Structured treatment plan generation and content integration in digital therapeutic applications

Abstract

Various examples, systems, and / or methods related to the development process and workflow of digital therapy. One or more processors can process natural language therapy input, generate structured treatment plans for use in digital therapy applications, retrieve content from a database to match the structured treatment plans, and arrange the content in a design system based on the structured treatment plans.

Description

[0001] Cross-references to related applications

[0002] This application claims the benefit and priority of U.S. Nonprovisional Application No. 19 / 056,168, filed February 18, 2025, entitled “STRUCTURED TREATMENT PLANGENERATION AND CONTENT INTEGRATION IN DIGITAL THERAPEUTIC APPLICATIONS,” which is incorporated herein by reference in its entirety. Technical Field

[0003] This disclosure relates to the field of digital therapy, and more specifically to structured treatment plan generation and content integration in digital therapy applications. Background Technology

[0004] Treatment planning for health conditions, including mental health disorders, chronic diseases, and / or rehabilitation, typically involves structured, multi-step interventions, exercise, and / or condition-specific treatment workflows. Common conditions for structured treatment planning include post-traumatic stress disorder (PTSD), substance use disorders, cardiovascular disease, obesity, migraines, multiple sclerosis, cancer-related conditions, and / or any other acute or chronic medical or psychological condition. Challenges in treatment planning stem from a variety of factors, such as variability in user needs, differences in provider approaches, limitations of existing digital therapy tools, and / or a lack of automation in treatment plan generation. For example, factors contributing to these challenges include non-standardized treatment methods (e.g., variations in the ordering of components in a treatment plan, variations in prescribing activities), insufficient personalization (e.g., failure to consider user-specific symptoms, comorbidities, and treatment responses), and / or accessibility barriers (e.g., limited provider availability, limitations in delivering treatment to different populations). Other factors include inefficiencies in manual treatment plan development (e.g., time-consuming development of structured treatment plans), a lack of integration between treatment planning tools and digital therapy applications, and / or existing treatment plans that are difficult to adapt to real-time user data.

[0005] The effectiveness and / or efficacy of treatment plans are often hampered by limitations in structure, adaptability, and accessibility. Users receiving treatment plans that are not structured to their individual progress, symptom severity, or engagement may experience delayed treatment outcomes, decreased adherence, increased frustration, and / or diminished motivation to complete components of the treatment plan. Furthermore, healthcare providers and digital therapy developers face challenges in constructing, modifying, and / or implementing structured therapies, which limits the scalability of personalized treatment. Users engaged in rigid and / or non-adaptive treatment plans are at greater risk of difficulty maintaining engagement, progressing through therapy practice, responding to situation-specific triggers, integrating treatment recommendations into daily life, and / or adhering to long-term health strategies. The overall efficacy of structured treatment plans is often reduced due to these limitations.

[0006] At the technical level, inefficiencies in treatment planning present challenges in designing structured execution frameworks, which provide for treatment plan sequencing, adaptation to user progress, and / or integration with digital therapy applications. Specifically, previous systems lacked mechanisms to translate natural language therapy goals into structured execution frameworks, limiting their ability to streamline treatment plan delivery. Furthermore, users and healthcare providers relying on non-automated treatment planning systems face increased risks of inconsistencies between treatment goals and prescribed treatments (e.g., lack of conditional logic for adjusting activities based on user progress), limited continuity of care across different health conditions (e.g., difficulties in integrating behavioral therapy with medical treatment plans), and / or reduced responsiveness to evolving user needs. For example, manually designed plans may fail to incorporate dynamic feedback loops, resulting in fixed pathways that fail to adapt to user-reported pain levels. In another example, structured weight management programs may lack adaptability based on real-time user-reported dietary adherence. Yet another example is a rehabilitation plan for postoperative recovery that fails to dynamically adjust exercise intensity based on functional progress assessments.

[0007] Traditional solutions suffer from inefficiencies due to a lack of automation and personalization mechanisms, and / or insufficient integration with the digital ecosystem. In particular, reliance on manual treatment plan generation, limited capabilities of digital tools for structured adaptive execution frameworks, and / or difficulty in integrating existing treatment modules into treatment workflows reduce the scalability and effectiveness of digital treatment planning. For example, previous workflow-based digital treatment solutions may have required providers to provide substantial manual input to define treatment pathways, thus reducing efficiency in clinical settings. In another example, traditional treatment planning software may be unable to use models to create structured, multi-step treatment processes based on user-specific input. Yet another example is the inability of previous digital applications to dynamically adjust content presentation, treatment activities, or user interactions based on predefined scheduling parameters and conditional logic. Failure to incorporate automation, structuring, and adaptive mechanisms into treatment planning solutions can lead to suboptimal treatment engagement and poor user outcomes. Summary of the Invention

[0008] This paper introduces a system and method for generating structured treatment plans from natural language input using generative artificial intelligence (genAI) models. The system described involves processing natural language therapeutic input, generating structured treatment plans for use in digital therapy applications, retrieving content from a database that matches the structured treatment plans, and arranging the content within a designed system based on the structured treatment plans. This allows for rapid iteration and deployment of a large number of personalized digital therapy applications, each individually customized for its respective user. The generative model (e.g., a model for structuring treatment plans, ordering treatment components, and / or generating conditional logic) generates a structured representation of the treatment plan based on user-provided or physician-provided input. For example, the system can process a request such as: “Create a 3-month treatment program for an individual with PTSD whose triggers occur in the afternoon, midday, and midweek, including breathing exercises based on predefined templates, grounding exercises, interactive therapeutic interventions, medication administration, and a schedule for the first five days of treatment,” and generate a treatment plan that defines the sequence of therapeutic actions the user must perform, retrieve relevant content from the database, arrange the content in the design system, and / or structure the treatment plan to support its presentation to the user in a personalized digital therapy application.

[0009] Specifically, this improved approach facilitates the automated structuring of large volumes of treatment plans, including dynamically generated execution logic, predefined treatment workflows, and modular content retrieval, aimed at improving treatment adherence, enhancing engagement, and simplifying intervention delivery. The system applies structured treatment goals to digital therapy applications, retrieving and structuring treatment components to present to users, such as the sequence of smoking cessation tasks based on a predefined execution framework. Structured data packages and / or structured execution frameworks provide a representation of the treatment plan, defining how treatment components are arranged, scheduled, and / or conditionally triggered within the digital therapy application.

[0010] Technological improvements can be achieved through structured, modular content retrieval and / or adaptive treatment sequencing of treatment plans executed in a digital execution environment (e.g., continuous, periodic, and / or user-progress-based). A structured execution framework manages treatment components to provide a structured representation of the treatment plan by defining placement, scheduling, and conditional logic. The system generates the structured execution framework by processing treatment plans, retrieving relevant content from a database, and organizing the retrieved content into a structured data representation within the design system. The structured execution framework defines how treatment components are linked, scheduled, and / or conditionally presented in a digital therapy application. For example, the system can structure the treatment plan by associating retrieved content with scheduled treatment actions, specifying transition conditions for when treatment activities are presented, and defining execution dependencies to support structured intervention delivery.

[0011] This integration represents a technological improvement over previous treatment planning techniques that relied on manual design, rigid treatment schedules, and / or predefined treatment pathways, lacking automated structuring and / or adaptive progression mechanisms. If an imbalance exists between the treatment structure and the user's specific treatment needs, the structured execution framework and / or modular content retrieval mechanism disclosed herein can improve personalization, adaptability, and / or efficiency in treatment plan generation. The combination of the structured execution framework and content retrieved from databases facilitates the scalable deployment of large numbers of digital therapeutic treatment plans and / or the automated adaptation of treatment strategies.

[0012] Furthermore, the digital therapy application described in this paper addresses the lack of structured treatment generation in previous methods by integrating real-time treatment plan generation, structured sorting logic, and / or modular content retrieval. The system uses a structured execution framework with predefined conditional logic to dynamically organize and / or arrange content retrieved from the database, ensuring that treatment plans align with a structured treatment workflow. This structured execution framework can incorporate structured rules to determine how treatment content is retrieved, arranged, and / or presented within the digital therapy application.

[0013] Therefore, digital therapy applications address inefficiencies in treatment planning by providing a structured execution framework that defines treatment progress, schedules treatment activities, and / or retrieves content based on structured queries of a database and incorporates it into the design system. The structured execution framework provides a structured format for arranging, linking, and presenting treatment content within digital therapy applications. By integrating generative AI, structured data frameworks, and / or modular content retrieval, the system improves the scalability and / or adaptability of a wide range of digital therapy application developments, thereby enhancing the delivery of structured treatment plans and increasing their accessibility. Furthermore, the digital therapy application system discussed in this paper addresses the challenge of creating generative AI-based platforms that are individually tailored for a large number of users requiring therapy across physical, cognitive, social, and / or behavioral domains.

[0014] Furthermore, the systems and methods described herein combine structured execution models, conditional logic frameworks, and / or dynamic treatment plan generation techniques to improve the efficiency of large-scale treatment plan structuring, reduce the complexity of manual design, and enhance integration with global digital platforms (e.g., interoperability with electronic health record systems, wearable devices, and / or other products and devices that can interact with the systems described herein). By providing structured, dynamically generated treatment pathways, the disclosed systems and methods improve consistency, structured content placement, and / or the execution of treatment workflows in digital therapeutic applications.

[0015] Some embodiments relate to systems including one or more processors coupled to memory. The one or more processors are configured to receive natural language input to generate a treatment plan for a digital therapy application to address a condition. The one or more processors are configured to apply the natural language input and the condition as input to at least one generative model to generate a treatment plan for the condition. In some embodiments, the treatment plan includes a list of treatment components and a schedule for displaying at least one treatment component in the list of treatment components. The one or more processors are configured to retrieve multiple content items from a database corresponding to at least one treatment component in the list of treatment components in the treatment plan. The one or more processors are configured to arrange the list of treatment components of the treatment plan into a structured data packet using the multiple content items and the schedule for display in a digital therapy application. The one or more processors are configured to provide the structured data packet.

[0016] In some embodiments, one or more processors coupled to the memory are further configured to generate a plurality of instructions using structured data packets, the plurality of instructions being configured to cause the digital therapy application to display at least one of a plurality of content items according to the arrangement of a list of treatment components of a treatment plan.

[0017] In some embodiments, a structured data packet corresponds to a mapping of multiple content items, schedules, and / or multiple execution dependencies of a structured data packet.

[0018] In some embodiments, a treatment plan includes a schedule of multiple conditional steps that identify at least one treatment component in a list of treatment components for the treatment process.

[0019] In some embodiments, the treatment schedule includes multiple conditional logics corresponding to at least one treatment component and initiating at least one subsequent treatment component. In some embodiments, the at least one subsequent treatment component is provided based at least on the completion or failure of at least one previous treatment component or the completion or non-compliance of predefined criteria for the at least one previous treatment component.

[0020] In some embodiments, one or more processors coupled to the memory are further configured to compile skeleton code retrieved from at least one database or generated using structured data packets. In some embodiments, the skeleton code includes at least one program construct corresponding to a list of treatment components and a schedule of the treatment plan.

[0021] In some embodiments, one or more processors coupled to the memory are further configured to provide a treatment plan or skeleton code to an administrator for review. In some embodiments, one or more processors coupled to the memory are further configured to receive at least one update to the treatment plan or skeleton code from the administrator before providing a structured data package or using the skeleton code in a digital therapy application.

[0022] Some embodiments relate to systems including one or more processors coupled to memory. The one or more processors are configured to receive natural language input to generate content for a digital therapy application to resolve a situation. The one or more processors are configured to apply the natural language input and the situation as input to at least one generative model to generate a directed process graph for the situation, the directed process graph including multiple actions and conditional logic, according to which at least one of the multiple actions is provided for presentation. The one or more processors are configured to identify at least one data object corresponding to at least one of the multiple actions included in the directed process graph. The one or more processors are configured to use the at least one data object and the directed process graph to generate a structured execution framework, the structured execution framework identifying (i) multiple content items for identifying at least one of the multiple actions and (ii) corresponding conditional logic, according to which at least one of the multiple actions is provided for presentation. The one or more processors are configured to provide the structured execution framework.

[0023] In some embodiments, one or more processors coupled to the memory are further configured to generate a plurality of instructions using a structured execution framework, the plurality of instructions being configured to cause the digital therapy application to display content including at least one of a plurality of content items according to a directed process graph.

[0024] In some embodiments, the structured execution framework corresponds to a mapping of multiple content items, corresponding conditional logic, multiple timing parameters, and / or multiple execution dependencies in a directed process graph.

[0025] In some embodiments, one or more processors coupled to memory are further configured to use a simulator to simulate the performance of multiple actions based on processing at least a subset of instructions according to a structured execution framework. In some embodiments, the multiple instructions include at least one operation for performing at least one of multiple actions corresponding to at least one timing parameter.

