Wearable Physiological Biomarkers for Disruptive Behavior Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing behavioral therapies for children, such as Parent-Child Interaction Therapy (PCIT), require consistent parental effort and interaction to be effective, and there is a need for accurate prediction of disruptive behavior to facilitate timely interventions.
Innovation Solution
A wearable device-based system using machine learning models, trained on motion, heart rate, and sleep data, to predict disruptive behavior in children, generating predictive feature data and alerts for parents or caregivers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If machine learning models are trained on comprehensive physiological data (motion, heart rate, sleep), then prediction accuracy for disruptive behavior is improved, but data collection complexity and device requirements increase
Solution Approach 1:
The patent employs a single wearable device that integrates multiple sensor types (accelerometer for motion, heart rate monitor, sleep tracker) to collect diverse physiological data. This multi-functional approach enables comprehensive prediction accuracy without requiring multiple separate devices, thus resolving the contradiction between measurement precision and device complexity.
Solution Approach 2:
The patent combines multiple data streams (motion data, heart rate data, sleep data) into a unified prediction model. By merging these diverse physiological measurements through a single machine learning framework, the system achieves high prediction accuracy while simplifying the overall system architecture compared to using separate specialized devices for each measurement type.
2Speed
If real-time monitoring and prediction alerts are implemented, then timely intervention capability is improved, but processing speed and computational requirements increase
Solution Approach 1:
The patent pre-trains machine learning models offline using historical data, creating optimized prediction algorithms that can be executed efficiently on wearable devices. This preliminary action of model training outside the real-time environment allows the system to make rapid predictions during actual use without requiring excessive computational power at the moment of prediction, thus resolving the contradiction between response speed and processing power requirements.
3Measurement precision
If multiple physiological parameters are monitored continuously, then behavior prediction accuracy is improved, but energy consumption increases
Solution Approach 1:
The patent implements periodic sampling of physiological data rather than continuous monitoring at maximum frequency. The system adjusts sampling rates based on activity context (e.g., higher frequency during sleep hours, lower frequency during awake periods), which maintains adequate prediction accuracy while significantly reducing overall energy consumption. This periodic action strategy resolves the contradiction between measurement precision and energy usage.
Data Source
AI summary
Impending disruptive behavior in an individual is predicted using a machine learning model that processes measurement recorded with a wearable device. Measurement data are received from a wearable device and input to a trained machine learning model, generating predictive feature data as an output. The predictive feature data may include predictive scores, classifications, or the like, of a likelihood of a subject having a disruptive behavior.