of the dose-reducing drug, the timing of administration of the dose-reducing drug, the frequency of administration, the route of administration, the dose-increase protocol, the circumstances of administration, or any combination thereof.

8. The method according to claim 1, further comprising identifying at least one of the expected outcomes: discontinuation of treatment, side effects, or therapeutic efficacy.

9. The method according to claim 8, wherein the aforementioned side effects are selected from nausea, vomiting, diarrhea, early satiety, loss of appetite, anorexia, dizziness, increased heart rate, indigestion, headache, hypoglycemia, kidney or ureteral stones, pancreatitis, diabetic retinopathy, depression, suicidal ideation or attempt, abdominal pain, acute kidney injury, muscle weakness and atrophy, constipation, or any combination thereof.

10. The method according to claim 1, wherein the step of generating the metric further includes the step of generating a plurality of metrics for a plurality of expected outcome parameters based on the step of applying the one or more physiological measurements to the machine learning model.

11. The method according to claim 8, wherein the expected outcome further comprises at least one expected outcome parameter, the expected outcome parameter further comprising at least one of the timing of termination, the cause of termination, the probability of termination, the mitigation of termination, or any combination thereof.

12. The aforementioned expected outcome further includes at least one expected outcome parameter, The method according to claim 8, wherein the expected outcome parameter includes at least one of the timing of the onset of the adverse event, the duration of the adverse event, the probability of the adverse event, the mitigation of the adverse event, or a combination thereof.

13. The method according to claim 8, wherein the expected outcome further comprises expected outcome parameters, the expected outcome parameters comprising at least one of weight loss, fat loss, fasting blood glucose, cholesterol levels, hormone levels, duration of fat loss, risk of weight regrowth, change in body mass index, or any combination thereof.

14. The method according to claim 1, further comprising the step of generating a simulation by one or more processors that identifies a plurality of expected outcomes across a plurality of corresponding time points.

15. The method according to claim 14, wherein the simulation includes a display of the multiple expected outcomes across the multiple corresponding points in time, the display including at least one of a timeline, graph, video, audio, or avatar.

16. The method according to claim 14, wherein the plurality of expected outcomes identified by the simulation include at least one expected outcome parameter, the expected outcome parameter includes at least one of the timing of termination, cause of termination, probability of termination, mitigation of termination, or any combination thereof.

17. The method according to claim 14, wherein the plurality of expected outcomes identified by the simulation include at least one expected outcome parameter, the expected outcome parameter includes at least one of the timing of the onset of an adverse event, the duration of an adverse event, the probability of an adverse event, the mitigation of an adverse event, or a combination thereof.

18. The method according to claim 14, wherein the plurality of expected outcomes identified by the simulation include expected outcome parameters, the expected outcome parameters include at least one of weight loss, fat loss, fasting blood glucose, cholesterol levels, hormone levels, duration of fat loss, risk of weight regain, change in body mass index, or any combination thereof.

19. The method according to claim 1, wherein the user has a BMI greater than 25, a body fat percentage greater than 20%, type 1 diabetes, type 2 diabetes, or non-alcoholic steatohepatitis (NASH).

20. The method according to claim 1, wherein the message is provided to the user device or the user's clinician.

21. The method according to claim 1, wherein the one or more physiological measurements of the user are obtained by at least one of the user device or an instrumentation device attached to the user.

22. The method according to claim 1, wherein the step of generating the metric indicating the expected outcome includes the step of one or more processors identifying the metric based on at least one of (i) the mean of the plurality of examples, (ii) a weighted combination of the plurality of examples, or (iii) a comparison with a dataset consisting of the plurality of examples.

23. The step of receiving one or more physiological measurements further includes the step of receiving the one or more physiological measurements from at least one of the user device or instrumentation device over a certain period of time by one or more processors, The step of applying the machine learning model further includes the step of applying the one or more physiological measurements to the machine learning model in response to the passage of a certain period of time using one or more processors, The method according to claim 1, wherein the step of generating the metric further includes the step of generating the metric, which indicates the expected outcome associated with the weight-loss drug over a subsequent period, using one or more processors.

