A team of researchers from the University of Malta and from Malta have successfully developed a prototype for a new prosthetic hand, based on an innovative approach that focuses on optimising the dexterity and intuitive usability of the system while simultaneously reducing the complexity of the overall design. This research is part of the MAProHand project, entitled “Development of the mechanical and control framework for a minimal anthropomorphic prosthetic hand” which was launched in March 2019 and will run until the end of November 2022.
The amputation of an upper limb, often following an accident, represents a catastrophic loss to the functionality in daily living, and often also to the self-image of the person involved. Over the years, several approaches have been taken to the development of prostheses to replace the missing limb. These range from purely aesthetic devices (with no dexterity); through open-close devices that are powered through an intact muscle of the user such as through a shoulder harness (with very limited dexterity); to highly complex and dexterous battery-powered five-fingered prostheses that are controlled through detection of faint electrical signals that are generated in the forearm of the user whenever the forearm muscles are clenched.
The dexterous prostheses are typically very expensive, rendering them financially inaccessible to many potential users, and are also typically quite heavy, leading to high rejection rates by the users. The use of the electrical signals (called surface electromyography, or sEMG, signals) from the forearm to control these hands is generally counter intuitive, where for example the user would clench and unclench the muscles a number of times to select the required grasp. This is of course not the way we use our natural hands!
This is where MAProHand came in. Early on, the local research team made a number of critical observations. Firstly, they had already noted in their own previously published work that the active use of only three digits (the thumb, and the index and middle fingers) was enough to attain a high degree of dexterity in manual function. This indicated that complexity could be reduced while maintaining dexterity. Secondly, they noted that when a person loses a hand, the remaining hand assumes the role of the dominant hand, irrespective of whether it had this role beforehand. This meant that the prosthesis was only required to attain the functionality of a non-dominant hand (e.g. the left hand of a right-handed person, used primarily for grasping during activities that require two hands) which reduced further the required complexity, as well as the number of distinct signals required to control the hand. And thirdly, they noted that sEMG signals were more heavily associated with muscle movements than with muscle positions, meaning that if the user aimed to try to control the missing hand as if it were there, to execute specific gestures, distinct sEMG patterns could perhaps be detected and applied in an intuitive way.
The MAProHand researchers first carried out a series of experiments where human subjects used their natural hands to grasp various objects that were carefully selected based on their importance in two- handed activities of daily living, while having their ring and middle fingers constrained, and while wearing a specially designed data glove to measure the various joint positions during each grasp.
The data were then mathematically analysed to extract three basic modes of hand motion that were sufficient to carry out all of the tasks. These results were then used to develop a sEMG detection and analysis system that is able to detect and distinguish between these specific hand motions, as well as one hand-opening motion, to a very high reliability, both in persons having a natural hand as well as in persons missing a hand either from birth or through amputation. Meanwhile, the researchers carried out further experimentation and analysis to determine the minimum forces required to be applied by the prosthetic hand, and then designed in detail the mechanical device and the motor and gear systems required to move it according to the required basic motions and to apply the required forces.
The prosthetic hand has five fingers, however the ring and little fingers move together with the middle finger rather than independently, so only three motors in total are required. The researchers also developed a clever system to read the forces being applied by the prosthetic hand through force sensors fitted inside the gear systems, and to transmit the information directly to the user using small flat motors that can vibrate against the skin, therefore giving the user touch and force sensations. The prototype was built and the system has been demonstrated successfully in the laboratory, and shown to compare favourably to other available devices. The researchers are now working to convert the system from a physically demonstrated concept to a commercial reality.
The project is led by Professor Michael A. Saliba of the Department of Mechanical Engineering of the University of Malta, and includes other researchers who have been employed by the department specifically for the project as well as academics from the Departments of Systems and Control Engineering and of Artificial Intelligence at UM. Orthopaedic Centre Malta has also employed and deployed researchers to work specifically on MAProHand. The project has been funded under the national Fusion research and innovation programme through the Malta Council for Science and Technology.
