Design, implementation and testing of a three degree of freedom adaptive grasp prosthetic hand

The objective of this study was to design three control degrees of freedom (DOF), adaptive grasp prosthetic hand. The proposed design leveraged the limitations in commercial prostheses which utilize only two control DOFs, lack of independent thumb control and no inter-phalangeal joint, limiting the prehension patterns and naturalness of movement of the prosthesis. A lack of visual force feedback in commercial prostheses renders the user incapable of estimating the grip force being exerted on the object that he/she is grasping. This feature is provided by other commercial devices only during training.;A control scheme was designed and emulated in real-time using LabVIEW 8.2RTM. Pre-amplifiers and motor drive circuits were designed and implemented. This scheme used three muscles to control the finger extension, finger flexion and thumb flexion independently. ABS polymer palm was designed to house the drive system of the prosthesis. A non-back drivable bi-directional clutch was designed for circuit protection and user safety. Centrally driven gearbox controlled the flexion-extension of the fingers through pulleys. Kevlar strings were used as tendons. Fingers were designed from SantopreneRTM and had one inter-phalangeal joint as a living hinge. A frictionless metacarpo-phalangeal joint and an elastomer inter-phalangeal joint allowed for shape adaption. A majority of mechanical components were fabricated from polymers to keep the hand light-weight. Force sensing resistors were used to obtain an on-screen artificial visual force feedback for the hand.;Human subject testing was conducted along with other tests to evaluate the performance of this prosthesis. Human subject testing (12 subjects) of the control system emulation proved that independent control of thumb motion was possible with this scheme. A visual analog scale rating of 9.28/10 made the system and the program user interface very easy to learn. The system was fairly easy to control with an average score of 6.3/10 and a marginally significant difference (p=0.057) was observed between genders, where females perceived the system to be easier to control than males. The control scheme had a response time of 0.29s (sampling frequency= 7 kHz). The hand weighs 315g and is very light in weight compared to existing commercial devices. The hand exerted a grip force of 3.5N which is very low compared with other prostheses. This hand should be tested on amputees using a cosmetic glove. Further work needs to be done to improve the performance standards of this hand prosthesis.