[0026] In some embodiments, the directed process diagram includes a hierarchical arrangement of components representing multiple actions, the arrangement being based on multiple timing parameters to identify the order and duration corresponding to at least one of the multiple actions.

[0027] In some embodiments, the at least one generative model is at least one of the following: (i) a deep learning model, (ii) a supervised learning model, and / or (iii) an unsupervised learning model. In some embodiments, causing the at least one generative model to generate a directed process graph includes processing natural language input to extract a plurality of executable elements, associating the plurality of executable elements with a plurality of actions and a plurality of conditional logics, and / or arranging the plurality of actions and the plurality of conditional logics into a hierarchical structure.

[0028] In some embodiments, the conditional logic of the directed process graph includes multiple conditional logics corresponding to at least one treatment component and initiating at least one subsequent treatment component. In some embodiments, the at least one subsequent treatment component is provided at least based on the completion or failure of at least one previous treatment component or the completion or non-completion of a predefined criterion for the at least one previous treatment component.

[0029] In some embodiments, natural language input includes instructions that identify at least one of the following: (i) treatment duration, (ii) at least one treatment component, (iii) at least one action corresponding to the at least one treatment component, and / or (iv) at least one criterion for completing the at least one treatment component.

[0030] In some embodiments, a directed process graph includes multiple execution paths. In some embodiments, the multiple execution paths of the directed process graph correspond to multiple treatment workflows of a treatment process. In some embodiments, each of the multiple treatment workflows includes a sequence of treatment actions to resolve or manage a condition.

[0031] In some embodiments, at least one of the plurality of actions corresponds to a treatment component or a response to a diagnostic query. In some embodiments, one or more processors coupled to the memory are further configured to update the activity record to include completion status, timestamp data, and / or data generated or acquired during the treatment component or diagnostic query in response to the completion of the treatment component or the provision of a response to a diagnostic query.

[0032] In some embodiments, one or more processors coupled to memory are further configured to compile skeleton code retrieved from at least one database, or to generate skeleton code including at least one program construct corresponding to multiple actions and corresponding conditional logic using a structured execution framework and a directed process graph.

[0033] In some embodiments, one or more processors coupled to memory are further configured to provide a directed process graph or skeleton code to an administrator for review. In some embodiments, one or more processors coupled to memory are further configured to receive at least one update to the directed process graph or skeleton code from the administrator before providing a structured execution framework or using the skeleton code in a digital therapy application.

[0034] In some embodiments, identifying the at least one data object includes querying a database that maintains the at least one data object. In some embodiments, querying the database includes using at least one application programming interface (API) to request the at least one data object corresponding to at least one historical treatment component or configuration.

[0035] In some embodiments, the at least one data object includes at least one content item from a plurality of content items and a corresponding parameter. In some embodiments, the corresponding parameter includes at least one of content item relationships, temporal parameters, and / or metadata for representing the at least one historical treatment component as a node or relationship in a directed process graph.

[0036] Some embodiments relate to a method. The method includes receiving natural language input via one or more processors to generate content for a digital therapy application to resolve a situation. The method includes applying the natural language input and the situation as input to at least one generative model via one or more processors to cause the at least one generative model to generate a directed process graph for the situation. In some embodiments, the directed process graph includes a plurality of actions and conditional logic, according to which at least one of the plurality of actions is provided for presentation. The method includes identifying, via one or more processors, at least one data object corresponding to at least one of the plurality of actions included in the directed process graph. The method includes generating the directed process graph using the at least one data object and the directed process graph via one or more processors. In some embodiments, a structured execution framework may identify (i) a plurality of content items for identifying at least one of the plurality of actions and (ii) corresponding conditional logic, according to which at least one of the plurality of actions is provided for presentation. The method includes providing the structured execution framework via one or more processors.

[0037] In some embodiments, the method includes generating a plurality of instructions using a structured execution framework via one or more processors, the plurality of instructions being configured to cause a digital therapy application to display content comprising at least one of a plurality of content items according to a directed process graph.

[0038] In some embodiments, the structured execution framework corresponds to a mapping of multiple content items, corresponding conditional logic, multiple timing parameters, and / or multiple execution dependencies in a directed process graph.

[0039] In some embodiments, the method includes simulating the performance of multiple actions by using a simulator through one or more processors to process at least a portion of instructions according to a structured execution framework. In some embodiments, the multiple instructions include at least one operation for performing at least one of the multiple actions corresponding to at least one timing parameter.

[0040] In some embodiments, at least one generative model is at least one of the following: (i) a deep learning model, (ii) a supervised learning model, and / or (iii) an unsupervised learning model. In some embodiments, causing the at least one generative model to generate a directed process graph includes processing natural language input to extract a plurality of executable elements, associating the plurality of executable elements with a plurality of actions and a plurality of conditional logics, and / or arranging the plurality of actions and the plurality of conditional logics into a hierarchical structure.

[0041] In some embodiments, the conditional logic of the directed process graph includes multiple conditional logics corresponding to at least one treatment component and initiating at least one subsequent treatment component. In some embodiments, the at least one subsequent treatment component is provided at least based on the completion or failure of at least one previous treatment component or the completion or non-completion of a predefined criterion for the at least one previous treatment component.

[0042] In some embodiments, natural language input includes instructions that identify at least one of the following: (i) treatment duration, (ii) at least one treatment component, (iii) at least one action corresponding to the at least one treatment component, and / or (iv) at least one criterion for completing the at least one treatment component.

[0043] In some embodiments, a directed process graph includes multiple execution paths. In some embodiments, the multiple execution paths of the directed process graph correspond to multiple treatment workflows of a treatment process. In some embodiments, each of the multiple treatment workflows includes a sequence of treatment components to resolve or manage a condition. Attached Figure Description

[0044] The system and method of the present invention for structured treatment plan generation and content integration in digital therapy applications are described in detail below with reference to the accompanying drawings, wherein:

[0045] Figure 1 This is a block diagram illustrating an example of a system according to some embodiments of the present disclosure;

[0046] Figure 2 This is a flowchart of an example method for generating a structured execution framework according to some embodiments of this disclosure;

[0047] Figures 3A-3B These are example interfaces for interacting with generative models to generate structured execution frameworks, according to some embodiments of this disclosure;

[0048] Figure 4 This is a block diagram of an example server system and an example administrator device according to some embodiments of this disclosure, and according to illustrative embodiments. Detailed Implementation

[0049] This disclosure relates to systems and methods for generating structured treatment plans from natural language input, retrieving content from databases, and providing structured execution frameworks or structured data packages for use in digital therapy applications. For example, the systems and methods according to this disclosure facilitate the planning and modeling of structured therapy workflows by processing natural language and / or free text descriptions, generating directed process graphs, and associating retrieved content with structured therapeutic treatment plans. The system may receive natural language input describing treatment goals, apply the input to at least one generative artificial intelligence (AI) model configured to generate directed process graphs, and retrieve corresponding content items from a database to generate a structured execution framework. The structured execution framework may define the layout, arrangement, and / or conditional logic to manage treatment components for presentation in digital therapy applications.

[0050] Some traditional treatment planning techniques rely on manually curated sequences of treatment modules, strict treatment schedules, and / or predefined activity paths, lacking flexibility in structuring and organizing treatment plans. These prior approaches fail to establish treatment progression with structured execution rules or define flexible relationships between treatment workflows and content sources. Furthermore, the lack of integrated mechanisms for structuring and retrieving treatment content based on predefined treatment plan models leads to inefficient digital therapy application development. For example, previous systems failed to define structured treatment plans, map retrieved content to structured treatment plan workflows, and / or establish treatment sequencing within model execution frameworks. The systems and methods described in this paper overcome these limitations by generating and / or otherwise modeling digital therapeutic treatment plans using various models and / or functions that structure the output based on natural language input, retrieved content, and predefined treatment parameters.

[0051] The systems and methods according to this disclosure include receiving natural language input from an administrator device during processing, corresponding to a treatment goal for a given condition. For example, the system may apply the natural language input to at least one AI model to generate a directed process graph, defining multiple actions and corresponding conditional logic. The system may identify at least one data object corresponding to at least one of the multiple actions and retrieve content items associated with the identified data object from a database. The retrieved content items may be used to generate a structured execution framework, which defines the arrangement, sequence, and / or transformation logic of treatment components in a digital therapy application.

[0052] The system can also process treatment plans by modeling the relationships between retrieved content items and treatment components within a structured execution framework. For example, retrieved content may include predefined treatment exercises, multimedia resources, and / or interactive interface elements. The structured execution framework organizes these components based on predefined timing parameters, user engagement rules, and / or conditional transitions derived from directed process graphs. The structured execution framework can also encode conditional logic to manage when and how treatment components are presented.

[0053] In some embodiments, the system can generate a structured execution framework that includes predefined timing attributes, execution dependencies, and / or ordering logic. The structured execution framework can define the hierarchical arrangement of treatment components. Treatment components can be arranged based on timing parameters specifying the order and duration of at least one of a plurality of treatment components. The structured execution framework can also define multiple execution paths corresponding to alternative treatment workflows.

[0054] Furthermore, the system and method include generating a structured treatment workflow that contains predefined conditional logic and structured relationships between treatment components. A structured execution framework may include multiple conditional logics associated with at least one treatment component and associate at least one subsequent treatment component with corresponding conditional logics. Subsequent treatment components may be linked to at least one previous treatment component based on predefined conditional logic corresponding to sequence constraints and / or execution dependencies. For example, the structured execution framework may associate breathing exercises with subsequent treatment components based on predefined conditional logic rather than user-specific completion states. In another example, the structured execution framework may specify transition conditional logic for when treatment components are displayed based on timing parameters or predefined relationships between treatment components. The structured execution framework can provide a structured representation of ordering without requiring direct user input to determine treatment progress.

[0055] The system can also process structured treatment workflows by retrieving historical treatment components and configurations from a database. The system can query the database to identify at least one data object corresponding to a treatment component. The retrieved data object may include at least one content item and corresponding parameters. The corresponding parameters may include content item relationships, temporal parameters, and / or metadata, which are used to structure the retrieved actions in a directed process graph. The retrieved content items can be incorporated into a structured execution framework to support digital treatment application modeling.

[0056] The systems and methods described herein improve upon traditional technical solutions by generating structured execution frameworks that facilitate efficient, scalable, and / or adaptive modeling of therapeutic workflows. The disclosed systems provide an improved technical solution for structured digital therapeutic treatment planning by using generative AI models to plan therapeutic workflows, retrieve structured content, and / or define structured data representations. For example, the integration of directed flowcharts, structured execution frameworks, and / or temporal logic ensures that the generated treatment plans match structured intervention models and can be integrated into digital therapeutic applications in a rapid, scalable, and user-person-person-person-person-person-personalized manner.

[0057] The systems and methods described herein can be used for a variety of purposes, including planning structured treatment plans, generating treatment workflows, modeling structured execution rules for treatment interventions, and / or integrating retrieved content into predefined treatment frameworks. Furthermore, these methods can improve the efficiency of digital treatment planning by reducing reliance on manually defined treatment planning workflows, allowing for structured modeling of treatment sequences, and / or providing technical consistency in treatment plan delivery.

[0058] refer to Figure 1 , Figure 1 This is a block diagram illustrating a system 100 according to some embodiments of the present disclosure. System 100 includes components that can be implemented as discrete hardware, distributed components, and / or combinations of hardware, firmware, and / or software. The components of system 100 can be configured to perform the functions described herein, and / or the arrangement of these components can vary depending on the embodiment. Figure 1 The arrangements shown are provided as examples and are not limiting. Other configurations and elements (e.g., machines, interfaces, functions, operating sequences, and / or functional groupings) may be included in addition to and / or in place of those described. Depending on operational requirements, some components may be omitted in a particular embodiment. Furthermore, some functional entities in system 100 may be grouped differently or implemented in alternative locations. The functions associated with the components of system 100 may be executed by one and / or more processors that execute instructions stored in memory. The instructions may instruct the processor to perform operations corresponding to one or more components of system 100. System 100 may also include interfaces or connections (e.g., wired and / or wireless communication technologies) that facilitate communication between components. Embodiments may vary based on the requirements of the system and its operating environment. In some embodiments, [the following may be used]... Figure 4 Example server system 400 and / or Figure 4 The example administrator device 414 (described as "client computing system 414") has similar components, features, and / or functions to perform the systems, methods, and / or processes described herein.

