Cosimulation of the index finger extensor apparatus with finite element and musculoskeletal models

Abstract

Musculoskeletal modeling has been effective for simulating dexterity and exploring the consequences of disability. While previous approaches have examined motor function using multibody dynamics, existing musculoskeletal models of the hand and fingers have difficulty simulating soft tissue such as the extensor mechanism of the fingers, which remains underexplored. To investigate the extensor mechanism and its impact on finger motor function, we developed a finite element model of the index finger extensor mechanism and a cosimulation method that combines the finite element model with a multibody dynamic model. The finite element model and cosimulation were validated through comparison with experimentally derived tissue strains and fingertip endpoint forces respectively. Tissue strains predicted by the finite element model were consistent with the experimentally observed strains of the 9 postures tested in cadaver specimens. Fingertip endpoint forces predicted using the cosimulation were well aligned in both force (difference within 0.60 N) and direction (difference within 30°with experimental results. Sensitivity of the extensor mechanism to changes in modulus and adhesion configuration were evaluated for ± 50% of experimental moduli, presence of the radial and ulnar adhesions, and joint capsule. Simulated strains and endpoint forces were found to be minimally sensitive to alterations in moduli and adhesions. These results are promising and demonstrate the ability of the cosimulation to predict global behavior of the extensor mechanism, while enabling measurement of stresses and strains within the structure itself. This model could be used in the future to predict the outcomes for different surgical repairs of the extensor mechanism.

Figure 1.

Cosimulation framework including a) rigid body dynamic modeling and b) finite element modeling. A custom MATLAB program (center) passes information between the two simulation platforms.

Figure 2.

Effective strain of the extensor mechanism (c.f. Posture 5, Table 2), with (a) 11.8 N load on the EDC tendon and (b) no load applied. Bars indicate measured strain region for CS (black) and TS (gray) (as in Equation 1).

Figure 3.

Simulated and experimental CS and TS strains across all postures. Error bars present on experimental strains (dark grey) represent ±1 standard deviation. Error bars on simulated strains (light grey) indicate maximum and minimum simulated range.

Figure 4.

Simulated fingertip contact force magnitudes for both the multibody dynamic model and cosimulation approach, in comparison to experimentally reported values (Qiu, 2014). Error bars for the experimentally reported values (black) indicate ±1 standard deviation. Error bars on cosimulation values (light grey) indicate maximum and minimum simulated range.

Figure 5.

Simulated fingertip force vectors for the EDC, EI, FPI, and LUM muscles in the sagittal plane for the multibody dynamic model, experimentally reported results (Qiu, 2014), and the cosimulation model. Error boundaries for experimental results indicate ±1 standard deviation. Error boxes present for the cosimulation represent maximum and minimum simulated range.

Figure 6.

Simulated fingertip force vectors for the EDC, EI, FPI, and LUM muscles in the sagittal plane in different adhesion configurations in comparison to experimentally reported results (Qiu, 2014). The abbreviations RU refer to the radial and ulnar adhesions present on the proximal phalange, and JC refers to the joint capsule present about the central slip.