This may be the author’s version of a work that was submitted/accepted for publication in the following source:

Bishop, Chris, Paul, Gunther, & Thewlis, Dominic (2013) The reliability, accuracy and minimal detectable difference of a multi- segment kinematic model of the foot-shoe complex. Gait and Posture, 37(4), pp. 552-557.

This file was downloaded from: https://eprints.qut.edu.au/49002/

c Consult author(s) regarding copyright matters

This work is covered by copyright. Unless the document is being made available under a Creative Commons Licence, you must assume that re-use is limited to personal use and that permission from the copyright owner must be obtained for all other uses. If the docu- ment is available under a Creative Commons License (or other specified license) then refer to the Licence for details of permitted re-use. It is a condition of access that users recog- nise and abide by the legal requirements associated with these rights. If you believe that this work infringes copyright please provide details by email to [email protected]

License: Creative Commons: Attribution-Noncommercial-No Derivative Works 2.5

Notice: Please note that this document may not be the Version of Record (i.e. published version) of the work. Author manuscript versions (as Sub- mitted for peer review or as Accepted for publication after peer review) can be identified by an absence of publisher branding and/or typeset appear- ance. If there is any doubt, please refer to the published source. https://doi.org/10.1016/j.gaitpost.2012.09.020

G Model

GAIPOS-3720; No. of Pages 6

Gait & Posture xxx (2012) xxx–xxx

Contents lists available at SciVerse ScienceDirect

Gait & Posture

jo urnal homepage: www.elsevier.com/locate/gaitpost

The reliability, accuracy and minimal detectable difference of a multi-segment

kinematic model of the foot–shoe complex

a, b,1 a,2

Chris Bishop *, Gunther Paul , Dominic Thewlis

a

School of Health Sciences, University of South Australia, City East Campus, North Terrace, Adelaide, SA 5000, Australia

b

School of Public Health and Social Work, QUT - Queensland University of Technology, Kelvin Grove, Queensland 4059, Australia

A R T I C L E I N F O A B S T R A C T

Article history: Kinematic models are commonly used to quantify foot and ankle kinematics, yet no marker sets or

Received 2 February 2012

models have been proven reliable or accurate when wearing shoes. Further, the minimal detectable

Received in revised form 17 May 2012

difference of a developed model is often not reported. We present a kinematic model that is reliable,

Accepted 20 September 2012

accurate and sensitive to describe the kinematics of the foot–shoe complex and lower leg during walking

gait. In order to achieve this, a new marker set was established, consisting of 25 markers applied on the

Keywords:

shoe and skin surface, which informed a four segment kinematic model of the foot–shoe complex and

Kinematics

lower leg. Three independent experiments were conducted to determine the reliability, accuracy and

Footwear

minimal detectable difference of the marker set and model. Inter-rater reliability of marker placement

Foot and ankle biomechanics

Reliability on the shoe was proven to be good to excellent (ICC = 0.75–0.98) indicating that markers could be

Accuracy applied reliably between raters. Intra-rater reliability was better for the experienced rater (ICC = 0.68–

0.99) than the inexperienced rater (ICC = 0.38–0.97). The accuracy of marker placement along each axis

was <6.7 mm for all markers studied. Minimal detectable difference (MDD90) thresholds were defined

for each joint; tibiocalcaneal joint – MDD90 = 2.17–9.368, tarsometatarsal joint – MDD90 = 1.03–9.298

and the metatarsophalangeal joint – MDD90 = 1.75–9.128. These thresholds proposed are specific for the

description of shod motion, and can be used in future research designed at comparing between different

footwear.

ß 2012 Elsevier B.V. All rights reserved.

1. Introduction possible to model a foot–shoe complex (i.e. the foot and shoe

together) rather than just a shoe or the foot in isolation. In this

Kinematic models are commonly used to quantify foot and example, the foot remains the major factor determining the

ankle kinematics during activities of daily living. However, most of kinematics, however the shoe changes the basic assumptions of

these models were developed with the intention of studying the segments.

barefoot biomechanics [1,2], when typically activities of daily While the concept of a shoe model appears relatively logical,

living are performed while wearing shoes. The translation of these there are methodological issues that must be dealt with. Firstly, one

models to the study of shod kinematics is problematic. It has been must define the segments of the model based on standard palpable

previously identified that tracking the motion of shoe-mounted anatomical landmarks. Palpation through the surface of a shoe

markers does not indicate the movement of the foot inside the shoe comes with a degree of complexity and a number of unknown