[0059] System 100 can implement at least a portion of a digital therapy workflow, such as a treatment planning workflow, content generation workflow, behavioral intervention workflow, and / or condition-specific treatment adaptation workflow. System 100 can generate structured execution frameworks to provide targeted therapeutic content for addressing various conditions. System 100 can generate execution frameworks for improving physical, cognitive, social, and / or behavioral skills and / or enhance such training through any of the various systems described herein, including confidence training systems, emotion training systems, professional communication systems, adaptive language competence systems, social skills enhancement systems, therapeutic intervention systems, role-based interactive platforms, and / or behavioral therapy support systems. Furthermore, System 100 can process natural language input using at least one generative model to create directed process graphs containing multiple actions and conditional logic. System 100 can be used to generate adaptive digital therapy plans for mental health interventions, cognitive behavioral therapy modules, personalized rehabilitation programs, guided exposure therapy sequences, digital coaching systems, automated user progress tracking systems, and / or any adaptive therapeutic response framework for real-time and / or near-real-time intervention adjustments.

[0060] Typically, a digital therapy workflow may include operations performed via system 100. A digital therapy workflow may include any one or more of an interface phase, a modeling phase, a content phase, a framework phase, and / or an output phase. Each phase of the digital therapy workflow includes one or more components of system 100 that perform the functions described herein. In some embodiments, one or more phases may be performed during the training phase of an AI model. Furthermore, one or more phases may be performed during the inference phase using an AI model (e.g., model 108).

[0061] System 100 (e.g., implementing a digital therapy process) may receive natural language input (e.g., input 102) to retrieve content from a database or produce content for a digital therapy application to address one or more conditions (e.g., migraine). System 100 implementing the digital therapy process may apply the natural language input and one or more conditions as input 102 to at least one generative model to generate a directed process graph for the conditions. The directed process graph includes multiple actions and conditional logic, according to which at least one of the multiple actions is provided for presentation. In some embodiments, implementing the digital therapy process may include system 100 identifying (e.g., via a database interface) at least one data object (e.g., a therapy component) corresponding to at least one of the multiple actions included in the directed process graph.

[0062] Furthermore, implementing a digital therapy workflow may include system 100 generating a structured execution framework using at least one data object and a directed process graph. The generated structured execution framework can identify multiple content items, each identifying at least one action among multiple actions and / or corresponding conditional logic, and providing at least one action among the multiple actions for presentation based on the conditional logic. In some embodiments, implementing a digital therapy workflow may include providing a structured execution framework. Therefore, by applying generative models to construct a structured execution framework for dynamic treatment plans, digital therapy workflows can provide technological improvements over previous therapy systems that relied on predefined, static treatment plan generation.

[0063] In some embodiments, the interface phase may be a phase in a digital therapy workflow where system 100 may receive natural language input and facilitate interaction with a large language model (LLM). System 100 may include at least one interface system 104. Interface system 104 may receive natural language input, structured prompts, specific therapy queries, user preferences, user interaction data, and / or any data related to therapy personalization or other user personalization to retrieve content from a database or generate content for a digital therapy application to address the condition. Interface system 104 may receive natural language input (e.g., input 102) to generate a treatment plan for a digital therapy application to address the condition. During the interface phase, interface system 104 may capture, acquire, and / or format free text input to match it with predefined LLM parameters (e.g., tokenization rules, response length limits, semantic weights, context preservation strategies, and / or any custom prompting engineering techniques).

[0064] In some embodiments, the interface system 104 may receive and / or otherwise interact with the administrator and / or user device by establishing a secure communication interface for data exchange. Free text may be queries, responses, comments, and / or contextual input provided by the administrator and / or user during the treatment plan generation process. Input 102 may be typed responses, speech-to-text conversion, structured feedback, specific conversational phrases, gesture-based input, biometric sensor data, and / or any other form of data. In some embodiments, input 102 may also include recordings (e.g., audio / video), user annotations, context-aware cues, emotion data, timestamps, interaction history, physiological response indicators, and / or any other data acquired before, during, or after the session. The interface system 104 may receive and / or otherwise acquire input 102 from the administrator. The interface system 104 may receive and / or otherwise acquire input 102 from the user device by polling the device during a treatment session or via a push notification system.

[0065] In some embodiments, interface system 104 may initiate a treatment plan generation process between system 100 and an administrator providing natural language input. Natural language input may include instructions identifying treatment goals, treatment duration, treatment activities, actions corresponding to treatment activities, criteria for completing treatment activities, and / or other instructions for generating a treatment plan for a digital therapy application (e.g., application 118). Input 102 may be high-level treatment details (e.g., goals, duration, exercises, parameters used for transitions between exercises, and / or other details). Natural language input may be prompts, such as “Create a 3-month treatment process for an individual with PTSD who experiences triggers in the afternoon, mid-week, and mid-week, including breathing exercises based on a predefined template, grounding exercises, interactive therapeutic interventions, medication administration, and a schedule for the first five days of treatment.” Furthermore, input may include other instructions, such as “The first 5 days of treatment will be based on our treatment template, where the treatment will transition on the next screen if the user answers yes.” In some embodiments, the administrator may provide input 102. Administrators may include operators, developers, clinicians, researchers, healthcare providers, regulatory experts, pharmaceutical personnel, and / or system inspectors who have authorized access to generate and / or improve treatment workflows and / or adjust the content presentation in digital treatment systems.

[0066] Furthermore, when the interface system 104 receives multiple data points and / or elements (e.g., multiple contexts, treatment processes with session goals, performance history, and / or ongoing metrics), the interface system 104 can prioritize processing the input based on predefined weights and / or hierarchical rules. The interface system 104 can resolve conflicting inputs and / or ambiguous data using decision algorithms and / or functions (e.g., machine learning models trained based on historical data, predefined priority rules, user-specific preferences, heuristic ranking methods, and / or any context-aware optimization techniques).

[0067] Furthermore, the digital therapy application (e.g., application 118) can be configured to provide input to the interface system 104. Application 118 can provide administrators and / or users with an intuitive and engaging interface to interact with system 100 to determine and / or generate personalized treatment goals, treatment processes, user preferences, cognitive abilities, and / or any relevant treatment plans. In some embodiments, application users can interact with system 100 through the digital therapy application to construct digital therapy processes.

[0068] The user device may run and / or otherwise implement digital therapy applications (e.g., mobile applications, web-based platforms, desktop software, and / or any compatible user interface). In some embodiments, the user device may be a smartphone, tablet, laptop, desktop computer, wearable device, and / or any internet-enabled device. The user device may be used as a medium for providing data to the treatment plan generation system. The user device may allow real-time and / or near real-time data exchange between the user and the interface system 104.

[0069] Natural language input may include input from a pre-generated treatment plan defined in a digital therapy application (e.g., application 118), which may communicate with interface system 104 to transmit natural input. Interface system 104 may transmit requests to interact with predefined data (e.g., context, session history, user preferences, and / or any behavioral patterns) or contextual information about the treatment plan or the session with the end user to address one or more conditions (e.g., nausea) through the digital therapy application.

[0070] In some embodiments, the modeling phase may be a phase in a digital therapy workflow where system 100 can apply input to generate a treatment plan (e.g., a directed process graph) for a digital therapy application. System 100 may include at least one therapy system 106. Therapy system 106 may apply input (e.g., natural language input) to at least one generative model (e.g., model 108) to generate a directed process graph for content system 110. The directed process graph may include multiple actions and conditional logic, according to which at least one of the multiple actions is provided for presentation. Therapy system 106 may process natural language input to enable the at least one generative model to generate a treatment plan. The treatment plan may include a list of treatment components and a schedule for presenting at least one treatment component from the list of treatment components that matches the treatment goals of the input.

[0071] At least one generative model can be a deep neural network, a language model, a large language model (LLM), a small language model (SLM), a visual language model (VLM), a multimodal language model (MMLM), a perceptual model, a tracking model, a fusion model, a transformer model, a diffusion model, an encoder-only model, a decoder-only model, an encoder-decoder model, a neural rendering field (NERF) model, a binarization model, and / or a transcription model, etc., to generate a directed process graph. Enabling model 108 to generate a directed process graph may include processing natural language input to extract multiple executable elements, associating multiple executable elements with multiple actions, and / or arranging multiple actions into a hierarchical structure. Model 108 can be a language model that processes natural language input to generate treatment plans for nicotine dependence, migraines for episodic and chronic migraines, multiple sclerosis for symptom management and relapse prevention, atopic dermatitis for controlling attacks and maintaining the skin barrier in the long term, obesity for lifestyle interventions and medication guidance, oncology for symptom management and psychological support, insomnia for sleep hygiene enhancement and behavioral therapy, acute coronary syndrome for post-event rehabilitation and secondary prevention, and / or any therapeutic treatment plan for a specific condition. Model 108 can extract multiple actionable elements, such as “identifying negative thought patterns,” “introducing cognitive restructuring exercises,” “performing breathing exercises,” “using grounding techniques,” “arranging emotion tracking,” “providing relapse prevention strategies,” “tracking medication adherence,” “assessing behavioral triggers,” “implementing progressive exposure therapy,” “arranging automated check-in,” and / or any structured treatment guidance components.

[0072] Model 108 can associate these executable elements with multiple actions, such as presenting psychoeducational lessons, prompting users to reflect and / or providing interactive exercises, offering personalized suggestions, generating progress reports, dynamically adjusting treatment pathways and / or any adaptive treatment interventions, while linking them to multiple conditional logics, such as completion of previous treatment activities, user-reported symptom changes, biometric feedback (e.g., heart rate variability), engagement duration, specific dates or times, user preferences, and / or other conditional logics. Model 108 can further arrange the multiple actions and multiple conditional logics into a hierarchical structure, organizing directed process diagrams to introduce psychoeducation, progress to interactive cognitive restructuring exercises, and / or adjust subsequent interventions based on user engagement and reported progress.

[0073] In some embodiments, multiple actions may correspond to therapeutic components, sessions, prompts, tasks, responses, and / or any interactive or passive therapeutic engagement methods. The system 114 is designed to generate data objects corresponding to diagnostic queries for providing treatment for smoking cessation, migraines, multiple sclerosis, atopic dermatitis, obesity, oncology, insomnia, acute coronary syndrome, and / or any health condition.

[0074] System 100 can construct data objects to include diagnostic screening processes or clinical scales or questionnaires, containing standardized question sequences, predefined scoring logic, and / or conditional branching rules. System 100 can generate data objects to determine symptom severity, determine whether further evaluation is needed, and / or dynamically adjust the progress graph by incorporating targeted interventions (e.g., behavioral activation exercises or guided cognitive restructuring sessions). Data objects may include logic for responding to the completion of a treatment component or providing a response to a diagnostic or clinical scale / questionnaire query, updating activity logs to include completion status, timestamp data, and / or data generated or acquired during treatment components or diagnostic or clinical scale / questionnaire queries. Data objects may include logic for linking record completion status, record participation duration, capturing biometric data (e.g., heart rate variability from wearable devices), and / or updating content based on this information.

[0075] In some embodiments, the treatment system 106 may maintain, execute, train, and / or update one or more machine learning models during the coding phase. In some embodiments, the machine learning models may include any type of generative and predictive machine learning model capable of processing natural language input to generate structured treatment workflows (e.g., deep learning models, transformer-based models, and / or sequence-to-sequence models) to produce directed process graphs representing treatment plans.

[0076] Machine learning models can be trained and / or updated to extract executable elements from a natural language description of a treatment plan, associate these elements with predefined treatment components, and / or generate other adaptive treatment plan generation tasks such as structured execution frameworks. Machine learning models can be or include transformer-based models (e.g., generative pre-trained transformer (GPT) models). In some embodiments, machine learning models can be or include reinforcement learning-based models.

[0077] Model 108 may include an encoder configured to execute a machine learning model to generate output. The encoder may receive data to provide as input to the machine learning model, which may include natural language input (e.g., from an operator). The encoder may process treatment goals and parameters provided by the operator, extract structured actions and conditional logic, and / or generate a directed process graph representing a sequence of treatment components corresponding to the conditional logic.

[0078] Furthermore, Model 108 can be a supervised learning model that processes labeled therapeutic session recordings to extract multiple executable elements for the procedure. Model 108 can be trained on an annotated dataset containing structured user-provider interactions to identify key therapeutic components, such as “introducing pain tolerance techniques,” “reinforcing mindfulness practice,” and “assessing progress in emotion regulation.” Model 108 can associate these executable elements with multiple actions, such as providing guided meditation, prompting self-assessment quizzes, generating personalized feedback, adjusting therapeutic suggestions, modifying session pacing, presenting interactive educational content, suggesting strategies, providing real-time adaptive interventions, modifying difficulty levels, and / or any behavior that promotes therapeutic engagement and / or disengagement, while linking them to multiple metrics, such as user engagement scores, assigned practice completion, and / or predefined behavioral milestones. Model 108 can further arrange multiple actions and multiple conditional logics into a hierarchical structure, such as introducing mindfulness techniques through structured directed process graphs, transitioning to pain tolerance practice based on engagement, and / or personalizing emotion regulation strategies based on user performance.