24. The steps include: receiving one or more subsequent physiological measurements of the user by one or more processors; The steps include: applying the one or more subsequent physiological measurements to the machine learning model in response to the reception of the one or more subsequent physiological measurements by the one or more processors; The steps include: generating a subsequent metric that indicates the expected subsequent outcome associated with the user's subsequent weight loss medication, based on the application of the one or more subsequent physiological measurements to the machine learning model, using one or more processors; The process further includes the step of providing the user device with a subsequent message within a defined period of time for the reception of the one or more subsequent physiological measurements by the one or more processors, wherein the subsequent message is: (a) Selection of a subsequent dose-reducing drug based on the subsequent metric indicating the expected outcome of the subsequent dose-reducing drug or (b) The method according to claim 1, which provides at least one of the subsequent expected outcomes associated with the subsequent weight-loss drug.

25. The method according to claim 24, wherein the defined period is in the range of 1 second to 1 hour.

26. A system comprising one or more processors, wherein the processors are Receive one or more physiological measurements of the user; The one or more measurements are applied to a machine learning model, the machine learning model is trained using multiple examples, each example including one or more sample physiological measurements of a sample user and a corresponding sample weight-loss drug administered to the sample user; Based on the application of one or more physiological measurements to the machine learning model, a metric is generated that shows the expected outcome associated with the user's weight loss medication; A system configured to provide a user device associated with the user with a selection of a weight-loss drug based on the metric indicating the expected outcome, or a message indicating at least one of the expected outcomes associated with the weight-loss drug.

27. The system according to claim 26, wherein the one or more processors are further configured to select a weight-loss drug from a plurality of weight-loss drugs based on a plurality of expected outcomes associated with the plurality of weight-loss drugs.

28. The system according to claim 26, wherein the one or more physiological measurements include at least one of the following: obesity index, body weight, blood pressure, heart rate, smoking status, glucose excretion, total metabolic panel, complete blood count, lipase level, thyroid panel, magnesium level, HgA1c, fasting glucose, energy expenditure, physical activity, hormone levels, body weight, body fat percentage, genetic labeling, assessment of gut microbiota, or energy intake.

29. The machine learning model includes one or more corresponding weights for generating one or more values, and each of these one or more corresponding weights includes at least one of binary weights or continuous weights; The system according to claim 26, wherein one or more processors are further configured to generate the metric based on the one or more values.

30. The system according to claim 26, wherein the weight-loss agent is selected from a GLP-1 receptor agonist or a GIP receptor agonist.

31. The system according to claim 30, wherein the GLP-1 receptor agonist is selected from one or more of semaglutide, liraglutide, exenatide, and dulaglutide, and the GIP receptor agonist comprises tylzepatide.

32. The system according to claim 26, wherein the expected outcome further comprises at least one administration parameter of the weight-reducing drug, the administration parameter further comprises at least one of the following: the dose of the weight-reducing drug, the timing of administration of the weight-reducing drug, the frequency of administration, the route of administration, the dose-increase protocol, the circumstances of administration, or any combination thereof.

33. The system according to claim 26, further comprising identifying at least one of the expected outcomes: discontinuation of treatment, side effects, or therapeutic efficacy.

34. The system according to claim 33, wherein the aforementioned side effects are selected from nausea, vomiting, diarrhea, early satiety, loss of appetite, anorexia, dizziness, increased heart rate, indigestion, headache, hypoglycemia, kidney or ureteral stones, pancreatitis, diabetic retinopathy, depression, suicidal ideation or attempt, abdominal pain, acute kidney injury, muscle weakness and atrophy, constipation, or any combination thereof.

35. The system according to claim 26, wherein, in order to generate the metrics, one or more processors are further configured to generate multiple metrics of multiple expected outcome parameters based on the application of one or more physiological measurements to the machine learning model.

