The First-Aid Robotic Glove for Stroke Patients
Everyday tasks can become incredibly difficult for individuals who have experienced neurotrauma, such as a stroke, due to decreased coordination and strength in their upper limbs. To address this challenge, researchers at Florida Atlantic University’s College of Engineering and Computer Science have developed a groundbreaking robotic glove. This soft exoskeleton glove, powered by artificial intelligence (AI), is specifically designed to improve hand dexterity in stroke survivors.
Piano Playing Made Possible Again
The unique robotic glove incorporates flexible tactile sensors, soft actuators, and AI technology. It is the first of its kind to accurately differentiate between correct and incorrect variations of a song, integrating all these features into a single hand exoskeleton.
The ability to play the piano demands complex and highly skilled hand movements. Relearning these tasks requires the restoration of specific movements and skills. The robotic glove, composed of soft and flexible materials, gently supports and assists stroke survivors in regaining their motor abilities.
Precise Force and Guidance
The researchers integrated special arrays of sensors into each fingertip of the robotic glove. Unlike previous exoskeleton devices, this new technology provides precise force and guidance, enabling stroke patients to recover the fine finger movements necessary for piano playing. By continuously monitoring and responding to the user’s movements, the glove offers real-time feedback and adjustments, making it easier to grasp the correct techniques.
Classification Accuracy and Customization
To demonstrate the glove’s capabilities, the researchers programmed it to identify errors in the well-known tune “Mary Had a Little Lamb” played on a piano. The robotic glove was trained using the Random Forest (RF), K-Nearest Neighbor (KNN), and Artificial Neural Network (ANN) algorithms to differentiate between correct and incorrect variations of the song played by the user or independently. The ANN algorithm achieved the highest classification accuracy of 97.13% with a human subject and 94.60% without one. It successfully identified errors and out-of-time key presses. These findings highlight the potential of the glove to assist individuals in relearning complex tasks like playing musical instruments.
Crafting a Revolution
The robotic glove was designed using 3D printed polyvinyl acid stents and hydrogel casting, allowing for the integration of five actuators into a single wearable device. This fabrication process is simpler than most designs since all the actuators and sensors are combined into a single molding process. Additionally, the form factor of the glove can be customized using 3D scanning technology or CT scans to meet the unique anatomy of each patient.
Clinicians can utilize the data gathered from the glove to develop personalized action plans that target patient weaknesses. These weaknesses may manifest as consistent errors in specific sections of the song, indicating areas in need of improvement. As patients progress, the rehabilitation team can prescribe more challenging songs in a game-like progression, providing a customizable path to recovery.
This groundbreaking technology developed by Florida Atlantic University is a game-changer for individuals with neuromuscular disorders and reduced limb functionality. While previous soft robotic actuators have also been used for playing the piano, this robotic glove is the only one capable of “feeling” the difference between correct and incorrect versions of the same song.
Sources: Frontiers in Robotics and AI