[0079] As used herein, a "directed process diagram" can refer to a structured representation of actions, conditional logic, and transitions in a treatment workflow, defining the relationships between treatment steps and their dependencies. A directed process diagram represents the sequence of execution of treatment activities, treatment plans, and / or user interactions, including any branching logic based on engagement or predefined criteria. A directed process diagram can specify initial modules that the user completes before entering treatment components, with conditional transitions based on user responses or evaluation results. Therefore, it should be understood that directed process diagrams structure treatment workflows to dynamically adapt based on user input, predefined logic, and / or evolving treatment requirements.

[0080] Furthermore, Model 108 can be an unsupervised learning model that can analyze unstructured user log data to extract multiple executable elements for situation-specific digital therapeutic plans. Model 108 can cluster user-generated text entries based on recurring themes such as “increased stress levels,” “negative self-talk,” and “avoidance behavior” without predefined labels. Model 108 can associate these executable elements with multiple therapeutic components such as suggesting relaxation exercises, providing cognitive restructuring techniques and / or guiding users to complete exposure tasks, while linking them to multiple conditional logics, such as detected emotional polarity, the frequency of stress-related keywords, and / or user engagement patterns. Model 108 can further arrange the multiple therapeutic components and multiple conditional logics into a hierarchical structure, generating a directed process graph that includes conditional logics that can be used to dynamically adjust the therapeutic process. Model 108 can arrange multiple therapeutic components and multiple conditional logics by identifying emerging anxiety patterns, personalized therapeutic exercises, and / or adjusting the therapeutic approach in real time based on constantly changing user input.

[0081] A directed process graph can correspond to a representation of a user interaction flow (e.g., a treatment plan), which includes treatment components and corresponding transitions, structured to define the treatment process over a period of time (e.g., daily sessions, weekly progress assessments, adaptive treatment plan checkpoints, phase-based treatment milestones, dynamically adjusted timelines, and / or any structured timeframe). A directed process graph can include a graph with nodes and edges for treatment plans tailored to various conditions. Nodes can represent individual treatment activities, assessment checkpoints, educational modules, user decision points, automated feedback triggers, and / or any treatment components. Edges can represent conditional transitions between activities, branching logic of adaptive treatment pathways, dependencies based on user progress, reinforcement loops for skill consolidation, and / or any structured relationships managing progress. The interaction flow over time can guide the user through progressive treatment steps (e.g., nodes). A directed process graph can include initial assessment phases pointing to treatment plan components (e.g., structured exposure exercises, coping strategy reinforcement, and / or periodic self-assessment check-ins and / or other modules). At least one (e.g., each) phase may include specific actions, such as guided breathing exercises, log prompts, virtual counseling sessions, mindfulness challenges, symptom tracking entries, role-playing simulations, psychoeducational courses, interactive scenario-based training, and / or any structured therapy component with conditional logic (e.g., timelines, completion criteria, user-reported results, adaptive difficulty adjustments, real-time and / or near-real-time performance monitoring, automated intervention escalation, personalized content recommendations, and / or any dynamic therapy adaptation logic) that determines progress based on user response and engagement, and other relevant examples.

[0082] In some embodiments, the treatment plan may include a timetable that identifies multiple conditional steps corresponding to at least one treatment component in a list of treatment components of the treatment process. The directed process graph may include a hierarchical arrangement of components representing multiple actions, and model 108 generates a hierarchical arrangement based on multiple temporal parameters that identify the order and duration corresponding to at least one of the multiple actions. Furthermore, the multiple conditional steps may correspond to conditional logic to be satisfied. Model 108 may generate a treatment plan for post-traumatic stress disorder (PTSD) that includes a timetable with conditional steps guiding the user through exposure therapy. Model 108 may structure the treatment process to include an initial educational phase followed by progressive exposure exercises. The treatment process may include nodes containing data about the initial educational phase and edges pointing to progressive exposure exercises to resolve the initial phase. The multiple conditional steps may correspond to conditional logic such as user-reported anxiety levels, completion of assigned exercises, and / or participation in coping strategies. If the user successfully completes a guided breathing session and reports reduced anxiety, model 108 may proceed to the next exposure task. Otherwise, the system may adjust the difficulty or provide alternative coping mechanisms before proceeding.

[0083] In some embodiments, the treatment plan's schedule may include multiple conditional logics that correspond to at least one treatment component and initiate at least one subsequent treatment component. The at least one subsequent treatment component may be provided based on the completion or failure of at least one previous treatment component. If the user successfully completes a mindfulness meditation session (e.g., maintaining focus for a specified duration), the graph may provide logic where scheduling logic for the plan can instruct the scheduling of more advanced mindfulness practices as subsequent treatment components. The graph may include nodes for treatment components, with edges pointing to future activities if the user completes the at least one previous treatment component, and edges pointing to different future activities if the user fails to complete the at least one previous treatment component.

[0084] This graph can provide conditional logic for whether a user successfully completes a migraine session (e.g., reporting a reduction in migraine frequency and severity within a predefined time period), where conditional logic for the treatment plan can instruct the entry into advanced pain management strategies and / or trigger subsequent assessments to monitor long-term effectiveness. The graph can include nodes for migraine sessions with edges pointing to actions used for migraine management.

[0085] This graph can provide conditional logic for whether a user successfully completes a smoking cessation session (e.g., maintaining nicotine abstinence for a predefined duration), where the conditional logic for the treatment plan can instruct reinforcement of relapse prevention techniques and / or the introduction of long-term behavioral maintenance strategies. The graph can include nodes for smoking cessation sessions with edges pointing to actions used for nicotine treatment.

[0086] This graph can provide conditional logic for whether a user successfully completes an obesity treatment session (e.g., achieving target levels of dietary adherence and physical activity participation). The conditional logic for the treatment plan can instruct the initiation of metabolic health monitoring and / or the introduction of new goal-setting modules for ongoing weight management. The graph can include nodes for the obesity treatment session with edges pointing to actions related to physical activity.

[0087] This graph can provide conditional logic for whether a user successfully completes an oncology treatment session (e.g., demonstrating adherence to symptom management techniques and reporting improvements in quality of life indicators), where conditional logic for the treatment plan can instruct transition to a survival care plan and / or inclusion of ongoing psychosocial support. The graph can include nodes for the oncology treatment session with edges pointing to actions used for symptom management.

[0088] In some embodiments, if a user fails to complete a session (e.g., reports difficulty maintaining attention) and / or fails to participate in assigned therapeutic exercises within a predefined time period, the knowledge graph may include conditional logic to repeat the session using additional guidance or support tools (e.g., shorter duration or increased guidance prompts). Furthermore, the at least one subsequent therapeutic component may be provided based on the completion, partial completion, and / or non-compliance of predefined criteria for the at least one therapeutic component. For example, if a user fails to meet predefined criteria for a daily reflection log (e.g., misses two consecutive entries), the knowledge graph may include conditional logic to introduce simplified log prompts or alternative reflection exercises to re-engage the user. If the user meets the criteria, the knowledge graph may direct a therapeutic plan to a node to integrate these reflections into cognitive restructuring exercises, thereby advancing the therapeutic process. Nodes in a directed process graph may include edges corresponding to transitions between therapeutic activities based on predefined criteria of completion, partial completion, and / or non-compliance. Each edge may represent a conditional path, where the next node in the therapeutic sequence is determined by real-time user engagement data and / or healthcare provider (HCP) recommendations. For example, if a user successfully completes a mindfulness session, an edge from the current node can point to a more advanced relaxation technique module. If the user leaves or fails to reach a predefined threshold (e.g., quits early or reports persistent stress), alternative edges can redirect the process graph to remedial techniques, such as guided videos or guided support sessions. Edges in the directed process graph can be combined with weighted probabilistics or reinforcement learning mechanisms to improve transitions over time.

[0089] As used herein, “conditional logic” can refer to rules that control transitions between treatment components based on user responses (including lack of response), behavioral patterns detected by the system, and / or other predefined criteria (e.g., the passage of time). Conditional logic can determine whether a user advances, repeats, and / or modifies a treatment component based on engagement, assessment scores, and / or other predefined thresholds. For example, if a user completes a breathing exercise with low engagement, the system can schedule an alternative relaxation technique before proceeding to the next stage. Therefore, it should be understood that conditional logic facilitates personalized treatment pathways (e.g., lateralization) by dynamically adjusting treatment steps based on real-time and / or near-real-time user interaction data and / or HCP recommendations.

[0090] As used herein, a “component” of a treatment plan can refer to a module, activity, intervention, and / or a group thereof. A component may consist of one or more therapeutic activities designed to address a specific aspect of treatment (e.g., symptom management, skills development, behavior modification, and / or other aspects). Components of a treatment plan may include sessions, activities, interventions, and / or other therapeutic steps related to the treatment plan. Therefore, it should be understood that a component represents one or more individual activities in the treatment plan workflow that are structured to support the progress of the overall treatment process.

[0091] As used herein, a “conversation” can refer to a discrete unit of therapeutic interaction within a treatment plan, structured to provide targeted intervention, assessment, and / or educational content. A session may consist of one or more therapeutic activities designed to be completed within a specific timeframe, often accompanied by progress tracking and engagement metrics. For example, a session may include guided mindfulness exercises, cognitive restructuring tasks, and / or emotion tracking questionnaires to be completed in a single session. Therefore, it should be understood that a session represents an individual activity within a structured treatment plan or a component of a structured treatment plan that facilitates progress toward therapeutic goals.

[0092] As used herein, “action” can refer to a specific therapeutic activity and / or system-driven process performed within a session to advance therapeutic goals. Actions can correspond to operational steps in a therapeutic workflow, such as displaying educational modules, triggering behavioral cues, and / or collecting user-reported data. Actions may include displaying psychoeducational videos, initiating guided breathing exercises, and / or dynamically adjusting content based on user engagement. Therefore, it should be understood that actions serve as functional components within a directed process diagram, facilitating the delivery of structured interventions and real-time and / or near-real-time adaptability.

[0093] As used in this article, "condition" can refer to a user's illness, disability, symptoms, or other condition. A condition can correspond to a specific underlying health condition that a treatment plan aims to address in a user.

[0094] In some embodiments, a directed process diagram may include multiple execution paths. These multiple execution paths may correspond to multiple therapeutic workflows within a therapeutic process. As an example of a digital therapeutic procedure for addressing insomnia, the multiple execution paths may correspond to different therapeutic workflows, such as a sleep hygiene education workflow, a cognitive restructuring workflow targeting negative sleep thoughts, and / or a stimulus control therapy workflow. At least one (e.g., each) execution path may be dynamically selected (e.g., via model 108) based on user-reported sleep patterns and participation in previous interventions. Furthermore, at least one (e.g., each) of the multiple therapeutic workflows may include a sequence of therapeutic components to address or manage the condition. A cognitive restructuring workflow for insomnia may include an initial sleep diary assessment, followed by guided exercises to challenge negative beliefs about sleep, personalized relaxation techniques, and / or pre-arranged sleep restriction interventions. Progress achieved through these actions can be determined by user-reported improvements in sleep efficiency and adherence to recommended interventions.

[0095] In some embodiments, a content stage may be a stage in a digital therapeutic process where system 100 can identify at least one data object corresponding to at least one of a plurality of actions contained in a directed process diagram. System 100 may include at least one content system 110. Content system 110 may interact with database 112 to apply actions contained in the directed process diagram. Content system 110 may retrieve from the database via an interface to database 112 a plurality of content items corresponding to at least one therapeutic component in a list of therapeutic components in a treatment plan.

[0096] Content system 110 can retrieve, for example, breathing exercise modules, progressive muscle relaxation sub-modules, behavior cessation modules, migraine modules, multiple sclerosis modules, atopic dermatitis modules, obesity modules, tumor modules, insomnia modules, acute coronary syndrome modules, and / or any therapeutic treatment plan modules corresponding to a structured treatment plan. Furthermore, content system 110 can retrieve and / or otherwise obtain sub-modules of various modules by accessing predefined content hierarchies, querying a database (e.g., database 112) for relevant treatment components, dynamically generating sub-modules based on treatment parameters, referencing previous user engagement data, and / or any adaptive content retrieval mechanism supporting modular treatment customization. In some embodiments, various templates can be obtained through the interface of database 112. Content system 110 can use directed process graphs to identify and retrieve relevant treatment content items for use by design system 114 to generate a structured execution framework.

[0097] Database 112 can be an internal or external database of system 100. In some embodiments, database 112 may correspond to an external repository (e.g., Figma, Adobe XD, GitHub, and / or any other repository) accessed via an interface (e.g., an application programming interface (API), webhook, SDK, GraphQL query, and / or any other interface) to retrieve existing modules, submodules, and / or configuration data for incorporation into the digital therapy application. Content system 110 may interact with digital experience platforms (e.g., Adobe XD, Sketch, Figma, InVision, Zeplin, and / or other platforms) to retrieve structured templates for therapy plans, including predefined UI components for psychoeducation, interactive exercises, and / or progress tracking. Content system 110 may query APIs to extract design tokens and layout structures. In some embodiments, database 112 may correspond to an internal repository included within system 100 for storing pre-configured therapy workflows, reusable therapy modules, and / or adaptive conditional logic rules within system 100 (e.g., application 118). Content system 110 can access internal repositories to retrieve and modify the structured execution framework without relying on external design tools.