36. The system according to claim 33, wherein the expected outcome further comprises at least one expected outcome parameter, the expected outcome parameter further comprising at least one of the timing of termination, the cause of termination, the probability of termination, the mitigation of termination, or any combination thereof.

37. The system according to claim 33, wherein the expected outcome further comprises at least one expected outcome parameter, the expected outcome parameter comprising at least one of the timing of the onset of the adverse event, the duration of the adverse event, the probability of the adverse event, the mitigation of the adverse event, or a combination thereof.

38. The system according to claim 33, wherein the expected outcome further comprises expected outcome parameters, the expected outcome parameters comprising at least one of weight loss, fat loss, fasting blood glucose, cholesterol levels, hormone levels, duration of fat loss, risk of weight regain, change in body mass index, or any combination thereof.

39. The system according to claim 26, wherein one or more processors are further configured to generate simulations that identify a plurality of expected outcomes across a plurality of corresponding time points.

40. The system according to claim 39, wherein the simulation includes a display of the multiple expected outcomes across the multiple corresponding points in time, the display including at least one of a timeline, graph, video, audio, or avatar.

41. The system according to claim 39, wherein the plurality of expected outcomes identified by the simulation include at least one expected outcome parameter, the expected outcome parameter includes at least one of the timing of termination, the cause of termination, the probability of termination, the mitigation of termination, or any combination thereof.

42. The system according to claim 39, wherein the plurality of expected outcomes identified by the simulation include at least one expected outcome parameter, the expected outcome parameter includes at least one of the timing of the onset of an adverse event, the duration of an adverse event, the probability of an adverse event, the mitigation of an adverse event, or a combination thereof.

43. The plurality of expected outcomes identified by the simulation include expected outcome parameters, The system according to claim 39, wherein the expected outcome parameters include at least one of weight loss, fat loss, fasting blood glucose, cholesterol levels, hormone levels, duration of fat loss, risk of weight regain, change in body mass index, or any combination thereof.

44. The system according to claim 26, wherein the user has a BMI greater than 25, a body fat percentage greater than 20%, type 1 diabetes, type 2 diabetes, or non-alcoholic steatohepatitis (NASH).

45. The system according to claim 26, wherein the message is provided to the user device or the user clinician.

46. The system according to claim 26, wherein the one or more physiological measurements of the user are acquired by at least one of the user device or an instrumentation device attached to the user.

47. The system according to claim 26, wherein, in order to generate the metric representing the expected outcome, one or more processors are configured to identify the metric based on at least one of (i) the mean of the plurality of examples, (ii) a weighted combination of the plurality of examples, or (iii) a comparison with a dataset comprising the plurality of examples.

48. In order to receive the one or more physiological measurements, the one or more processors are further configured to receive the one or more physiological measurements from at least one of the user device or instrumentation device over a period of time. In order to apply the one or more physiological measurements to the machine learning model, the one or more processors are further configured to apply the one or more physiological measurements to the machine learning model in response to the passage of a certain period of time. To generate the metric, one or more processors are further configured to generate the metric indicating the expected outcome associated with the weight-loss drug over a subsequent period, the system according to claim 26.

49. The aforementioned one or more processors: Receiving one or more subsequent physiological measurements from the user; In response to the reception of one or more subsequent physiological measurements, the one or more subsequent physiological measurements are applied to the machine learning model; Based on the application of the one or more subsequent physiological measurements to the machine learning model, a subsequent metric is generated that shows the expected subsequent outcome associated with the user's subsequent weight loss medication; The system is further configured to provide the user device with a subsequent message within a defined period of time in response to the reception of one or more subsequent physiological measurements, wherein the subsequent message is: (a) Selection of a subsequent dose-reducing drug based on the subsequent metric indicating the expected outcome of the subsequent dose-reducing drug or (b) The system according to claim 26, which shows at least one of the subsequent expected outcomes associated with the subsequent weight-loss drug.

50. The system according to claim 49, wherein the defined period is in the range of 1 second to 1 hour.

Images