Enabling Independent Living for Individuals with Cervical Spinal Cord Injury via High-Density Electromyography Controlled Robotic Systems

Carnegie Mellon University, Pittsburgh PA

Investigators

Abstract

In the United States, there are over 100,000 individuals living with a spinal cord injury who have lost the use of their hands. The loss of hand function can dramatically change a person’s daily life, forcing them to rely on caregivers to provide assistance. This project aims to develop sensor-embedded garments designed to monitor muscle activity in individuals with tetraplegia resulting from spinal cord injuries. This sensorized garment detects muscle activity in the upper body, even for many people with spinal cord injuries who have lost hand mobility. By decoding and translating these muscle activities, the garment empowers individuals to take control of an assistive robot to perform a variety of household and self-care tasks, such as preparing and eating meals independently and conducting household chores. This research is poised to impact society by restoring a degree of autonomy and agency to individuals with spinal cord injury, promoting open-source development in wearable sensing, integrating research findings into classroom materials, and fostering inclusive education and mentorship for high school students from communities underserved in STEM. This collaborative project between Carnegie Mellon University and the University of Pittsburgh will introduce a high-density electromyography (HDEMG) wearable to capture neuromuscular signals and enable individuals with spinal cord injury (SCI) to embody a physically assistive mobile manipulator. Towards this goal, the research project will make key foundational contributions along three thrusts: (1) The design of a sensorized upper-body garment embedded with HDEMG arrays, allowing the capture of residual myoelectric activity and recognition of gesture intent in individuals with SCI. The research team will develop learning-based pattern recognition techniques to interpret the myoelectric activity patterns of attempted arm, wrist, and hand gestures and predict the underlying motor function intent. (2) Development of an interface that enables individuals with SCI to embody an assistive robot by mapping myoelectric signals and inferred gestures to actuator control. The sensorized garment is designed to be comfortable and wearable throughout an entire day, which also necessitates the development of real-time calibration algorithms that allow for accurate myoelectric activity sensing over extended periods of time. (3) The team will conduct a series of in-lab evaluations with stakeholders in collaboration with our community partners. Individuals with SCI will use the wearable to teleoperate an assistive mobile manipulator and perform a range of physical household and self-care activities. These studies will quantify technology performance and efficacy, assess safety, and support iterative refinement and improvements to the key technology components with stakeholder feedback. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.