[3,4]. Although methods have been developed to apply markers on factors. Palpation guidelines, such as those documented by Van Sint

the skin surface through the shoe, the analysis of in-shoe foot Jan [6], that are commonly followed in surface marker sets were not

kinematics is not always required [5]. In this case, it is essential to developed for use through clothing or shoes. By following these

consider the segments of the shoe, which are conceptually similar methods, it is likely the application of markers on the shoe surface

to the anatomy of the foot, rather than the foot itself. When the will come with a degree of inaccuracy in regards to the location of the

structure of the shoe is added to the complexity of the foot, it is underlying anatomical landmark. This will have a direct impact on

the accuracy of anatomical frame definition. We have previously

described a method for palpating anatomical landmarks through the

shoe upper [7]. Using these guidelines we have proposed a set of

* Corresponding author. Tel.: +61 8 83022425.

offset values, which account for the difference between the shoe

E-mail addresses: [email protected] (C. Bishop),

upper and anatomical landmark [7]. The second problem encoun-

[email protected] (G. Paul), [email protected] (D. Thewlis).

1 tered is that of technical componentry and material thickness

Tel.: +61 7 31385795.

2

Tel.: +61 8 83021540. common to shoes such as rigid heel counters.

0966-6362/$ – see front matter ß 2012 Elsevier B.V. All rights reserved.

http://dx.doi.org/10.1016/j.gaitpost.2012.09.020

Please cite this article in press as: Bishop C, et al. The reliability, accuracy and minimal detectable difference of a multi-segment

kinematic model of the foot–shoe complex. Gait Posture (2012), http://dx.doi.org/10.1016/j.gaitpost.2012.09.020

G Model

GAIPOS-3720; No. of Pages 6

2 C. Bishop et al. / Gait & Posture xxx (2012) xxx–xxx

coordinate systems (LCS) were used to define each segment. Each joint was

Both single-segment and multi-segment marker based models

modelled with three degrees of rotational freedom. The joint pose was then

of the foot have been published, with demonstrated good intra-

estimated using the global optimisation technique described by Lu and O’Connor

and inter-rater reliability of marker placement and model output

[18]. Rotation of the tibiocalcaneal joint around the x-axis (medial–lateral) was

[1,2,8–13]. A recent review highlighted that the current trend is to flexion/extension, about the y-axis (anterior–posterior) was abduction/adduction

consider the foot as a number of segments [14]. This trend can be and about the z-axis (superior–inferior) was inversion/eversion. To ensure the z-

axis was aligned with the long axis of the segment, the axes of rotations changed

extrapolated to the study of footwear kinematics as shoes are

when describing the distal foot joints. Rotation around the x-axis (medial–lateral)

typically thought to consist of a heel counter, mid-sole and toe box.

remained flexion/extension, but the y-axis (superior–inferior) changed to represent

Despite demonstrated reliability, the absence of marker placement

abduction/adduction and the z-axis (anterior–posterior) changed to represent

accuracy assessment makes it difficult to interpret the accuracy of inversion/eversion. An XYZ cardan sequence was used to represent the order of

rotations for each joint [19].

anatomical frame definition relative to the underlying bone

embedded frame [15]. In addition to reporting static or dynamic

2.1.3. Anatomical landmarks

measures of marker placement reliability and accuracy, it is

A marker set consisting of 14, 10 mm retro-reflective markers and one hallux

important to determine the minimal detectable difference and

trihedron (three, 10 mm markers) was applied to the shoe surface and four, 10 mm

understand the limitations of a given kinematic model. Therefore,

markers and one plate-mounted marker cluster (four, 15 mm markers) were

the aim of the study was to demonstrate the reliability and static applied to the skin surface of the leg to define anatomical frames (Fig. 1). To apply

markers on the skin surface, previously developed guidelines were used [6]. To

and dynamic accuracy of a marker set we have previously

palpate anatomical landmarks through the shoe upper, custom guidelines were

proposed [7] (albeit to describe the offsets between the shoe

developed (Additional File 1). A technical cluster consisting of at least three non-

upper and anatomical landmark), which could be used in future

collinear markers was placed on each segment.

research investigating the effectiveness of interventions that can

only be tested under shod conditions, e.g. foot orthotics. The

2.1.4. Anatomical reference frames

minimal detectable differences of the model will be defined to The leg:

avoid over emphasis on small differences, which might otherwise

The midpoint between the MFE and LFE;

be considered type I errors. Oleg

The x-axis joints the origin and the LFE marker; Xleg

2. Methodology

The y-axis is orthogonal to the x-axis and lies in the coronal plane;

Yleg

The z-axis is orthogonal to the xy plane.

