@article {1723, title = {Interaction control capabilities of an MR-compatible compliant actuator for wrist sensorimotor protocols during fMRI}, journal = {IEEE/ASME Transactions on Mechatronics}, volume = {20}, number = {6}, year = {2015}, pages = {2678-2690}, abstract = {

This paper describes the mechatronic design and characterization of a novel MR-compatible actuation system designed for a parallel force-feedback exoskeleton for measurement and/or assistance of wrist pointing movements during functional neuroimaging. The developed actuator is based on the interposition of custom compliant elements in series between a non-backdrivable MR-compatible ultrasonic piezoelectric motor and the actuator output. The inclusion of physical compliance allows estimation of interaction force, enabling force-feedback control and stable rendering of a wide range of haptic environments during continuous scanning. Through accurate inner-loop

velocity compensation and force-feedback control, the actuator is capable of displaying both a low-impedance, subject-in-charge mode, and a high stiffness mode. These modes enable the execution of shared haptic protocols during continuous fMRI.\ 

The detailed experimental characterization of the actuation system is presented, including a backdrivability analysis, demonstrating an achievable impedance range of 22 dB, within a bandwidth of 4 Hz (for low stiffness). The stiffness control bandwidth depends on the specific value of stiffness: a bandwidth of 4 Hz is achieved at low stiffness (10\% of the physical springs stiffness), while 8 Hz is demonstrated at higher stiffness. Moreover, coupled stability is demonstrated also for stiffness values substantially (25\%) higher than the physical stiffness of the spring. Finally, compatibility tests conducted in a 3T scanner are presented, validating the potential of inclusion of the actuator in an exoskeleton system for support of wrist movements during continuous MR scanning, without significant reduction in image quality.

}, keywords = {compliant actuators., Force control, functional MRI (fMRI), MR-compatible robotics}, doi = {10.1109/TMECH.2015.2389222}, attachments = {https://mahilab.rice.edu/sites/default/files/publications/MR-compatible_actuator_v3.pdf}, author = {Fabrizio Sergi and Andrew Erwin and Marcia K. O{\textquoteright}Malley} }