Hand amputation greatly affects the ability of a person to perform activities of daily living (ADLs). For this reason, prosthetic hands should present grasping characteristics to allow the manipulation of objects of different shapes and dimensions. This is the case of the Hannes prosthetic hand, an under-actuated myoelectric prosthesis able to adapt the grasping configuration to the object shape using the actuation of a single motor and the differential mechanism that characterize this device. In this paper, we present the development of a multibody and multi-domain model of the Hannes hand, which was experimentally validated using an external grasp force sensor to compare the model outcomes to the actual results. The model is used to investigate the correlation between available measurements from the prosthesis and the stiffness of the grasped objects, passing by the exploration of one of the most challenging aspects for tendon-driven under-actuated prosthetic devices: the friction experienced in the transmission mechanism. Therefore, the current analysis leads to the development of a novel control strategy through the use of an object stiffness classifier. The work provides an alternative model-based approach to overcome the absence of force sensors in under-actuated prosthetic hands for object recognition tasks. The obtained results are promising with contained percentage errors.

IntroductionIn recent years, hand prostheses achieved relevant improvements in term of both motor and functional recovery. However, the rate of devices abandonment, also due to their poor embodiment, is still high. The embodiment defines the integration of an external object – in this case a prosthetic device – into the body scheme of an individual. One of the limiting factors causing lack of embodiment is the absence of a direct interaction between user and environment. Many studies focused on the extraction of tactile information via custom electronic skin technologies coupled with dedicated haptic feedback, though increasing the complexity of the prosthetic system. Contrary wise, this paper stems from the authors' preliminary works on multi-body prosthetic hand modeling and the identification of possible intrinsic information to assess object stiffness during interaction.MethodsBased on these initial findings, this work presents the design, implementation and clinical validation of a novel real-time stiffness detection strategy, without ad-hoc sensing, based on a Non-linear Logistic Regression (NLR) classifier. This exploits the minimum grasp information available from an under-sensorized and under-actuated myoelectric prosthetic hand, Hannes. The NLR algorithm takes as input motor-side current, encoder position, and reference position of the hand and provides as output a classification of the grasped object (no-object, rigid object, and soft object). This information is then transmitted to the user via vibratory feedback to close the loop between user control and prosthesis interaction. This implementation was validated through a user study conducted both on able bodied subjects and amputees.ResultsThe classifier achieved excellent performance in terms of F1Score (94.93%). Further, the able-bodied subjects and amputees were able to successfully detect the objects' stiffness with a F1Score of 94.08% and 86.41%, respectively, by using our proposed feedback strategy. This strategy allowed amputees to quickly recognize the objects' stiffness (response time of 2.82 s), indicating high intuitiveness, and it was overall appreciated as demonstrated by the questionnaire. Furthermore, an embodiment improvement was also obtained as highlighted by the proprioceptive drift toward the prosthesis (0.7 cm).