Anatomical exoskeleton for movement approximation with neural networks

Abstract

A lower extremity exoskeleton for the right leg is assembled using 3D printed soft plastic parts within a semi-rigid frame and structure balancing the rigid and soft flexible components. One leg has been assembled at this time allowing us to test the function of the mechanical system's frame and structure. The comfort and safety of the exoskeleton is important for increasing the time the exoskeleton can be tolerated by a patient. The knee is the largest and most complicated of the lower extremity joints. This exoskeleton joint accommodates the rotational and sliding movement of the femur at the knee to increase the comfort level by allowing for the inherent anatomical motion of the knee. This is often not taken into consideration for most lower extremity orthopedic braces, exoskeletons, or prosthetics. Surface EMGs and IMUs are utilized as input sensors for the exoskeleton control system. Myo-Ware EMG/IMU sensors from Thalmic labs originally designed for the upper extremity must be adapted for our lower extremity exoskeleton. Pneumatic artificial muscles and Bowdoin cables controlled by electric motors are incorporated to assist movement as needed. The EMG/IMU input data is modified and synchronized by a supervised machine learning sensor fusion algorithm. EMG and IMU characteristics help reduce noise and synchronize input data. A reinforcement unsupervised learning (RL) algorithm determines intention and control the exoskeleton actuators. This depends on adequately implementing the RL network algorithm, thus testing the activities of daily living (standing, sitting, squatting while maintaining balance) are needed.

Deep reinforcement learning to assess lower extremity movement intention and assist a rehabilitation exoskeleton

“…The sensor design for the unilateral lower extremity exoskeleton plays a crucial role in bridging the gap between user intention and the exoskeleton's responsive movements. The integration of surface electromyography (sEMG) and inertial measurement units (IMU) into a cohesive sensor fusion framework allows for real-time, nuanced control of the exoskeleton, ensuring movements are both intuitive and aligned with the user's natural biomechanics [3]. This setup not only facilitates precise control over the PAMs but also underscores the potential for significant enhancements in rehabilitation technologies.…”

“…The integration of LSTM networks into our exoskeleton control system enables the creation of assistive devices with improved responsiveness, individualized to the user needs. By learning from the temporal sequences of sensor data, the LSTM model can adapt to the user's unique movement patterns, offering personalized support that improves over time [3]. This adaptability is further enhanced by training the LSTM model on a labeled dataset of movements, allowing it to recognize and respond to a wide range of activities.…”

“…Our neural network model marries the temporal precision of LSTM networks with the inherent decision-making ability of PPO, creating a system that is adept at interpreting and responding to sequential sensor data in real-time. The LSTM component is designed to process and remember sequences of data over time, making it perfectly suited for tasks that involve understanding the continuity and progression of human movement [3]. This capability allows our model to forecast the user's next move based on historical data, ensuring that the exoskeleton's response is both timely and attuned to the user's current needs.…”