0
Research Papers

# Freebal: Design of a Dedicated Weight-Support System for Upper-Extremity Rehabilitation

[+] Author and Article Information
Arno H. A. Stienen

Biomechanical Engineering, University of Twente, Enschede 7500 AE, The Netherlands; Research Associate of Physical Therapy and Human Movement Sciences, Northwestern University, Chicago, IL 60611arnostienen@gmail.com

Edsko E. G. Hekman

Biomechanical Engineering, University of Twente, Enschede 7500 AE, The Netherlands

Gerdienke B. Prange, Michiel J. A. Jannink

Roessingh Research and Development, Enschede 7500 AH, The Netherlands

Frans C. T. van der Helm

Biomechanical Engineering, Delft University of Technology, Delft 2628 CD, The Netherlands

Herman van der Kooij

Biomechanical Engineering, University of Twente, Enschede 7500 AE, The Netherlands; Biomechanical Engineering, Delft University of Technology, Delft 2628 CD, The Netherlands

J. Med. Devices 3(4), 041009 (Dec 09, 2009) (9 pages) doi:10.1115/1.4000493 History: Received October 29, 2008; Revised October 10, 2009; Published December 09, 2009; Online December 09, 2009

## Abstract

Most rehabilitation devices for the upper extremities include a weight-support system. In recent publications, weight support is shown to be effective for stroke rehabilitation. But current devices are often complex, have significant movement inertia, and/or limit the movement range. The goal of this study is to improve on current designs by introducing a novel, dedicated weight-support device, the Freebal. This passive mechanical device uses balanced spring mechanisms for constant-but-scalable forces to support the arm. It has a large workspace of roughly $1 m3$, low movement impedance, and independent support at the elbow and wrist of up to 5 kg. An explorative cross-sectional study with eight patients shows the Freebal to instantly extend the range of motion of the affected arm by 7%. In conclusion, most requirements are met for patients to benefit from therapy with the Freebal, potentially progressing earlier to more motivating, functional training.

<>
Copyright © 2009 by American Society of Mechanical Engineers
Your Session has timed out. Please sign back in to continue.

## Figures

Figure 1

The Freebal. The system generates the weight supporting forces with almost inertia-free balanced spring mechanisms (see Fig. 2). The wrist and elbow are supported by two slings connected via cabling to these independent mechanisms. In the figure, the overhanging beam is shown lowered for display reasons. During normal use, it can extend up to 3.5 m above ground level dependent on available space and the work volume needed.

Figure 2

The balanced spring mechanism, in theory (a) and as found in the final design (b). The Freebal has two of these mechanisms, connected to the wrist and elbow. The supporting force Fc,b at the beam end is independent of the angle β, because the vertical component of the spring force Fsp,z is always equal to distance A times spring-stiffness k (see Eq. 1). The support can be adjusted by changing the spring-attachment distance R1. Cable angles and friction influence the effective weight support at the sling Fc,s (see Figs.  34). To get the zero-length spring behavior with stock extension springs, the spring is placed in the vertical tube and connected via a cable (see inset).

Figure 3

Effective weight support as measured in the balanced spring mechanism (see Fig. 2). With the arm moving down, and thus the beam end moving up, the mean effective amount of support is given by the dark, solid line (thin lines are the standard deviation). The light, stripped line gives the amount of support with the arm moving up. The effective support differs from the desired force output line due to friction of the spring cable running over the small (10 mm diameter) guiding pulley (see Fig. 2). When measured, the friction was 12% of the weight support, although this was not felt manually.

Figure 4

Cross-sectional view of the theoretical nonlinearities caused by cable angles and spring mechanism. The work volume (1 m diameter) is 1 m above ground level beneath the foremost top pulley. The Freebal is extended to a height lt of 3.5 m. For Eq. 3: cable length la, vertical length lv, and sling-cable angle θ. In the top figure, the effective weight support Fc,s is given as percentage of the vertical beam force Fc,b. In the bottom figure, the horizontal inward directed forces at the sling Fh,s are given as percentage of Fc,b.

Figure 5

Experimental setup, with the Freebal connected to the wrist and elbow. On the arm the optical markers from the Vicon system are visible. Four circular dots of 0.1 m diameter on the table assist in the task execution; a base dot, from were all movements start, and the three target dots in medial, central, and lateral directions. The base dot is located directly under the hand in front of the elbow when the subject has 90 deg shoulder plane of elevation, 0 deg of both shoulder elevation and humerus long axis rotation, and 90 deg of elbow flexion. The target dots are 0.35 m from the base dot, with the medial and lateral dots at 45 deg angles to the base to central dot axis.

Figure 6

Typical circular range-of-motion wrist paths. In top figure (a) the corrected wrist paths of a typical healthy subject are shown, and in top figure (b) of a typical stroke patient. The wrist paths are corrected for shoulder translation by subtracting the shoulder positions at each time step. The finger to wrist length is accounted for by repositioning the base dot underneath the wrist at the start of each trial. The gray solid lines are without weight support, each consisting of five repetitions, and the black stripped lines are with full weight support. In the latter analysis, only the wrist paths inside the triangle of dots (shown with a gradient) are used.

Figure 7

Average normalized maximum shoulder-wrist distances. Left three groups: maximum stretch trials in the medial, central, and lateral directions. Right two groups: range-of-motion trials in clockwise and counterclockwise directions. Only the results grouped over all directions were significantly different between the supported and unsupported arms of stroke subjects ( ∗∗=p<0.001).

Figure 8

Freebal supporting the arm during interactive sessions. Interacting with a virtual environment or playing games has been used by many rehabilitation devices as a way to increase the patient’s motivation.

Figure 9

A possibly better spring cable guiding mechanism. By using three pulleys (Po, Pr, and Pa) of equal diameter, though larger (40 mm) as compared with the original Pa pulley (10 mm), the friction is removed. Lowering the spring cable tension by using a slacker spring and increasing the A and/or R1 distances further reduce the friction, although this needs more space for spring deflection and longer springs. The amount of support cannot be adjusted by changing R1 anymore, as the change in cable length between Po and Pa negates the balanced spring mechanism. Instead, A now needs to be lowered or heightened, together with the spring and spring tube.

Figure 10

Alternative weight-support mechanism. Compare with the Freebal mechanism in Fig. 2. The spring beam is now split, eliminating the (small) nonlinearities of the beam-end point horizontal translations. Furthermore, the cabling beam is vertically hinged roughly above the human shoulder and has a vertical slider underneath the cable beam. This can position the vertical cable exactly above the wrist and elbow and reduce the nonlinearity due to the angles of the cable with the vertical. However, both changes make the mechanisms more complex and more susceptible to friction and undesirable dynamics. For example, the hinged beam may swing with frequencies close to the eigenfrequencies.

## Discussions

Some tools below are only available to our subscribers or users with an online account.

### Related Content

Customize your page view by dragging and repositioning the boxes below.

Topic Collections