Michele Canepa

Giulio Maria Bianco, Alberto Dellacasa Bellingegni, Federico Mereu et al.

To provide multimode sensory feedback and motion control, bidirectional bionic interfaces for advanced prosthetic systems require continuous and secure energy delivery to implantable electronics and integration in the sensing WBAN (Wireless Body Area Network) of the patient. However, powering such interfaces is still an open issue. Wireless Power Transfer (WPT) avoids implanted batteries and transcutaneous connections, but its design is constrained by stringent requirements on electromagnetic safety, implant size, voltage compliance, and coexistence with sensitive bio-signal acquisition and stimulation circuitry. This paper presents the design and testing of a low-frequency (127 kHz) inductive WPT link for an implantable bidirectional bionic interface. The system includes an autoresonant driving control to maintain operation at resonance under varying coupling and load conditions of the cyber-physical prosthesis. Starting from the requirements of the bionic interface, the wireless body-area sensing system is designed by selecting the working frequency, drawing the electrical schemes, and checking its safety and regulatory compliance. Preliminary WPT prototypes can provide up to ~140 mA and ~20 V, achieving a maximum power transfer efficiency higher than 40% and satisfying the project requirements up to a 2 cm implantation depth.

Dario Di Domenico, Nicolò Boccardo, Andrea Marinelli et al.

Noninvasive human-machine interfaces such as surface electromyography (sEMG) have long been employed for controlling robotic prostheses. However, classical controllers are limited to few degrees of freedom (DoF). More recently, machine learning methods have been proposed to learn personalized controllers from user data. While promising, they often suffer from distribution shift during long-term usage, requiring costly model re-training. Moreover, most prosthetic sEMG sensors have low spatial density, which limits accuracy and the number of controllable motions. In this work, we address both challenges by introducing a novel myoelectric prosthetic system integrating a high density-sEMG (HD-sEMG) setup and incremental learning methods to accurately control 7 motions of the Hannes prosthesis. First, we present a newly designed, compact HD-sEMG interface equipped with 64 dry electrodes positioned over the forearm. Then, we introduce an efficient incremental learning system enabling model adaptation on a stream of data. We thoroughly analyze multiple learning algorithms across 7 subjects, including one with limb absence, and 6 sessions held in different days covering an extended period of several months. The size and time span of the collected data represent a relevant contribution for studying long-term myocontrol performance. Therefore, we release the DELTA dataset together with our experimental code.