"Compliant Magnetic FMG Array Dataset"

"Control of powered prosthetic hands could benefit from improved techniques to infer the amputees\u2019 desired grasp intentions. Force myography (FMG) has recently emerged as a potential alternative to electromyography, which is traditionally used in clinical practice.  Due to their practical and robust attributes, we introduce innovative compliant magnetic FMG sensor arrays for forearm muscle pattern recognition. Sensor arrays with 18 or 24 compliant magnetic FMG sensor modules were created, demonstrating the customizability of our Scan2Make process for creating individualized sockets with integrated FMG sensor arrays for varying limb differences. Real-time control experiments with subject-specific artificial neural networks (ANN) showed that non-amputees were able to produce 15.43 \u00b1 2.37 grasp gesture classes with over 90% accuracy. The three amputees produced an average of 9.00 \u00b1 3.61 gestures with over 90% accuracy in real-time. Additionally, principal component analysis (PCA) was used to both reduce sensor count and improve classification accuracy in subsequent real-time playback analyses. Seven out of ten subjects achieved their highest accuracy with a reduced number of compliant magnetic sensor modules. To our best knowledge, this study is the first to use arrays of compliant magnetic sensors for forearm muscle pattern recognition to enable real-time control of a robotic hand. Furthermore, this is the first effort to use a 3D scanning and printing approach to create custom prosthetic sockets with any kind of integrated FMG sensor. Additionally, we are providing the first open-source database of FMG signals from 3 amputees and 7 non-amputees using the compliant magnetic sensor arrays as a resource for the research community. These novel compliant magnetic FMG sensor arrays have the potential to advance the state of the art for prosthetic hand control and could be used broadly in the fields of tactile sensing, haptics, teleoperation, and rehabilitation. "