[0098] In some embodiments, identifying at least one data object may include querying a database 112 that maintains the at least one data object, and / or querying database 112 may include using at least one API request to retrieve the at least one data object corresponding to at least one treatment action or configuration. Content system 110 may query database 112 to retrieve previously configured treatment plans. Content system 110 may send an API request to database 112 to retrieve a data object corresponding to a treatment action (e.g., a structured exposure therapy session) to be performed by the user. The retrieved data object may include session structure, pacing guide, user interface components for presenting treatment content, interactive elements for user engagement, transition logic for navigating between treatment steps, UI state parameters for dynamic content adaptation, accessibility configuration, and / or conditional branching logic based on user progress and / or HCP recommendations.

[0099] In some embodiments, at least one data object includes at least one content item from a plurality of content items and corresponding parameters. The data object may be a content item including corresponding parameters such as duration, difficulty level, and / or instructional text. The content system 110 may retrieve the data object and adjust its parameters based on treatment phases, user preferences, or other examples. Furthermore, the corresponding parameters may include at least one of content item relationships, temporal parameters, and metadata used to represent the at least one treatment action as a node or relationship in a directed process graph. Treatment actions, such as graded exposure modules for treatment, may be represented in the directed process graph as nodes connected to multiple branching paths. These paths may correspond to different exposure levels; for example, progress may depend on user-reported anxiety reduction. The system may use metadata, such as previous users' completion rates and effectiveness scores, to dynamically determine the optimal exposure level. Content may be personalized based on user preferences, such as the user indicating (or the system detecting) that summer scenes are more calming than winter scenes, or that urban landscapes are more appealing than natural landscapes, or that personas representing different demographic groups of the user are more inspiring when presenting psychoeducational lessons.

[0100] In some embodiments, the framework phase may be a phase in a digital therapy workflow where system 100 can generate a structured execution framework. System 100 may include at least one design system 114. Design system 114 may include any one or more artificial intelligence models (e.g., machine learning models, supervised models, language models, LLM, SLM, VLM, MMLM, neural network models, deep neural network models), rules, heuristics, algorithms, functions, and / or various combinations thereof, to perform operations including receiving, applying, recognizing, causing, generating, managing, and / or providing the structured execution framework. Design system 114 may include neural networks to generate the structured execution framework using at least one data object and a directed process graph.

[0101] The design system 114 can generate a structured execution framework to organize and / or otherwise arrange multiple content items that identify at least one of multiple actions. Furthermore, the design system 114 can generate a structured execution framework to arrange corresponding conditional logic, based on which at least one of multiple treatment components is provided for presentation. The design system 114 can use multiple content items and / or schedules to arrange a list of treatment components of a treatment plan into a structured data package for presentation in a digital therapy application.

[0102] In some embodiments, the design system 114 may maintain, execute, train, and / or update one or more machine learning models. In some embodiments, the machine learning models may include any type of deep learning-based, sequence-to-sequence, and / or rule-based machine learning model capable of generating structured execution frameworks (e.g., pseudocode, skeleton code, workflow diagrams, execution state diagrams, hierarchical action maps, declarative profiles, API interaction patterns, and / or any structured representation that facilitates the execution of treatment plans) to structure adaptive digital therapy applications based on natural language treatment goals and parameters. The design system 114 may implement one or more machine learning models to improve conditional logic for treatment progress and / or generate executable workflow representations, as well as other tasks such as improving content personalization, adjusting treatment workflows based on user engagement, and / or predicting optimal intervention sequences to deploy the structured execution framework.

[0103] As used herein, "structured execution framework" and / or "structured data package" can refer to a representation of organized data, relationships, and / or execution parameters that defines how treatment plans are structured, processed, and implemented in a digital therapy application. A structured execution framework provides a formalized structure for encoding actions, conditions, and / or dependencies, thereby allowing for processing, execution sequencing, and / or integration with external systems (e.g., application 118). A structured execution framework can specify sequences of treatment activities, corresponding transitions, timing constraints, and / or conditional logic for adjusting treatment delivery. Therefore, it should be understood that a structured execution framework supports the dynamic and structured execution of treatment workflows, thereby facilitating modularity, adaptability, and / or integration across a variety of digital therapy applications.

[0104] The machine learning model may be or include a transformer-based model (e.g., a generative pre-trained transformer (GPT) model). In some embodiments, the machine learning model may be or include a reinforcement learning-based model. System 114 is designed to execute the machine learning model to generate output. System 114 may receive data as input to the machine learning model. In some embodiments, system 100 may configure (e.g., train, update, fine-tune, apply transfer learning) the model of system 114 by modifying or updating one or more parameters (e.g., weights and / or biases) of the individual nodes of the model in response to evaluation of the estimated output of the model.

[0105] In the framework phase, the design system 114 can process treatment plans defined by natural language and generate a structured execution framework by extracting treatment components from a database. The design system 114 can use directed flowcharts to associate exercises, activities, and / or tasks with corresponding conditional logic (such as user engagement metrics and predefined success criteria).

[0106] In some embodiments, the structured execution framework and / or structured data packets correspond to a representation of a treatment plan generated from a directed process graph and / or data objects. The structured execution framework may correspond to multiple treatment components, conditional logic, and / or timing attributes, organized to support the generation of executable instructions for a structured digital therapy application. The design system 114 can convert the structured execution framework into data packets by encoding multiple actions, conditional logic, state transitions, user interaction parameters, execution dependencies, and / or timing attributes into JSON, XML, YAML, protobuf, and / or any configuration file. The structured data packets can serve as an intermediate representation for the digital therapy application to interpret and present treatment sessions. The JSON file can define sequences of treatment components, conditional logic for transitions between activities, and / or timing constraints, allowing the system to dynamically generate user-facing interfaces. The configuration file can define the UI layout, specify conditional logic for presenting dynamic content updates, establish rules for adaptive content presentation, encode accessibility settings, and / or facilitate integration with external data sources or APIs.

[0107] Design system 114 can generate a structured execution framework by processing directed process graphs and acquired data objects to define an executable treatment workflow. Design system 114 can traverse the directed process graph to identify nodes representing treatment components, extract relevant conditional logic, and link these elements to relevant data objects. Design system 114 can arrange actions, logic, elements, and / or other node objects into structured data formats (e.g., JSON, XML, YAML, and / or other data formats). For example, if the directed process graph includes a user assessment node followed by a relaxation module, design system 114 can retrieve a predefined content template (e.g., a bootstrapping module from a database) and encode its execution parameters (e.g., duration, user progress criteria, and transition conditional logic) into the structured execution framework.

[0108] The design system 114 can use a hierarchical structured approach to organize the execution framework based on the relationships between directed process graph components and their corresponding data objects. The design system 114 can analyze execution dependencies and conditional logic within the directed process graph to determine the sequencing of the treatment workflow. Then, the design system 114 can assign timing parameters, state transitions, and adaptive logic rules to generate an execution model that supports adjustments. For example, if the treatment process includes progressive exposure treatment tasks, the design system 114 can retrieve intensity-adjusting task parameters from structured data objects, encode the progression logic into the execution framework, and specify dynamically updated UI configurations based on user behavior.

[0109] In some embodiments, the structured execution framework and / or structured data packets correspond to off-the-shelf formats and / or schemes (e.g., machine-readable, API-compatible, platform-independent, extensible) that are structured to present a treatment process that defines how at least one (e.g., each) content item is presented and specifies the order and timing of the content items. The structured execution framework can be formatted as a standardized schema, such as a state machine model, thereby defining how content items are presented over time. The structured execution framework can specify whether a treatment component is presented before or after another treatment component based on real-time and / or near-real-time user performance, administrator control, and / or HCP recommendations. Design system 114 can generate this format as a structured object containing execution rules, metadata, and / or user progress paths.

[0110] In some embodiments, the design system 114 may use a structured execution framework to generate multiple instructions configured to cause a digital therapy application to display content comprising at least one of multiple content items according to a directed process graph. The design system 114 can generate executable instructions for the digital therapy application, for example, transforming the structured execution framework into an API-driven workflow for dynamically loading and displaying treatment content. The design system 114 can formulate treatment plans, generate corresponding JavaScript and Python instructions, and / or provide an interactive user experience through an adaptive digital interface. The system can compile the structured execution framework into executable functions that trigger content modules and dynamically adjust them based on user progress, administrator control, and / or HCP recommendations.

[0111] In some embodiments, the structured execution framework corresponds to a mapping of multiple content items, corresponding conditional logic (e.g., schedules), multiple temporal parameters, and / or multiple execution dependencies in a directed process graph. As an example of a smoking cessation treatment plan, design system 114 can generate a structured execution framework that includes a mapping of content items (e.g., motivational coaching videos, craving management exercises, behavioral alternative cues), corresponding conditional logic (e.g., progress-related branching paths based on reported cravings), temporal parameters (e.g., daily self-assessment check-ins, weekly behavioral goal-setting sessions), and / or execution dependencies (e.g., completing nicotine replacement therapy education before introducing behavioral coping strategies) for coherent and personalized treatment workflows in a digital therapy application.

[0112] In some embodiments, the design system 114 may use a simulator to simulate the performance of multiple actions based on processing at least a portion of a plurality of instructions according to a structured execution framework. The plurality of instructions may include at least one operation for performing at least one of a plurality of actions corresponding to at least one timing parameter or conditional logic parameter. As an example, the design system 114 may use a reinforcement learning-based model and simulator to simulate the execution of a PTSD treatment process. System 100 may process the structured execution framework in a sandbox environment to predict how a user will interact with the treatment content. The simulator may analyze engagement trends, test different conditional approaches, and / or improve the execution framework before deploying it to a real-time digital therapy application.

[0113] In some embodiments, the design system 114 may retrieve skeleton code from a database or generate skeleton code using a structured execution framework and / or a directed process graph. This skeleton code includes at least one program construct corresponding to multiple actions and corresponding conditional logic. The design system 114 may retrieve skeleton code from a database or generate skeleton code from a directed process graph by extracting a sequence of treatment actions to be performed by the user, mapping them to predefined functional treatment components, and structuring them into executable program constructs. The design system 114 may analyze the nodes and edges of the directed process graph to identify dependencies between treatment components, state transitions, and conditional logic, and then convert this information into a structured code representation. As an example, if the directed process graph specifies a treatment sequence with conditional branches based on user-reported data, the design system 114 may generate JavaScript functions that define the interactive session logic, such as event handlers for user responses, API calls for progress tracking, and UI state updates that dynamically adjust the treatment workflow.

[0114] Furthermore, the design system 114 can compile skeleton code retrieved from at least one database. The design system 114 can retrieve skeleton code from a backend database containing predefined UI components for treatment modules. The system can structure the skeleton code to match it with a directed process diagram, thereby generating a base codebase for a digital therapy application (e.g., application 118). The design system 114 can retrieve template-based UI structures of modules from the database, extract relevant functional components (e.g., log input fields, guided breathing exercise templates), and integrate them with logic derived from the directed process diagram. The system can compile the retrieved skeleton code using logic dynamically generated from a structured execution framework. If the directed process diagram specifies conditional logic for transitions between treatment modules, the design system 114 can embed corresponding state management logic into the skeleton code.

[0115] In some embodiments, using a structured execution framework, system 114 can generate HTML templates that define the order of treatment sessions and interactive prompts. Using directed flowcharts, system 114 can dynamically structure JavaScript functions that adjust content delivery based on user behavior, administrator controls, and / or HCP recommendations. Furthermore, system 114 can retrieve or generate skeleton code from a database, including code for handling user interactions, managing state transitions, dynamically rendering treatment components, integrating API calls for progress tracking, and / or other JavaScript implementations within the digital therapy application, or CSS code for styling treatment session UI elements, providing responsive design across devices, customizing animations for interactive treatment components, defining accessibility features to enhance usability, and / or other implementations for developing or updating the digital therapy application.

[0116] In some embodiments, the design system 114 may provide structured data packages (e.g., content, treatment plans) and / or skeleton code to an administrator for review. In some embodiments, the design system 114 may receive at least one update to the treatment plan or skeleton code from the administrator before providing the structured execution framework to a digital therapy application (e.g., application 118). For example, an administrator reviewing an AI-generated exposure therapy sequence for phobia treatment may adjust the intensity level of the initial exposure task before the design system 114 finalizes the structured execution framework for deployment in the digital therapy application. In some embodiments, the design system 114 may receive at least one update to the treatment plan or skeleton code from the administrator before using the skeleton code in the digital therapy application. Systems engineers may update the skeleton code to integrate additional progress tracking APIs before deploying the code to the digital therapy application.

