Wearable Joint Monitoring via Hybrid Sensor Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current wearable sensor systems for monitoring joint and muscle dynamics in orthopedics require complex arrays of sensors, which are cumbersome and discourage use, while existing techniques either rely too heavily on physics-based simulations that need many sensors or machine learning methods that fail to characterize individual muscle contraction dynamics effectively.
Innovation Solution
A hybrid technique using a minimal array of wearable sensors that combines physics-based and probabilistic models, specifically Gaussian process synergy functions, to estimate muscle excitations and joint moments, allowing for the reduction of sensor numbers and improving generalizability, with sensors embedded in braces or sleeves for practical deployment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If complex sensor arrays are used to accurately measure muscle and joint dynamics, then measurement precision is improved, but device complexity increases and ease of operation deteriorates
Solution Approach 1:
The patent extracts and eliminates redundant sensors from the measurement system. By using a minimal array of only two inertial sensors instead of complex sensor arrays, the system removes unnecessary components while maintaining the ability to measure joint and muscle dynamics through intelligent signal processing and biomechanical modeling.
Solution Approach 2:
The inertial sensors are designed to perform multiple functions: measuring segment acceleration, orientation, and angular velocity, as well as providing input for computing joint kinetics and muscle forces. This multi-functionality reduces the total number of sensors needed while maintaining comprehensive measurement capabilities.
2Reliability
If physics-based simulation methods are used to estimate muscle forces and joint moments, then reliability is improved, but device complexity increases due to required sensor数量
Solution Approach 1:
The patent changes the input parameters required for physics-based simulation from numerous sensor measurements to a minimal set of inertial sensor data combined with biomechanical models. By transforming the problem from direct measurement to model-based estimation using fewer parameters, the system maintains simulation reliability while reducing sensor requirements.
3Device complexity
If machine learning methods are used to reduce sensor numbers, then device complexity is reduced, but measurement precision deteriorates due to inability to characterize individual muscle contraction dynamics
Solution Approach 1:
The patent introduces biomechanical models and signal processing algorithms as intermediaries between the minimal sensor data and the muscle force estimates. These intermediaries transform the raw inertial measurements into meaningful physiological parameters while preserving individual muscle contraction dynamics that pure machine learning approaches would miss.
Data Source
AI summary
A system for determining dynamics of a joint of an individual includes a first muscle contraction sensor configured to measure an excitation of a first muscle adjacent to a joint; a first movement sensor configured to measure movement on a first side of the joint; a second muscle contraction sensor configured to measure an excitation of a second muscle located adjacent to the joint; and a second movement sensor configured to measure movement on a second side of the joint. A machine learning processor is trained to determine a full set of excitation values based on an excitation value from the first muscle contraction sensor and an excitation value from the second muscle contraction sensor. A processor is configured to determine a joint moment based on values from the first and second movement sensors and the set of excitation values from the machine learning processor.