Kinematic design of linkage-based haptic interfaces for medical applications: a review

Abstract

A haptic interface recreates haptic feedback from virtual environments or haptic teleoperation systems that engages the user’s sense of touch. High-fidelity haptic feedback is critical to the safety and success of any interaction with human beings. Such interactions can be seen in haptic systems utilized in medical fields, such as for surgical training, robotic tele-surgery, and tele-rehabilitation, which require appropriate haptic interface design and control. In order to recreate high-fidelity soft and stiff contact experiences for the user in the intended application, different designs strike different trade-offs between the desirable characteristics of an interface, such as back-drivability, low apparent inertia and low friction for the best perception of small reflected forces, large intrinsic stiffness and force feedback capability for the best perception of large reflected forces, a large-enough workspace for exploring the remote or virtual environment, and the uniformity of haptic feedback and its adequate sensitivity over the workspace. Meeting all of the requirements simultaneously is impossible, and different application-driven compromises need to be made. This paper reviews how various kinematic designs have helped address these trade-offs in desired specifications. First, we investigate the required characteristics of linkage-based haptic interfaces and inevitable trade-offs between them. Then, we study the state of the art in the kinematic design of haptic interfaces and their advantages and limitations. In all sections, we consider the applications of the intended haptic interfaces in medical scenarios. Non-linkage-based haptic interfaces are also shortly discussed to show the broad range of haptic technologies in the area. The potentials of kinematic redundancy to address the design trade-offs are introduced. Current challenges and future directions of haptic interface designs for medical applications are shortly discussed, which is finally followed by the conclusion.

Workspace and dexterity analysis of the hybrid mechanism master robot in Sinaflex robotic telesurgery system: An in vivo experiment

“…The serial mechanisms have easy mathematics for forward and inverse kinematic calculation, a large workspace and high dexterity, while their high inertia and small stiffness lead to a large end effector position error. 5,7 Three well-known examples are phantom omni (touch), 8 phantom premium, 9 and phantom desktop (touch X) 10 as the 6-DoF serial mechanisms. The master robot of the da Vinci system 11 is an 8-dof serial mechanism.…”

Workspace and Dexterity Investigation of the Master Robot in Sina flex Robotic Telesurgery System: An in Vivo Validation

“…Robotic surgery systems are composed of the surgeon side and patient side subsystems. To fulfill the surgeon desires to access to all points of the surgery sites, a good master robot should provide three important characteristics: enough degree of freedom (DoF), a large enough workspace, and high dexterity [4,5]. In this regard, various master robots or haptic interfaces are developed [6].…”

“…Each of the robot mechanisms has its own pros and cons. The serial mechanisms have easy mathematics, a large workspace and high dexterity, while their high inertial and small stiffness leads to a large end effector position error [4,6]. Two well-known examples are phantom omni [7], and phantom premium [8] as the 6-DoF serial mechanisms.…”

“…In contrast to the serial structures, the parallel mechanisms are accurate in end effector positioning due to their low inertia and high stiffness. However, their workspace and dexterity are limited and the solutions for forward and inverse kinematics are complicated [4,6]. As the commercial haptic interfaces with parallel mechanisms, Omega.7 [11] and Sigma.7 [12] (7-DoF), can be mentioned.…”

“…A 7-DoF parallel mechanism is presented in [14] whose workspace size is 480×450×250 mm 3 . Hybrid mechanisms stand somewhere in the middle where provide desirable workspace and dexterity along with low end effector position error [4,6]. In these mechanisms, positions and orientation of the end effector are usually set independently based on the two different mechanisms.…”