US20250302653
2025-10-02
Human necessities
A61F5/0123
The knee brace system integrates advanced sensor technologies, including inertial measurement units (IMUs) and electromyography (EMG) sensors, with machine learning models to prevent injuries to the anterior cruciate ligament (ACL) and posterior cruciate ligament (PCL). It collects real-time biomechanical and neuromuscular data, analyzing these inputs to detect high-risk movements. Upon identifying potential injury risks, the system can issue alerts or activate a hybrid actuation system designed to reduce joint loading and prevent injury.
Machine learning techniques such as supervised, personalized, or federated learning are employed to enhance the system's predictive capabilities. These models are trained to recognize patterns indicative of injury risk, allowing for personalized adaptation through calibration activities and transfer learning. Federated learning ensures privacy while improving performance by aggregating data from multiple users without sharing individual data.
Feedback mechanisms include visual, auditory, or haptic interfaces, providing users with real-time information about their movement patterns and potential risks. The system can be integrated with rehabilitation tools or mobile applications, making it suitable for use in various environments such as sports, dance, clinical settings, or rehabilitation. This integration helps in optimizing recovery and enhancing performance while reducing injury risks.
The knee brace utilizes a combination of high-force, low-displacement actuators and low-force, high-displacement actuators. Piezoelectric actuators are highlighted for their precision and quick response times. These actuators absorb excess torque detected by sensors, converting it into mechanical energy to mitigate potential damage. The system disengages immediately after activation to allow uninterrupted user activity.
EMG sensors measure neuromuscular activity by detecting electrical impulses from muscle contractions. IMUs measure linear acceleration and angular rotation, providing comprehensive motion metrics. Both sensor types contribute data for machine learning analysis, enabling the prediction of hazardous movements. This predictive capability is crucial for preventing knee injuries by allowing timely interventions based on real-time data.