An assistive lower limb exoskeleton for people with neurological gait disorders.

Lower limb exoskeletons have already proven the capability to give back mobility to people suffering from spinal cord injury (SCI). Other important populations such as people with multiple sclerosis or muscular dystrophy, frail elderly and stroke victims, suffer from severe gait impairments and could benefit from similar technology. The work presented in the current paper describes a novel design of a 6-actuated degrees of freedom (DOFs) assistive lower limb exoskeleton for people with moderate mobility impairments. The electrical actuators are all remotely located on the back of the user for a more compact design with high dynamics. Cable driven solutions are used to transmit the flexion/extension of the hip and knee joints, while a powerful ballscrew carries out the hip adduction/abduction. The design of this exoskeleton, named AUTONOMYO, follows the key specifications of being highly back-drivable and able to perform dynamic motions at low energy consumption. AUTONOMYO is capable to assist the user's balance by providing complementary torques at the hip and the knee. Results show that the projected level of assistance for sit-to-stand transition varies from 50% to 100% in function of the user's bodyweight and height while higher level of assistance are reached for walking and stairs climbing activities.

Visuo-tactile integration and body ownership during self-generated action.

Although there is increasing knowledge about how visual and tactile cues from the hands are integrated, little is known about how self-generated hand movements affect such multisensory integration. Visuo-tactile integration often occurs under highly dynamic conditions requiring sensorimotor updating. Here, we quantified visuo-tactile integration by measuring cross-modal congruency effects (CCEs) in different bimanual hand movement conditions with the use of a robotic platform. We found that classical CCEs also occurred during bimanual self-generated hand movements, and that such movements lowered the magnitude of visuo-tactile CCEs as compared to static conditions. Visuo-tactile integration, body ownership and the sense of agency were decreased by adding a temporal visuo-motor delay between hand movements and visual feedback. These data show that visual stimuli interfere less with the perception of tactile stimuli during movement than during static conditions, especially when decoupled from predictive motor information. The results suggest that current models of visuo-tactile integration need to be extended to account for multisensory integration in dynamic conditions.

Design and Evaluation of a Novel Haptic Interface for Endoscopic Simulation.

Inspection of the colon with an endoscope for early signs of cancer (colonoscopy) has become an extremely widespread procedure, since early treatment radically improves the outlook of patients. The procedure requires a close coordination between the sense of touch and vision to navigate the endoscope along the colon. This raises the need to develop efficient training methods for physicians. Training simulators based on virtual reality, where realistic graphics are combined with a mechatronic system providing haptic feedback, are alternative to traditional training methods. To provide physicians with realistic haptic sensations of an endoscopic procedure, we have designed a haptic interface, instrumented a clinical endoscope and combined them with a simulation software for colonoscopy. In this contribution, we present the mechatronic components of the simulator. The haptic interface is able to generate high forces using the combination of electrical motors and brakes in a compact design. Experiments were performed to determine the characteristics of the device. A model-based control has been implemented and the results show that the control successfully compensates for the device nonlinearities, such as friction. The proposed haptic interface, together with the virtual reality, form a highly realistic training simulator for endoscopic surgeons, applicable not only to colonoscopy, but also to similar interventions.

A computer-based real-time simulation of interventional radiology.

Interventional radiology is a minimally invasive procedure where thin instruments, guidewires and catheters or stents are steered through the patient's vascular system under X-ray imaging for treatment of vascular diseases. The complexity of these procedures makes training in order to master hand-eye coordination, instrument manipulation and procedure protocols for each radiologist mandatory. In this paper, we present a computer-based real-time simulation of interventional radiology procedures, which deploys a very efficient physics-based thread model to simulate the elastic behavior of guidewires and catheters. A fast collision detection scheme provides continuous collision response, which reveals more details of arterial walls than a centerline approach. Furthermore rendering techniques for realistic X-ray effect have been implemented. Our simulation structure is updated at a haptic rate of 500 Hz, thus contributing to physical realism.

An electrostatic glass actuator for ultrahigh vacuum: A rotating light trap for continuous beams of laser-cooled atoms.

This article describes the design, characterization, and performance of an electrostatic glass actuator adapted to an ultrahigh vacuum environment (10(-8) mbar). The three-phase rotary motor is used to drive a turbine that acts as a velocity-selective light trap for a slow continuous beam of laser-cooled atoms. This simple, compact, and nonmagnetic device should find applications in the realm of time and frequency metrology, as well as in other areas of atomic, molecular physics and elsewhere.

Multi-joint arm movements to investigate motor control with FMRI.

Performing multi-joint arm movements in controllable dynamic environments during functional magnetic resonance imaging (fMRI) could provide important insights into the brain mechanisms involved in human motor control and related dysfunctions. In order to obtain useful data, these movements must be possible and comfortable for the subject within the narrow bore of the scanner and should not create any movement artifacts in the image. We found that commonly studied arm movements involving the shoulder create movement artifacts, and investigated alternative multijoint arm movements within a mock-up of an MR scanner. We selected movements involving the elbow and wrist joints, with an extension attached to the hand, and propose a dedicated kinematic structure using the MR compatible actuators we have previously developed.

A Haptic Interface Based on Potential Mechanical Energy to Investigate Human Motor Control using fMRI.

This paper describes a mechanical interface to use in conjunction with fMRI, in order to infer the brain mechanisms of human motor learning. Innovative mechanical concepts based on gravity and elastic forces were used to generate typical stable and unstable dynamic interactions at the hand during multijoint arm movements. Two designs were retained and implemented from MR compatible materials. The first uses a spring constrained between two specially designed surfaces and the other a capstan to transform the force induced by a groove carved on a shaft. These two degree-of-freedom mechanical interfaces have been constructed and tested. The use of a capstan mechanism was found to be limited by excessive friction, however, the method using a machined surface provides a simple and effective interface to investigate human motor control.

Knot-tying with visual and force feedback for VR laparoscopic training.

Real-time simulation of thread and knot-tying with visual and force feedback is an essential part of virtual reality laparoscopic training. This paper presents a physics-based thread simulator that enables realistic knot tying at haptic rendering rate. The virtual thread follows Newton's law and behaves naturally. The model considers main mechanical properties of the real thread such as stretching, compressing, bending and twisting, as well as contact forces due to self-collision and interaction with the environment, and the effect of gravity. The structure of the system has essential advantages over geometrically based approaches, as was illustrated in an implementation on the Xitact simulator.

Measurement of elastic properties of blood vessels.

This paper is related to the measurements of the modulus of elasticity of an artery by studying the deformations due to the inflation of an angioplasty balloon catheter used for Interventional Radiology (IR) procedures. Various types of balloons are studied in order to characterize and compare their behaviors at the time of inflation. A test bench, consisting of an angioplasty balloon, a Polyvinyl alcohol model and an actuator used to inflate a balloon, is developed for the realization of the experiments. The pressure-volume curve during the inflation of a balloon is observed. Elasticity modulus are derived with an analytical model of the measurement system. The results are then analyzed and compared to existing data from literature.

Measurement of internal constraints during an interventional radiology procedure.

This paper presents a novel method for measurements of internal constraints during an Interventional Radiology procedure. Fiber-optic strain gauge, operating as a Fabry-Perot interferometer, is inserted and navigated inside a polyvinyl alcohol (PVA) model. The information obtained from deformation of the Fabry-Perot cavity is analyzed and the results are then rescaled through a calibration procedure for catheters. The measures help to locate the maximum interaction forces between catheter and blood vessel wall at approximately 1.3 N and to observe the shape of dynamic interaction during manipulation.