Forward dynamic optimization of handle path and muscle activity for handle based isokinetic wheelchair propulsion: A simulation study

Abstract

Push-rim wheelchair propulsion is biomechanically inefficient and physiologically stressful to the musculoskeletal structure of human body. This study focuses to obtain a new, optimized propulsion shape for wheelchair users, which is within the ergonomic ranges of joint motion, thus reducing the probability of injuries. To identify the propulsion movement, forward dynamic optimization was performed on a 3D human musculoskeletal model linked to a handle based propulsion mechanism, having shape and muscle excitations as optimization variables. The optimization resulted in a handle path shape with a circularity ratio of 0.95, and produced a net propulsion power of 34.7 watts for an isokinetic propulsion cycle at 50 rpm. Compared to push-rim propulsion, the compact design of the new propulsion mechanism along with the ergonomically optimized propulsion shape may help to reduce the risk of injuries and thus improve the quality of life for wheelchair users.

Keywords: Wheelchair propulsion; dynamic optimization; handle-based propulsion; musculoskeletal modelling.

Figure 1.

(a) Musculoskeletal model with right hand linked to the propulsion mechanism, with the 15 muscle actuators, Delt1(AnteriorDeltoid), Delt2(MiddleDeltoid), Delt3(PosteriorDeltoid), BicLong(BicepsLong), BicShort(BicepsShort), TriLong(TricepsLong), TriLat(TricepsLateral), Bracs(Brachialis), FCR(FlexorCarpiRadialis),FCU(FlexorCarpiUlnaris),PecM(Pectoralis Major) and Rotator cuff muscles(Supraspinatus (SUPRA), subscapularis (SUBSC), infraspinatus (INFRA), teresminor (TMIN)) with the major DOF such as Elevation Plane, Elevation angle, Elbow flexion, Shoulder rotation, Wrist deviation and flexion. (b) Kinematic components of the propulsion mechanism, with major DOF such as crank angle (ɵ), effective crank length (C_L), tilt angle (β) and handle angle (ɳ).

Figure 2.

Dynamically optimized propulsion path with the centre C for HBP in the parasagittal plane defined by the wheels.

Figure 3.

Optimized muscle activity patterns (only muscles for which a comparison to push rim propulsion is available), with the dark solid lines (muscle activations) and the dotted lines (muscle excitations) over one full propulsion cycle. The shaded regions indicate the phases in which the respective muscles were active during push-rim propulsion (Mulroy et al. 1996). The shaded bars below the diagrams show the propulsion zones.

Figure 4.

Net work done by the upper limb muscles (in joules) during the four zones of propulsion with the optimized handle path.

Figure 5.

Comparison of the joint ranges of motion (ROM) of the upper extremity between push-rim propulsion and the HBP (handle-based propulsion) technique (Rankin et al. ; Morrow et al. 2014). Shoulder rotation Int/Ext(+/−).Wrist deviation Ulna/Radial(+/−) and wrist flexion Ext/Flex(−/+). The physiological range represents the anatomical joint range. The figures on the right side indicate the joint motion during HBP.

Figure 6.

Normalized fiber length and normalized fiber shortening velocity of the muscles spanning the elbow joint during one full crank rotation. The dotted lines represent the optimal fiber length (L_opt). In the normalized fiber velocity graph the muscle contraction is negative and muscle lengthening is positive. The maximum shortening velocity of each muscle was assumed to be 10 optimal fiber lengths per second (V_max = 10 L_opt s⁻¹). The dark shaded areas in the graphs represent the regions with more that 20 percent of muscle activation.

Figure 7.

Peak muscle forces obtained from the computational simulation of HBP compared to dynamic optimization results for push rim propulsion (Morrow et al. 2014).