Neural Network Estimating Knee Joint Force from Insole Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for estimating medial tibiofemoral joint reaction force are complex, expensive, and limited to laboratory settings, making it difficult to monitor joint forces in real-world scenarios, particularly for early prediction and diagnosis of knee osteoarthritis.
Innovation Solution
A system and method using instrumented insoles and a trained neural network model to estimate medial joint contact force by processing force plate and motion capture data, allowing for real-time monitoring outside of laboratory settings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If complex musculoskeletal models and laboratory equipment are used to estimate medial tibiofemoral joint reaction force, then measurement precision is improved, but device complexity and cost increase significantly
Solution Approach 1:
The patent creates a simplified computational model that copies the essential function of complex musculoskeletal models. Instead of using full-scale musculoskeletal modeling requiring motion capture systems and force plates, the invention develops a machine learning model trained on such data that replicates joint reaction force estimation using only wearable sensor inputs, thereby maintaining measurement precision while dramatically reducing device complexity
Solution Approach 2:
The patent replaces the mechanical measurement system (force plates, motion capture cameras, and complex musculoskeletal modeling hardware) with an electronic/computational system. Wearable sensors embedded in clothing or accessories collect biomechanical data, which is then processed by a machine learning model to estimate joint reaction forces, substituting mechanical measurement infrastructure with electronic sensing and computational analysis
2Measurement precision
If complex musculoskeletal models and laboratory equipment are used to estimate medial tibiofemoral joint reaction force, then measurement precision is improved, but cost increases significantly
Solution Approach 1:
The patent creates a simplified computational model that copies the essential function of complex musculoskeletal models. Instead of using full-scale musculoskeletal modeling requiring motion capture systems and force plates, the invention develops a machine learning model trained on such data that replicates joint reaction force estimation using only wearable sensor inputs, thereby maintaining measurement precision while dramatically reducing device complexity
Solution Approach 2:
The patent replaces expensive, permanent laboratory infrastructure with inexpensive, disposable or reusable wearable sensors. The machine learning model can be deployed on consumer-grade devices, eliminating the need for costly force plates, motion capture systems, and specialized laboratory equipment, making the technology economically viable for widespread use
3Measurement precision
If complex musculoskeletal models and laboratory equipment are used to estimate medial tibiofemoral joint reaction force, then measurement precision is improved, but ease of operation is worsened due to requiring trained researchers and lab settings
Solution Approach 1:
The patent creates a simplified computational model that copies the essential function of complex musculoskeletal models. Instead of using full-scale musculoskeletal modeling requiring motion capture systems and force plates, the invention develops a machine learning model trained on such data that replicates joint reaction force estimation using only wearable sensor inputs, thereby maintaining measurement precision while dramatically reducing device complexity
Solution Approach 2:
The patent enables the system to automatically perform joint reaction force estimation without requiring trained researchers to operate complex equipment. The wearable sensors autonomously collect data, and the machine learning model automatically processes this data to provide joint force estimates, eliminating the need for specialized laboratory operators and making the system easy for end-users to operate independently
4Measurement precision
If traditional laboratory-based methods are used, then measurement precision is improved, but adaptability to real-world scenarios is reduced due to being restricted to lab settings
Solution Approach 1:
The patent transitions from static laboratory measurements to dynamic real-world monitoring. The wearable sensors continuously collect biomechanical data during natural activities in everyday environments, allowing the machine learning model to adapt to varying conditions, terrains, and activity types while maintaining measurement accuracy, thereby achieving both precision and real-world applicability
Solution Approach 2:
The patent creates a universal system that functions across multiple settings and activities. The wearable sensors and machine learning model can operate in laboratories, clinical settings, and everyday environments, supporting various activities from walking to sports, making the technology versatile and adaptable to real-world scenarios while maintaining measurement precision
Data Source
AI summary
Methods and systems are provided for determining medial joint contact force. For example, methods and systems described herein may train a neural network model to determine medial joint contact force. A plurality of training data is obtained and a plurality of heel strike to toe-off time periods are extracted from the training data. A gait categorization label is assigned. A ground reaction force and joint contact forces are determined. An input training array and an output training array is provided to the neural network. The neural network is trained to output a predicted joint contact force value corresponding to a calculated joint contact force measured from force plate data, based only on insole data. The trained neural network model is stored after it has been validated.