Elsevier

Gait & Posture

Volume 33, Issue 4, April 2011, Pages 645-650
Gait & Posture

Foot bone kinematics as measured in a cadaveric robotic gait simulator

https://doi.org/10.1016/j.gaitpost.2011.02.011Get rights and content

Abstract

The bony motion of the foot during the stance phase of gait is useful to further our understanding of joint function, disease etiology, injury prevention and surgical intervention. In this study, we used a 10-segment in vitro foot model with anatomical coordinate systems and a robotic gait simulator (RGS) to measure the kinematics of the tibia, talus, calcaneus, cuboid, navicular, medial cuneiform, first metatarsal, hallux, third metatarsal, and fifth metatarsal from six cadaveric feet. The RGS accurately reproduced in vivo vertical ground reaction force (5.9% body weight RMS error) and tibia to ground kinematics. The kinematic data from the foot model generally agree with invasive in vivo descriptions of bony motion and provides the most realistic description of bony motion currently available for an in vitro model. These data help to clarify the function of several joints that are difficult to study in vivo; for example, the combined range of motion of the talonavicular, naviculocuneiform, metatarsocuneiform joints provided more sagittal plane mobility (27.4°) than the talotibial joint alone (23.2°). Additionally, the anatomical coordinate systems made it easier to meaningfully determine bone-to-bone motion, describing uniplanar motion as rotation about a single axis rather than about three. The data provided in this study allow for many kinematic interpretations to be made about dynamic foot bone motion, and the methodology presents a means to explore many invasive foot biomechanics questions under near-physiologic conditions.

Introduction

An accurate description of the bony motion of the foot during normal gait can aid in injury prevention, identification of foot abnormalities, surgical correction, and implant design. While in vivo foot models with skin-mounted retro-reflective markers are commonly used to describe foot kinematics [1], [2], [3], [4], these models suffer from inaccuracies related to skin-motion artifact and rigid body assumptions. Movement of tarsal and midtarsal bones such as the talus, navicular, cuboid, and cuneiforms is often lumped together or ignored since individual bones are very difficult to access; however, these bones have been shown to have significant individual movement in normal feet [5], [6], [7]. For example, Nester et al. [5] demonstrated greater motion between the cuneiforms and navicular and between the cuneiforms and cuboid than previously reported in the literature. Making rigid body assumptions within the foot can lead to inaccurate descriptions of the specific joints where motion occurs. Although in vivo models using temporary, surgically placed bone pins have provided valuable foot kinematic data [6], [7], the highly invasive procedure greatly restricts the use and application of these models with living subjects. Bone pins used in cadaveric models that accurately simulate gait would reduce the need for invasive studies.

In vitro models used with cadaveric feet allow for invasive techniques and access to individual bones of the foot. The main limitation of in vitro models is their inability to accurately reproduce physiologic gait. To address this issue, dynamic gait simulators have been used in conjunction with in vitro models [5], [8], [9]. While these simulators are valuable tools in understanding foot bony motion, their accuracy has been affected by the following: non-physiologic ground reaction forces (GRFs) [5], simplified tibial kinematics [5], [8], [9], low velocity of simulation [8], [9], low vertical GRF (vGRF) magnitude [5], [8], [9], exclusion of bones [9], and technical, rather than anatomical, based coordinate systems [5], [8], [9]. Our group has developed a cadaveric gait simulator (i.e., the robotic gait simulator or RGS) that has begun to address these issues [10], [11], [12], with the intent to provide a more accurate and realistic description of foot kinematics during walking. The aim of this work was to provide a description of the bony motion of the foot during gait and present a methodology (the RGS) that addresses many of the limitations associated with dynamic in vitro foot and ankle models of gait.

Section snippets

Methods

For this IRB-approved study, six fresh-frozen cadaveric lower limb specimens (age: 75.8 ± 6.1 yrs, 5 male, 1 female, two pairs, two unpaired) with neutral bony alignment, transected approximately 12 cm proximal to the ankle joint, were identified by an orthopedic surgeon using X-rays. Approximately 10 cm of each of nine extrinsic ankle tendons were dissected and attached to aluminum or plastic tendon clamps. After applying the foot model marker set described below, the cadaveric specimen was mounted

Results

The fuzzy logic vGRF controller demonstrated its ability to track the target vGRF with high fidelity for six cadaveric specimens [12]. The average RMS error between the target in vivo and actual in vitro vGRF was 5.9% body weight (BW) across all 18 final trials. The RGS was able to replicate the in vivo kinematics of TIB with respect to GND. All three fixed angles of the tibia with respect to the ground were almost entirely within ±1 SD of those found in vivo for all feet except for the sagittal

Discussion

Quantifying the bony motion of the foot and ankle during gait has the potential to advance our understanding of its normal and pathologic function and assist clinical decision making. In general, we found good agreement between our kinematic data and data from in vivo studies, especially Lundgren et al.’s study employing bone pins [6], which will be considered the gold standard for in vivo bony motion during gait. The total ROM reported here was within ±1 SD of the data reported by Lundgren et

Conflict of interest statement

The authors have no conflicts to report.

Acknowledgements

This work was funded in part by the Department of Veterans Affairs Rehabilitation Research and Development Service grants A3923R and A6669R.

References (22)

  • C. Hurschler et al.

    In vitro simulation of stance phase gait – part I: model verification

    Foot Ankle Int

    (2003)
  • Cited by (60)

    • Naviculocuneiform Arthrodesis for Treatment of Adult-Acquired Flatfoot Deformity

      2023, Clinics in Podiatric Medicine and Surgery
      Citation Excerpt :

      There is evidence that the range of motion obtained through the NC joint is likely underappreciated, bringing into question the theory of the joint being nonessential. Multiple studies have evaluated a sagittal plane range of motion of the medial column, and it has been found that the NC joint not only provides 50% of the total sagittal plane range of motion, but the NC joint also shows a 12.2° of sagittal plane range of motion, which is significantly more than both the TN and TMT joints.11,12 As noted, the primary plane of motion of the NC joint is the sagittal plane, although some frontal plane range of motion does occur as it was found on weight-bearing CT analysis that the medial cuneiform pronates 6.1° relative to the navicular at the NC joint.

    • Does temporary bridge plate fixation preserve joint motion after an unstable Lisfranc injury?

      2023, Foot and Ankle Surgery
      Citation Excerpt :

      Its kinematic function is generally described as nonessential, being mainly for stability. Yet, biomechanical studies have shown a mean total range of motion of 5.3° [28,31]. Temporary bridge plate fixation is in theory a motion-preserving osteosynthesis.

    • Biomechanical evaluation of naviculocuneiform fixation with lag screw and locking plates

      2021, Foot and Ankle Surgery
      Citation Excerpt :

      Precise anatomic and biomechanical knowledge is the basis for successful foot and ankle surgery [1–3].

    • Finite element analysis of secondary effect of midfoot fusions on the spring ligament in the management of adult acquired flatfoot

      2020, Clinical Biomechanics
      Citation Excerpt :

      Furthermore, long term clinical data regarding the effects of medial column arthrodesis on the Spring ligament have not been established. Whittaker et al. (2011) have reported previous studies using cadaver modeling to evaluate the relative contribution of the various joints of the foot to overall foot dorsiflexion. The NC joint is responsible for most of the dorsiflexion of the foot and at least as much as the TN joint.

    View all citing articles on Scopus
    View full text