[0117] In some embodiments, the output stage may be a stage in a digital therapy workflow where system 100 can provide a structured execution framework. System 100 may include at least one data packet 116. In some embodiments, data packet 116 may be a structured execution framework, a structured data packet, skeleton code, compiled workflow instructions, UI component mappings, and / or any other data packet / object discussed herein that design system 114 can provide (e.g., provided to application 118). Design system 114 may generate and send output to guide application 118. Design system 114 may provide data packet 116 to a digital therapy application that parses the structured execution framework to dynamically construct executable code to provide interactive modules. Application 118 may use the packet to configure session flow, adjust content presentation based on conditional logic, and / or generate UI components for therapy components. Application 118 may process the structured data packet to determine the order of therapy components, adjust interactive elements based on predefined conditional logic, and / or organize content presentation according to therapy parameters. For example, if the data packet specifies a progressive relaxation procedure, the application 118 can dynamically structure the session by generating introductory breathing exercises, followed by a guided body scan and adjusting subsequent activities based on user engagement.

[0118] Application 118 may be installed and / or otherwise maintained by a user device (e.g., mobile phone, tablet, wearable device, desktop computer, smart speaker, and / or any compatible hardware). Application 118 may receive data packets 116 for parsing a structured execution framework, extracting treatment plan parameters, and configuring application behavior accordingly. Application 118 may interpret the structured data packets in a runtime environment to organize and present the treatment workflow based on system-generated execution dependencies. As an example, data packet 116 may include a JSON-encoded treatment structure definition specifying a multi-step migraine relief procedure. Upon receiving the packet, application 118 may assemble a structured sequence of treatment components, adjusting UI elements and interactive components according to the encoded treatment plan. Data packet 116 may include media references and guidance prompts for guiding relaxation sessions. Application 118 may dynamically retrieve and present corresponding audio, video, and / or text content, structuring the user experience based on predefined arrangement parameters rather than executing predefined application logic.

[0119] Overall, System 100 addresses the technical challenge of providing effective and dynamic solutions to a large number of users by facilitating adaptive therapy workflows based on real-time and / or near-real-time engagement and structured execution frameworks to personalize digital therapeutic treatment plans. Furthermore, System 100 resolves the development delays in digital therapy applications caused by the manual work required to design, structure, and / or validate treatment plans. By leveraging APIs from digital experience platforms such as Adobe XD, Sketch, Figma, InVision, Zeplin, and / or other platforms to integrate proprietary databases, and / or generating structured execution frameworks from natural language input, System 100 significantly reduces the time and effort required to build and deploy functional digital therapy applications. Developers are not required to manually code at least one (e.g., each) treatment process, but can provide high-level natural language descriptions of treatment goals, and / or System 100 can generate structured execution frameworks, compile skeleton code, and / or integrate it with existing UI components, allowing for rapid iteration and deployment of large numbers of personalized digital therapy applications.

[0120] System 100 addresses the challenge of creating a generative AI-based platform that is individually tailored to a large number of users requiring therapy in the physical, cognitive, social, and / or behavioral domains. System 100 provides the application with a structured execution framework to dynamically adjust personalized treatment plans based on user progress and external data sources, such as HCP recommendations. For example, if a user reports high stress levels, the application can modify the execution path to prioritize relaxation exercises before cognitive restructuring interventions.

[0121] In some embodiments, system 100 can process natural language input to generate structured treatment plans, thereby overcoming significant barriers to describing user processes in digital therapy using NLP. Traditional systems often struggle to transform unstructured input from users, administrators, or healthcare providers into structured execution frameworks. System 100 can apply at least one generative model to extract key treatment components, associate them with predefined actions, and / or generate directed process graphs that structure user interactions over time. For example, if an operator provides a natural language description of the treatment process, system 100 can identify necessary treatment modules, perform logical transformations based on user progress constraints, and / or structure adaptive digital therapy workflows.

[0122] In some embodiments, system 100 can interact with a centralized database 112 to dynamically retrieve, modify, and / or deliver treatment content, thereby overcoming a major obstacle to centralization in the design of digital therapy applications. Traditional therapy systems often require manually structuring content, leading to inefficient and inconsistent treatment plan implementation. System 100 can integrate with external and internal repositories (e.g., Figma, proprietary databases) to systematically query, retrieve, and / or update UI components, treatment templates, and / or structured content assets. System 100 can retrieve structured module templates from digital experience platforms (e.g., Adobe XD, Sketch, Figma, InVision, Zeplin, and / or other platforms) to adapt its content based on an updated therapy framework.

[0123] In some embodiments, system 100 can use generative AI to translate natural language into a plan. In some embodiments, the plan can invoke a group of APIs that generate front-end JavaScript code (and / or any other tag-based code) and CSS code (and / or any other styled code), thereby providing a digital therapy application that can be further refined by practitioners. System 100 can use proprietary APIs to generate a large portion of the front-end code (e.g., UI). System 100 can add a database 112 to a UI building / digital experience platform (e.g., Adobe XD, Sketch, Figma, InVision, Zeplin, and / or other platforms). System 100 can iterate through the design structures generated by database 112 to combine one or more structures into the desired front-end code. System 100 can build groups of digital therapy components with improved accuracy by leveraging natural language input and proprietary APIs. System 100 can provide shortened timelines in application 118 feature creation, automated UI component assembly, dynamic layout generation, predictable code quality in front-end interactions, consistent UI / UX, and / or any framework-matched development workflow optimizations.

[0124] In some embodiments, natural language input can be used to generate treatment plans for digital therapy applications to address a variety of conditions, such as neurological disorders and mental health conditions, metabolic syndrome and cardiovascular disease, chronic pain conditions and rehabilitation programs, and / or any palliative health condition that uses structured interventions.

[0125] Systems and methods for generating structured execution frameworks in the content generation process

[0126] refer to Figure 2The example flowchart illustrates methods for retrieving content from a database or generating and providing content in a digital therapy workflow, according to some embodiments of this disclosure. It should be understood that the embodiments described herein and other embodiments are examples. Alternative configurations, elements (e.g., machines, interfaces, functions, sequences, and / or groupings) and / or omissions are possible. Many elements are functional and can be implemented as discrete or distributed components, combined with other elements, and / or located in various configurations. Functions can be performed using hardware, firmware, and / or software (e.g., a processor executing instructions stored in memory). Systems, methods, and / or processes can be used with... Figure 4 It has similar components and functions to the server system 400 and the client computing system 414.

[0127] exist Figure 2 In this method, each block of method 200 represents a computational process that can be executed using hardware, firmware, and / or software (e.g., a processor executing instructions stored in memory). The method can also be implemented as computer-readable instructions on a storage medium, provided as a standalone application, service, microservice via API, and / or plugin. Method 200 Reference Figure 1 The system described herein may also be implemented through any other system or combination of systems described herein.

[0128] Figure 2 This is a flowchart illustrating a method 200 for receiving, applying, identifying, initiating, generating, managing, and / or providing operations according to some embodiments of this disclosure. Various operations of method 200 may relate to improving the personalization and generation of digital therapeutic applications. Existing systems typically rely on and / or use predefined content (e.g., treatment plans), which can lead to limited adaptability and reduced treatment effectiveness. These technical problems arise when these prior systems fail to provide dynamic and adequate treatment plans for use by the application, resulting in reduced engagement and invalid outcomes. Figure 2 Method 200 addresses these technical challenges by implementing personalized AI-driven content generation for each individual user, thereby improving the efficacy of the application and downstream treatments.

[0129] Method 200, at block 210, includes receiving natural language input to retrieve or generate content for a digital therapy application to address a condition. In some embodiments, processing circuitry may receive natural language input to generate a treatment plan for the digital therapy application to address the condition. In some embodiments, the natural language input may include instructions identifying the duration of treatment (e.g., a week, 30 days, a three-month intervention, a maintenance plan, and / or other durations). In some embodiments, the natural language input may include instructions identifying at least one treatment component to be performed by the user (e.g., guided meditation, exposure therapy, cognitive restructuring exercises, and / or other behavioral activation tasks). In some embodiments, the natural language input may include instructions identifying at least one action corresponding to at least one treatment component (e.g., displaying a breathing exercise module, prompting log reflection, adjusting the task difficulty level, scheduling follow-up sessions, and / or other actions). In some embodiments, the natural language input may include instructions identifying at least one criterion for completing the at least one treatment component (e.g., successfully completing three consecutive sessions, self-reported stress reduction below a threshold, participation rate above 80%, adherence to the prescribed intervention schedule, and / or other criteria). Input may be high-level treatment details (e.g., goals, duration, exercises, transition parameters between exercises, and / or other details).

[0130] At block 220, method 200 includes applying natural language input and a condition and / or multiple conditions (e.g., migraine) as input to at least one generative model, causing the at least one generative model to generate a directed process graph for the condition. In some embodiments, the directed process graph may include multiple actions and conditional logic, according to which at least one of the multiple actions is provided for presentation. In some embodiments, one or more processing circuits may apply natural language input and a condition and / or multiple conditions as input to at least one generative model, causing the at least one generative model to generate a treatment plan for the condition. The treatment plan may include a list of treatment components and a schedule for presenting at least one treatment component in the list of treatment components. The treatment plan may include a schedule of multiple conditional steps that identify at least one treatment component in the list of treatment components corresponding to the treatment process. Furthermore, each of the multiple conditional steps may correspond to conditional logic to be satisfied.

[0131] In some embodiments, the treatment schedule may include multiple conditional logics corresponding to at least one treatment component and initiating at least one subsequent treatment component. The at least one subsequent treatment component may be provided based at least on the completion or failure of at least one previous treatment component, or the completion or non-completion of predefined criteria for the at least one previous treatment component. In some embodiments, the directed process graph may include a hierarchical arrangement of treatment components representing multiple actions to be performed by the user, the hierarchical arrangement being based on multiple timing parameters identifying the order and duration corresponding to at least one of the multiple actions. In some embodiments, the directed process graph may include multiple execution paths corresponding to multiple treatment workflows of a treatment process. Furthermore, at least one (e.g., each) treatment workflow may include a sequence of treatment actions to be performed by the user to address or manage a condition (e.g., obesity).

[0132] In some embodiments, the at least one generative model may be a deep learning model, a supervised learning model, and / or an unsupervised learning model. Furthermore, generating a directed process graph from the at least one generative model may include processing natural language input to extract multiple executable elements, associating the multiple executable elements with multiple actions and multiple conditional logics, and / or arranging the multiple actions and multiple conditional logics into a hierarchical structure. The conditional logic of the directed process graph may correspond to at least one treatment component and initiate at least one subsequent treatment component. The at least one subsequent treatment component is provided at least based on the completion or failure of at least one previous treatment component or the completion or non-completion of a predefined criterion for the at least one previous treatment component.

[0133] At block 230, method 200 can identify at least one data object. The processing circuitry can identify the data object through a database interface. In some embodiments, the data object may correspond to at least one of a plurality of actions included in a directed process diagram. In some embodiments, the processing circuitry can retrieve a plurality of content items from a database through a database interface. The plurality of content items may correspond to at least one treatment component in a list of treatment components in a treatment plan. In some embodiments, identifying the at least one data object includes querying a database that maintains the at least one data object. Querying the database may include using at least one API request to retrieve the at least one data object corresponding to at least one treatment action or configuration.

[0134] At block 240, method 200 includes generating a structured execution framework. The processing circuitry can generate the structured execution framework using at least one data object and a directed process graph. The structured execution framework can identify multiple content items, each content item identifying at least one action among multiple actions. Furthermore, the structured execution framework can identify corresponding conditional logic, based on which at least one action among multiple actions is provided for presentation. In some embodiments, the processing circuitry can use multiple content items and a schedule to arrange a list of treatment components of a treatment plan into a structured data package for presentation in a digital therapy application.

[0135] In some embodiments, the processing circuitry may use structured data packets to generate multiple instructions configured to cause a digital therapy application to display at least one of a plurality of content items according to the arrangement of a list of treatment components in a treatment plan. The structured data packets may correspond to a mapping of multiple content items, a timeline, and / or multiple execution dependencies. Furthermore, the processing circuitry may use a structured execution framework to generate multiple instructions configured to cause the digital therapy application to display content including at least one of a plurality of content items according to a directed process graph. The structured execution framework corresponds to a mapping of multiple content items, corresponding conditional logic, multiple timing parameters, and / or multiple execution dependencies in a directed process graph.

[0136] In some embodiments, the processing circuitry may use a simulator to simulate the performance of multiple actions based on processing at least a portion of a plurality of instructions according to a structured execution framework. The plurality of instructions may include at least one operation for performing at least one of a plurality of actions corresponding to at least one timing parameter. In some embodiments, the processing circuitry may use a structured execution framework and a directed process graph to compile skeleton code retrieved or generated from at least one database. The skeleton code may include at least one program construct corresponding to the plurality of actions and corresponding conditional logic. In some embodiments, the processing circuitry may use a structured data package to compile skeleton code retrieved or generated from at least one database, the skeleton code including at least one program construct corresponding to a list of treatment components and a schedule of a treatment plan.

