LLM-Generated Personalized Feedback for Digital Therapy

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Traditional motion analysis systems in digital therapy lack personalization and adaptability, providing repetitive and generic feedback that fails to account for a patient's nuanced movements and therapeutic journey, disrupting the session flow and engagement.

Innovation Solution

Integration of large language models (LLMs) to analyze movement statistics and generate personalized, context-aware feedback, converted to real-time audio for enhanced interaction.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If hardcoded rules are used for feedback, then the system is simple to implement, but the feedback lacks personalization and adaptability

Engineering Contradiction:

Improvesystem implementation simplicityVSAvoidfeedback personalization

Core Design Contradiction:

Ease of manufactureVSAdaptability or versatility

Solution Approach 1:

The patent replaces the mechanical hardcoded rule-based system with an AI language model that processes patient data and generates personalized feedback dynamically. This substitution enables the system to adapt to individual patient needs while maintaining ease of implementation through a unified AI-driven architecture.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The feedback system transitions from static hardcoded rules to dynamic AI-generated responses that adapt in real-time based on patient performance, session context, and therapeutic goals. This allows the system to provide personalized feedback without requiring complex pre-programming for every scenario.

Inventive Principle:

Principle #15Dynamics

2Loss of information

If text feedback is provided, then detailed information can be conveyed, but patient focus shifts away from exercise

Engineering Contradiction:

Improvefeedback information completenessVSAvoidsession flow continuity

Core Design Contradiction:

Loss of informationVSEase of operation

Solution Approach 1:

The patent introduces audio feedback as an intermediary medium that conveys detailed therapeutic information without requiring the patient's visual attention. This allows comprehensive feedback to be delivered while maintaining continuous visual focus on the exercise demonstration and proper form.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system replaces text-based feedback delivery with audio-based feedback, substituting the visual-textual interface with an auditory interface that does not compete for the patient's visual attention during exercise execution.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Device complexity

If generic feedback is provided, then the system is easier to maintain, but patient engagement decreases

Engineering Contradiction:

Improvesystem maintenance complexityVSAvoidpatient engagement

Core Design Contradiction:

Device complexityVSProductivity

Solution Approach 1:

The feedback system evolves from static generic responses to dynamic personalized feedback that adapts to each patient's progress, preferences, and real-time performance. This increases engagement while the underlying AI model maintains manageable complexity through standardized processing frameworks.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The AI language model enables the system to generate personalized feedback autonomously without requiring manual customization for each patient. This self-service capability maintains system simplicity while delivering engaging personalized content at scale.

Inventive Principle:

Principle #25Self-service

4Speed

If feedback is based solely on current session movements, then real-time responsiveness is achieved, but context from past performance is lost

Engineering Contradiction:

Improvefeedback responsivenessVSAvoidtherapeutic journey context

Core Design Contradiction:

SpeedVSLoss of information

Solution Approach 1:

The system performs preliminary analysis by integrating historical performance data and therapeutic context before generating real-time feedback. This allows the AI to provide responsive feedback that is informed by the patient's broader therapeutic journey, combining immediate movement analysis with past performance patterns.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent merges real-time movement data with historical performance data and therapeutic context into a unified feedback generation process. This integration enables the system to maintain real-time responsiveness while incorporating comprehensive contextual information from the patient's therapeutic journey.

Inventive Principle:

Principle #5Merging (Combining)

Data Source

PatentUS20250312651A1Personalized communication in a digital therapy platform

Publication Date: 2025.10.09 SWORD HEALTH SA

AI summary

Systems and methods in the present disclosure relate to technology for automated and personalized message generation. A user interface on a user device presents instructions for performing a therapeutic activity. During a session, one or more sensors capture activity data for a user while the user performs the therapeutic activity. The activity data is processed to track performance of the therapeutic activity and generate session data. A prompt for a language model is automatically generated using a prompt data structure that is updated in real-time based on at least one of the session data or historical data associated with the user. A personalized message, as generated in real-time by the language model, is presented to the user via the user interface or via audio hardware associated with the user device.