In order to define the accuracy and reliability of marker placement and the Zleg

minimal detectable difference of the model, three experiments were conducted and

are reported independently:

The heel segment:

Experiment 1 – marker placement accuracy;

 The midpoint between the MM and LM;

Oheel segment

Experiment 2 – marker placement reliability;

 The x-axis joins the origin and LM marker;

xheel segment

Experiment 3 – determination of the model’s minimal detectable difference.

 The z-axis is orthogonal to the x-axis and lies in the transverse plane;

zheel segment

The y-axis is orthogonal to the xz plane.

Each of the experiments recruited a sample of university staff and students. yheel segment

Participants were included if they were aged between 18 and 40 years old, had a

normal arched right foot [16] and reported no medical history that could adversely The midfoot segment:

affect gait. Written informed consent was obtained [17]. The Human Research

Ethics Committee of the University of South Australia approved the protocol

The midpoint between the NT and SP;

Omidfoot segment

(protocol number 0000020984).

The x-axis joins the origin and SP marker;

A basic, commercially available shoe (Mexico 66, ASICS, Japan) was used in this xmidfoot segment

The z-axis is orthogonal to the x-axis and lies in the transverse study. This shoe is a straight-lasted, enclosed leather shoe with a midsole and

zmidfoot segment

plane;

outsole, yet no rigid heel counter. Shoe sizes were measured using a Brannock

The y-axis is orthogonal to the xz plane. device (The Brannock Device Co. Inc, USA) and fitted by a qualified podiatrist. All

ymidfoot segment

shoelaces were tightened to a self-reported comfortable fit. A Shimadzu computed

radiography machine (Shimadzu, Japan) was used to take all weight-bearing X-rays.

The toe box segment:

To capture kinematic data, a 12 camera Vicon MX-F20 system (Vicon, Oxford, UK)

was used. All kinematic data were sampled at 100 Hz. The use of each of these

The midpoint between the 1MTH and 2MTH;

O

instruments/systems is expanded further in the following sections. toe box segment

The x-axis joins the origin and the 2MTH marker;

xtoe box segment

The z-axis is orthogonal to the x-axis and lies in the transverse

2.1. Kinematic model ztoe box segment

plane;

2.1.1. Model segments The y-axis is orthogonal to the xz plane.

ytoe box segment

A four-segment model of the foot–shoe complex and lower leg was developed.

The four segments were assumed to act as rigid bodies and were defined as:

2.2. Experiment 1: reliability tests

 leg (consisting of the tibia and fibula); Twelve participants (mean age of 21.3 years [95% CI = 20.7–24.0 years], height of

1.77 m [95% CI = 1.74–1.85 m], Euro shoe size 41.7 (95% CI = 40.1–43.2) and body

 heel (consisting of the hindfoot of the foot and the rear of the shoe);

mass of 73.0 kg [95% CI = 65.9–81.8 kg]) were recruited. Each participant attended

 midfoot (consisting of the midfoot and metatarsals 1–5 of the foot and the middle

two data collection sessions one week apart and was provided with shoes. Two

third of the shoe);

raters were used to investigate intra- and inter-rater reliability. The raters were

 toe box (consisting of toes 1–5 and the toe box).

classified as experienced (>5 years) and inexperienced (<5 years) based on the

number of years of experience in musculoskeletal science. Prior to data collection,

2.1.2. Model joints an independent rater affixed three markers on the rear sole of the shoe to define a

The articulation between two segments defined a joint: shoe coordinate system (SCS). The markers defining the SCS were not removed

between sessions. The marker set was applied to each participant and a static trial

was captured to define the position of each marker in the global coordinate system

 tibiocalcaneal joint was defined as the articulation between the leg and the heel

segment; (GCS) relative to the origin of the SCS. The markers were removed and the process

repeated for the second rater. Each rater was blinded to the positioning of the

 tarsometatarsal joint was defined as the articulation between the heel segment

previous rater. To define inter-rater reliability, participants returned one week later

and the midfoot segment;

and both raters repeated the process. To describe linear displacement, the distance

 metatarsophalangeal joint was defined as the articulation between the midfoot

between each marker and the origin of the SCS was calculated for each marker in

segment and the toe box segment.

each plane. The Euclidean distance of each marker from the origin of the SCS was

The median position between the proximal medial and lateral markers defining then calculated. Intra-class correlation coefficients were used as the measure of

the segment was defined as a joint centre. Right-handed, bone-embedded local intra- and inter-rater reliability [20].