[0137] In some embodiments, the processing circuitry may provide a treatment plan or skeleton code to an administrator for review. Furthermore, the processing circuitry may receive at least one update to the treatment plan or skeleton code before providing structured data packets or using the skeleton code in a digital therapy application. The processing circuitry may receive updates from the administrator, HCP, or other external sources.

[0138] In some embodiments, at least one of the plurality of actions corresponds to a treatment component or a response to a diagnostic or clinical scale / questionnaire query. In some embodiments, the processing circuitry may update the activity log to include completion status, timestamp data, and / or data generated or acquired during the treatment component or diagnostic or clinical scale / questionnaire query in response to the completion of the treatment component or the provision of a response to the diagnostic or clinical scale / questionnaire query.

[0139] At block 250, method 200 includes providing a structured execution framework. Processing circuitry can provide structured data packets. Processing circuitry can provide the structured execution framework to an administrator for review. Furthermore, processing circuitry can receive updates to the structured execution framework. Processing circuitry can provide the structured execution framework or skeleton code to a digital therapy application for use.

[0140] Now for reference Figure 3A This is an example interface, according to some embodiments of the present disclosure, for providing a way to receive natural language input to retrieve content from a database or generate content for a digital therapy application to address a condition (e.g., nausea). System 100 interacts with device 302 from an administrator, digital therapy application user, and / or other sources via the interface to generate a structured execution framework. System 100 can provide an interface 304 for model interaction 310. In this scenario, a model (e.g., model 108 and / or the model designing system 114) is configured to process natural language input and dynamically generate a structured execution framework. Model interaction 310 involves instructions input by the administrator for model 108 to create a treatment process. The administrator begins “creating a 3-month treatment process for an individual suffering from PTSD who experiences triggers in the afternoon, the middle of the workday, and several days in the middle of the workweek. This process includes breathing exercises based on predefined templates, grounding exercises, interactive therapeutic interventions, medication administration, and a schedule for the first five days of treatment.” Model 108 and / or the model designing system 114 generates a chatbot response: “This is a structured execution framework,” including an additional structured execution framework. In the background, the processing circuitry of System 100 processes the input through Model 108 to generate a directed process graph that includes multiple actions and conditional logic. The processing circuitry can use the directed process graph to generate a structured execution framework. The administrator responds: “Modify the treatment plan to introduce new exercises after successfully completing 80% of the previous exercises.” Metrics such as input clarity and scenario-specific goals can be tracked to personalize interactions and facilitate dynamic framework generation and adaptive responses.

[0141] A structured execution framework can be a structured representation of a treatment process generated from natural language input, comprising multiple actions, conditional logic, and / or temporal attributes organized to support the generation of executable instructions for a structured digital therapy application. A structured execution framework can be a JavaScript configuration file containing structured JSON data defining the treatment session flow, conditional triggers, and / or interactive components. The JavaScript file can instantiate dynamic UI components, register event listeners to track user interactions, and / or execute conditional logic that determines content ordering based on user progress. For example, a treatment plan can adjust its execution path in real-time and / or near real-time by evaluating engagement metrics and correspondingly transitioning the user to the appropriate treatment module. A structured execution framework can be a pattern-driven representation that uses the API of design tools to query all existing modules and submodules in the digital experience platform space. The system can retrieve predefined UI components, such as treatment session layouts, button interactions, and / or progress indicators, and integrate them into the structured execution framework. Databases and / or orchestration tools can compile these UI elements into the final digital therapy application. The compiled UI and execution logic can be deployed in a simulator for testing, allowing administrators to visualize content transformations, validate user interactions, and / or improve the overall therapy experience before deployment. The structured execution framework can include a drag-and-drop file containing a CSS code shell to be imported into the front-end development environment, where predefined style components are automatically mapped to corresponding UI elements in the digital therapy application. When developers drag and drop the file into a web-based editor or IDE, the system can parse the CSS structure, apply class-based style rules, and / or generate responsive layouts that dynamically adjust to different screen sizes and accessibility settings.

[0142] Now for reference Figure 3B This is an example interface used to provide a framework according to some embodiments of this disclosure. System 100 interacts with device 302 from administrators, digital therapy application users, and / or other sources via the interface to generate a structured execution framework. System 100 can provide an interface 304 for model interaction 310. Model 108 can generate prompts and display messages to provide additional information to the administrator for creating the treatment process. Model 108 can generate prompts and display additional messages in response to administrator input. Additional interactions between model 108 and the administrator can provide additional input to the model to generate a treatment plan.

[0143] In this scenario, Model 108 is configured to process natural language input and dynamically generate a structured execution framework. Model interaction 310 involves instructions input by the administrator for Model 108 to create a treatment process. The administrator begins, “Create a treatment process for an individual suffering from PTSD who experiences triggers in the afternoon, the middle of the workday, and several days in the middle of the workweek. This process includes a schedule of breathing exercises.” Model 108 generates a chatbot response: “Okay. How many days would you like the plan to last?” The administrator replies, “Please make it 5 days.” System 100 applies Model 108 and / or designs the model of System 114 to process this input and generates a response: “Okay. Do you have any other instructions?” The administrator continues, “Based on our template, where a support team member will contact the user if user engagement reaches 40%.” System 100 applies Model 108 and / or designs the model of System 114 to process this input and generates a response: “This is a structured execution framework,” including additional structured execution frameworks.

[0144] System 100 can apply model 108 and / or design the model of system 114 to generate a structured execution framework from other natural language prompts. As shown, system 100 can apply model 108 to generate prompts for administrator input. System 100 can apply model 108 to interact with the administrator to provide prompts and / or other prompts regarding duration, exercise type, treatment difficulty level, conditional logic branches, user progress tracking methods, adaptive intervention adjustments, etc. System 100 can apply model 108 to generate a process diagram based on the interaction. Furthermore, system 100, which generates additional prompts for administrator input, can allow for dynamic updates to the treatment plan's process diagram by incorporating newly provided details (e.g., updated HCP recommendations). If the administrator modifies the treatment duration, adds new treatment activities, and / or specifies additional conditional transitions, system 100 can regenerate the directed process diagram to reflect these changes in the continuously improving treatment workflow, where at least one (e.g., each) modification can be incorporated into the execution framework.

[0145] System 100 can generate feedback of conditional logic represented between process graph modules to further create directed graphs of user interaction flows over arbitrary time periods (e.g., days, weeks, months, years). In some embodiments, an administrator can examine the output of model 108 to design the final treatment procedure.

[0146] The various operations described in this article can be implemented on a computer system. Figure 4A simplified block diagram of a representative server system 400, user computer system 414, and / or network 426 that can be used to implement certain embodiments of this disclosure is shown. In various embodiments, server system 400 or similar systems may implement the services or servers or portions thereof described herein. System 100 described herein may be similar to server system 400. Server system 400 may have a modular design that incorporates multiple modules 402 (e.g., blades in a blade server embodiment); although two modules 402 are shown, any number of modules 402 may be provided. At least one (e.g., each) module 402 may include a processing unit 404 and local storage 406.

[0147] Processing unit 404 may include a single processor, which may have one or more cores, or multiple processors. In some embodiments, processing unit 404 may include a general-purpose main processor and one or more dedicated coprocessors, such as a graphics processor, a digital signal processor, etc. In some embodiments, some or all of processing unit 404 may be implemented using custom circuitry, such as an application-specific integrated circuit (ASIC) or a field-programmable gate array (FPGA). In some embodiments, such an integrated circuit executes instructions stored on the circuit itself. In other embodiments, processing unit 404 may execute instructions stored in local memory 406. Processors of any type and in any combination may be included in processing unit 404.

[0148] Local storage 406 may include volatile storage media (e.g., DRAM, SRAM, SDRAM, etc.) or non-volatile storage media (e.g., disk or optical disk, flash memory, etc.). The storage media incorporated in local storage 406 may be fixed, removable, or upgradeable as needed. Local storage 406 may be physically or logically divided into various sub-units, such as system memory, read-only memory (ROM), and / or permanent storage devices. System memory may be a read-write memory device or a volatile read-write memory, such as dynamic random access memory. System memory may store some or all of the instructions and data required by processing unit 404 during operation. ROM may store static data and instructions required by processing unit 404. Permanent storage devices may be non-volatile read-write memory devices that can store instructions and data even when module 402 is powered off. The term "storage media" as used herein includes any medium that can store data indefinitely (subject to overlay, electrical interference, power outages, etc.), but excludes carrier waves and transient electronic signals propagated via wireless or wired connections.

[0149] In some embodiments, local storage 406 may store one or more software programs executed by processing unit 404, such as operating systems and / or programs that implement various server functions (such as the functions of system 100 or any other system described herein, and / or any other server associated with system 130 or any other system described herein).

[0150] "Software" generally refers to a sequence of instructions that, when executed by processing unit 404, cause server system 400 (or a portion thereof) to perform various operations, thereby defining one or more specific machine embodiments for executing and implementing software program operations. Instructions may be stored as firmware residing in read-only memory and / or program code stored in non-volatile storage media, which can be read into volatile working memory for execution by processing unit 404. Software may be implemented as a single program or a collection of separate programs or program modules that interact as needed. From local storage 406 (or non-local storage described below), processing unit 404 may retrieve program instructions to be executed and data to be processed in order to perform the various operations described above.

[0151] In some server systems 400, multiple modules 402 can be interconnected via a bus or other interconnect 408 to form a local area network (LAN) that supports communication between the modules 402 and other components of the server system 400. The interconnect 408 can be implemented using various technologies, including server racks, hubs, routers, etc.

[0152] Wide area network (WAN) interface 410 can provide data communication capabilities between a LAN (e.g., via interconnection 408) and a network 426 (such as the Internet). Other technologies can be used to couple the server system to the network 426 for communication, including wired (e.g., Ethernet, IEEE 802.3 standard) or wireless technologies (e.g., Wi-Fi, IEEE 802.11 standard).

[0153] In some embodiments, local storage 406 is designed to provide working memory for processing unit 404, providing fast access to programs and / or data to be processed while reducing traffic on interconnect 408. Storage of large amounts of data over a LAN can be provided through one or more mass storage subsystems 412 that can be connected to interconnect 408. Mass storage subsystem 412 can be based on magnetic, optical, semiconductor, and / or other data storage media. Direct-attached storage, storage area networks, network-attached storage, etc., can be used. Any data storage or other data sets generated, used, and / or maintained by a service or server as described herein can be stored in mass storage subsystem 412. In some embodiments, additional data storage resources (potentially with increased latency) can be accessed via WAN interface 410.

[0154] Server system 400 can operate in response to requests received via WAN interface 410. At least one of modules 402 can implement monitoring functions and, in response to received requests, distribute discrete tasks to other modules 402. Work assignment techniques can be used. While processing a request, results can be returned to the requester via WAN interface 410. This operation can typically be automated. Furthermore, in some embodiments, WAN interface 410 can connect multiple server systems 400 to at least one (e.g., each) other server system 400, thereby providing a scalable system capable of managing a large volume of activity. Other techniques for managing server systems and server farms (a collection of cooperating server systems) can be used, including dynamic resource allocation and reallocation.

[0155] Server system 400 can interact with various user-owned or user-operated devices via WAN (such as the Internet). Figure 4 An example of a user-operated device, namely user computing system 414, is shown. User computing system 414 can be implemented as a consumer device, such as a smartphone, other mobile phone, tablet computer, wearable user device (e.g., smartwatch, glasses), desktop computer, laptop computer, etc. User computing system 414 can communicate via WAN interface 410. User computing system 414 may include computer components such as processing unit 416, storage device 418, network interface 420, user input device 422, and / or user output device 424. User computing system 414 can be a user device implemented in various forms, such as a desktop computer, laptop computer, tablet computer, smartphone, other mobile user device, wearable user device, etc.

[0156] Processing unit 416 and storage device 418 may be similar to processing unit 404 and local storage 406 described above. Appropriate devices may be selected based on the requirements to be arranged on user computing system 414; user computing system 414 may be implemented as a "thin" user device with limited processing capabilities or a high-performance user device. User computing system 414 may be equipped with program code executable by processing unit 416 to allow various interactions with server system 400.

[0157] Network interface 420 can provide connectivity to network 426, such as a WAN (e.g., the Internet), to which the WAN interface 410 of server system 400 is also connected. In various embodiments, network interface 420 may include a wired interface (e.g., Ethernet) and / or a wireless interface that implements various RF data communication standards (e.g., Wi-Fi, Bluetooth, and / or cellular data network standards (e.g., 3G, 4G, LTE, 5G, etc.)).

[0158] User input device 422 may include any device (or multiple devices) through which a user provides signals to user computing system 414; user computing system 414 may interpret the signals as indications of user requests or information. In various embodiments, user input device 422 may include at least one of a keyboard, touchpad, touchscreen, mouse and / or other pointing device, scroll wheel, click wheel, dial pad, button, switch, keypad, microphone, etc.

[0159] User output device 424 may include any device through which user computing system 414 can provide information to the user. User output device 424 may include display-to-display images generated by or transmitted to user computing system 414. 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, cathode ray tubes (CRTs), etc., and supporting electronic devices (e.g., digital-to-analog converters or analog-to-digital converters, signal processors, etc.). Some embodiments may include devices that function as both input and output devices (such as touchscreens). In some embodiments, other user output devices 424 may be provided in addition to or in lieu of a display. Examples include indicator lights, speakers, haptic "display" devices, printers, etc.

[0160] Some embodiments include electronic components, such as microprocessors, storage, and / or memory, that store computer program instructions in a computer-readable storage medium. Many of the features described herein can be implemented as processes specified as a set of program instructions encoded on a computer-readable storage medium. When one or more processing units execute these program instructions, they cause the processing units to perform various operations indicated in the program instructions. Examples of program instructions or computer code include machine code, such as machine code generated by a compiler, and / or files containing higher-level code executed by a computer, electronic components, and / or microprocessor using an interpreter. With appropriate programming, processing units 404 and 416 can provide various functions for server system 400 and user computing system 414, including any of the functions and / or other functions described herein that are performed by a server or user.

[0161] It should be understood that server system 400 and user computing system 414 are illustrative and can be varied and modified. Computer systems used in conjunction with embodiments of this disclosure may have additional functionalities not specifically described herein. Furthermore, while server system 400 and user computing system 414 are described with reference to specific blocks, it should be understood that these blocks are defined for ease of description and are not intended to imply a specific physical arrangement of component portions. For example, different blocks may, but do not need to, be on the same facility, the same server rack, and / or the same motherboard. Moreover, blocks do not necessarily correspond to physically different components. Blocks may be configured to perform various operations, such as by programming a processor or providing appropriate control circuitry, and / or various blocks may or may not be reconfigurable depending on how the initial configuration is obtained. Embodiments of this disclosure can be implemented in a variety of devices, including electronic devices implemented using any combination of circuitry and software.

[0162] While this disclosure has described specific embodiments, those skilled in the art will recognize that various modifications can be made. Embodiments of this disclosure can be implemented using various computer systems and communication technologies, including but not limited to the specific examples described herein. Embodiments of this disclosure can be implemented using any combination of dedicated components and / or programmable processors and / or other programmable devices. The various processes described herein can be implemented in any combination on the same or different processors. When a component is described as being configured to perform certain operations, such configuration can be implemented, for example, by designing electronic circuits to perform the operations, by programming programmable electronic circuits (e.g., microprocessors) to perform the operations, and / or any combination thereof. Furthermore, while the above embodiments may refer to specific hardware and software components, those skilled in the art will recognize that different combinations of hardware and / or software components can also be used, and specific operations described as implemented in hardware can also be implemented in software, and vice versa.

[0163] Computer programs incorporating the various features of this disclosure can be encoded and stored on a variety of computer-readable storage media; suitable media include magnetic disks or magnetic tapes, optical storage media (e.g., optical discs (CDs) or digital multifunction discs (DVDs), flash memory, and / or other non-transitory media). Computer-readable media encoding program code can be packaged with a compatible electronic device, and / or the program code can be provided separately from the electronic device (e.g., downloaded via the Internet or as a separately packaged computer-readable storage medium).

[0164] Therefore, although this disclosure has been described with reference to specific embodiments, it should be understood that this disclosure is intended to cover all modifications and equivalents within the scope of the following claims.

Claims

1. A system for structured treatment plan generation and content integration in digital therapy applications, comprising: One or more processors, said one or more processors coupled to memory, said one or more processors being configured to: It receives natural language input to generate treatment plans for digital therapy applications to address the condition; The natural language input and the situation are used as input to apply to at least one generative model to generate a treatment plan for the situation, the treatment plan including a list of treatment components and a timetable for displaying at least one treatment component in the list of treatment components; Retrieve from the database multiple content items corresponding to at least one treatment component in the list of treatment components in the treatment plan; The list of treatment components of the treatment plan is arranged into a structured data package using the multiple content items and the timetable for presentation in the digital therapy application; as well as Provide the structured data packet.

2. The system according to claim 1, wherein, The one or more processors are further configured to: The structured data packet is used to generate multiple instructions, which are configured to cause the digital therapy application to display at least one of the multiple content items according to the arrangement of the list of treatment components in the treatment plan.

3. The system according to claim 1, wherein, The structured data packet corresponds to the mapping of the plurality of content items, the time schedule, or the plurality of execution dependencies of the structured data packet.

4. The system according to claim 1, wherein, The treatment plan includes a timetable of multiple conditional steps that identify at least one treatment component from the list of treatment components in the treatment process.

5. The system according to claim 1, wherein, The treatment plan's timeline includes multiple conditional logics corresponding to at least one treatment component and initiating at least one subsequent treatment component, wherein the at least one subsequent treatment component is provided at least based on the completion or failure of at least one previous treatment component or the completion or failure to meet predefined criteria of the at least one previous treatment component.

6. The system according to claim 1, wherein, The natural language input includes instructions that recognize at least one of the following: (i) treatment duration, (ii) at least one treatment component, (iii) at least one action corresponding to the at least one treatment component, or (iv) at least one criterion for completing the at least one treatment component.

7. The system according to claim 1, wherein, The one or more processors are further configured to: Compile skeleton code retrieved from at least one database or generated using the structured data package, the skeleton code including at least one program construct corresponding to the list of treatment components and the timetable of the treatment plan.

8. The system according to claim 7, wherein, The one or more processors are further configured to: Provide the administrator with the treatment plan or the skeleton code for review; and Before providing the structured data package or using the skeleton code in the digital therapy application, the administrator shall receive at least one update to the treatment plan or the skeleton code.

9. A system comprising: One or more processors, said one or more processors coupled to memory, said one or more processors being configured to: It receives natural language input to generate content for digital therapy applications to address the condition; The natural language input and the situation are used as input to apply to at least one generative model to generate a directed process graph for the situation. The directed process graph includes multiple actions and conditional logic, and at least one of the multiple actions is provided for display according to the conditional logic. Identify at least one data object corresponding to at least one of the plurality of actions contained in the directed process graph; A structured execution framework is generated using the at least one data object and the directed process graph, wherein the structured execution framework identifies (i) multiple content items for identifying at least one of the multiple actions and (ii) corresponding conditional logic, and provides at least one of the multiple actions for display based on the conditional logic; as well as Provide the structured execution framework.

10. The system according to claim 9, wherein, The one or more processors are further configured to: The structured execution framework is used to generate a plurality of instructions, which are configured to cause the digital therapy application to display content including at least one of the plurality of content items according to the directed process graph.

11. The system according to claim 10, wherein, The structured execution framework corresponds to the mapping of the multiple content items, the corresponding conditional logic, multiple timing parameters, or multiple execution dependencies in the directed process graph.

12. The system according to claim 10, wherein, The one or more processors are further configured to: The simulator simulates the performance of the plurality of actions based on processing at least a portion of the plurality of instructions according to the structured execution framework, wherein the plurality of instructions include at least one operation for performing at least one of the plurality of actions corresponding to at least one timing parameter.

13. The system according to claim 11, wherein, The directed process graph includes a hierarchical arrangement of components representing the plurality of actions, the hierarchical arrangement being based on the plurality of timing parameters to identify the order and duration corresponding to at least one of the plurality of actions.

14. The system according to claim 13, wherein, The at least one generative model is at least one of the following: (i) a deep learning model, (ii) a supervised learning model, or (iii) an unsupervised learning model. The process of generating the directed process graph by the at least one generative model includes processing the natural language input to extract a plurality of executable elements, associating the plurality of executable elements with a plurality of actions and a plurality of conditional logics, and arranging the plurality of actions and the plurality of conditional logics into a hierarchical structure.

15. The system according to claim 9, wherein, The conditional logic of the directed process graph includes multiple conditional logics corresponding to at least one treatment component and initiating at least one subsequent treatment component, wherein the at least one subsequent treatment component is provided based at least on the completion or failure of at least one previous treatment component or the completion or non-completion of a predefined criterion of the at least one previous treatment component.

16. The system according to claim 9, wherein, The natural language input includes instructions that recognize at least one of the following: (i) treatment duration, (ii) at least one treatment component, (iii) at least one action corresponding to the at least one treatment component, or (iv) at least one criterion for completing the at least one treatment component.

17. The system according to claim 16, wherein, The directed process graph includes multiple execution paths, and the multiple execution paths of the directed process graph correspond to multiple treatment workflows of the treatment process, wherein each of the multiple treatment workflows includes a sequence of treatment actions to resolve or manage the condition.

18. The system according to claim 9, wherein, At least one of the plurality of actions corresponds to a treatment component or a response to a diagnostic query, and wherein the one or more processors are further configured to: In response to the completion of the treatment component or the provision of a response to a diagnostic query, the activity log is updated to include completion status, timestamp data, or data generated or acquired during the treatment component or diagnostic query.

19. The system according to claim 9, wherein, The one or more processors are further configured to: Compile skeleton code retrieved from at least one database, or generate skeleton code including at least one program construct corresponding to the plurality of actions and corresponding conditional logic using the structured execution framework and the directed process graph.

20. The system according to claim 19, wherein, The one or more processors are further configured to: Provide the administrator with the directed process diagram or the skeleton code for review; and Before providing the structured execution framework or using the skeleton code in the digital therapy application, the system shall receive at least one update to the directed process graph or the skeleton code from the administrator.

21. The system according to claim 9, wherein, Identifying the at least one data object includes querying a database that maintains the at least one data object, and wherein querying the database includes using at least one application programming interface (API) to request the at least one data object corresponding to at least one treatment action or configuration.

22. The system according to claim 21, wherein, The at least one data object includes at least one content item from the plurality of content items and a corresponding parameter, wherein the corresponding parameter includes at least one of content item relations, timing parameters or metadata for representing the at least one treatment action as a node or relation in the directed process graph.

23. A method comprising: Natural language input is received through one or more processors to generate content for digital therapy applications to address the condition; The natural language input and the situation are applied as input to at least one generative model by the one or more processors to generate a directed process graph for the situation, the directed process graph including multiple actions and conditional logic, and providing at least one of the multiple actions for display according to the conditional logic; The one or more processors identify at least one data object corresponding to at least one of the plurality of actions contained in the directed process graph; A structured execution framework is generated by the one or more processors using the at least one data object and the directed process graph. The structured execution framework identifies (i) multiple content items for identifying at least one of the multiple actions and (ii) corresponding conditional logic, and provides at least one of the multiple actions for display based on the conditional logic. as well as The structured execution framework is provided by one or more processors.

24. The method of claim 23, further comprising: The one or more processors use the structured execution framework to generate a plurality of instructions, the plurality of instructions being configured to cause the digital therapy application to display the content, including at least one of the plurality of content items, according to the directed process graph.

25. The method according to claim 24, wherein, The structured execution framework corresponds to the mapping of the multiple content items, the corresponding conditional logic, multiple timing parameters, or multiple execution dependencies in the directed process graph.

26. The method of claim 24, further comprising: The simulator uses one or more processors to simulate the performance of the plurality of actions based on processing at least a portion of the plurality of instructions according to the structured execution framework, wherein the plurality of instructions include at least one operation for performing at least one of the plurality of actions corresponding to at least one timing parameter.

27. The method according to claim 26, wherein, The at least one generative model is at least one of the following: (i) a deep learning model, (ii) a supervised learning model, or (iii) an unsupervised learning model. The process of generating the directed process graph by the at least one generative model includes processing the natural language input to extract a plurality of executable elements, associating the plurality of executable elements with a plurality of actions and a plurality of conditional logics, and arranging the plurality of actions and the plurality of conditional logics into a hierarchical structure.

28. The method according to claim 23, wherein, The conditional logic of the directed process graph includes multiple conditional logics corresponding to at least one treatment component and initiating at least one subsequent treatment component, wherein the at least one subsequent treatment component is provided based at least on the completion or failure of at least one previous treatment component or the completion or non-completion of a predefined criterion of the at least one previous treatment component.

29. The method according to claim 23, wherein, The natural language input includes instructions that recognize at least one of the following: (i) treatment duration, (ii) at least one treatment component, (iii) at least one action corresponding to the at least one treatment component, or (iv) at least one criterion for completing the at least one treatment component.

30. The method according to claim 29, wherein, The directed process graph includes multiple execution paths, and the multiple execution paths of the directed process graph correspond to multiple treatment workflows of the treatment process, wherein each of the multiple treatment workflows includes a sequence of treatment components to resolve or manage the condition.

Images