Please cite this article in press as: Bishop C, et al. The reliability, accuracy and minimal detectable difference of a multi-segment

kinematic model of the foot–shoe complex. Gait Posture (2012), http://dx.doi.org/10.1016/j.gaitpost.2012.09.020

G Model

GAIPOS-3720; No. of Pages 6

C. Bishop et al. / Gait & Posture xxx (2012) xxx–xxx 3

and anterior–posterior weight-bearing X-ray was taken of the right foot.

The marker set was removed from the skin, the shoe applied and the

marker set reapplied on the shoe. A second series of X-rays was taken of

the foot–shoe complex. Marker placement accuracy was defined using the

method described by Bishop et al. [7]. The mean displacement between skin-

mounted and shoe-mounted markers from the anatomical landmark they were

purported to represent was considered a measure of marker placement accuracy

on the shoe.

2.4. Experiment 3: determination of the minimal detectable difference

Twenty-four participants (12 males and 12 females) with a mean age of

22.6 years (95% CI = 20.0–23.6 years), height of 1.76 m (95% CI = 1.71–1.80 m),

Euro shoe size 41.4 (95% CI = 40.3–42.5) and body mass of 72.5 kg (95% CI = 65.5–

76.4 kg) were recruited. The marker set was applied to the foot–shoe complex

and leg by the experienced rater. A static reference trial was captured to define

the static pose of the segments. Each participant was required to perform five

walking trials along an 8-m walkway. Marker trajectory data were captured,

labelled and tracked in Vicon Nexus (Vicon, UK, Version 1.6). After tracking and

labelling of markers, all data were saved to C3D format and imported to Visual3D

(C-Motion, Inc., MA, USA) for post-processing. All kinematic data were filtered

using a lowpass 4th order recursive Butterworth filter with a cut-off frequency of

7 Hz [21]. Initial contact and toe off were defined based on the velocity of foot

marker trajectories [22]. Additional gait events were created at 15% stance, 50%

stance and the time of peak toe box dorsiflexion (PHD). Joint angles were

extracted at each gait event for each of the three joints. Inter-rater ICCs were

used as the reliability coefficient to calculate the standard error of measurement

(SEM) and from these the minimal detectable difference (MDD90) of the model

was calculated [23].

3. Results

3.1. Experiment 1: reliability

The intra-rater reliability of applying markers on the skin and

shoe was consistently better and the ICC range less for the

experienced rater (ICC = 0.70–0.99) than the inexperienced rater

(ICC = 0.38–0.97, see Table 1). Apart from one marker (inexperi-

enced rater – medial femoral epicondyle [mean-difference

(MD) = 12.63 mm]) the mean-difference in marker placement

between sessions was <8.0 mm. Inter-rater reliability resulted in

ICC values >0.75 for all markers (ICC = 0.75–0.98). When describ-

ing marker placement error by plane, the largest differences in

marker placement between raters were identified in the coronal

plane (MD = 13.0–20.8 mm), followed by the sagittal (MD = 10.1–

14.3 mm) and transverse planes (MD = 0.9–7.5 mm).

3.2. Experiment 2: accuracy

The 2-D accuracy of marker placement on the medial–lateral

and anterior–posterior X-ray images are presented in detail in

Bishop et al. [7], with the results indicating all markers were placed

on the shoe <5 mm further away from the anatomical landmark

relative to the marker placed on the skin [7]. Table 1 reports marker

placement accuracy along each axis. On the medial–lateral and

anterior–posterior axis, all markers were placed 6.7 mm of the

Fig. 1. Foot–shoe complex marker set. The foot–shoe complex depicting the

anatomical landmarks and corresponding anatomical reference frames. The anatomical landmark. Despite the presence of a toe box on the

landmarks are defined as the medial malleolus (MM), medial malleolus medial–lateral X-ray view, all markers were placed 5.6 mm on

projection (C1), postero-dorsal heel (C2), postero-plantar heel (C3), lateral

the dorso-plantar axis.

malleolus projection (C4), lateral malleolus (LM), styloid process (SP), dorso-

lateral 4th metatarsal shaft (4MTS), 5th metatarsal head (5MTH), dorsal 2nd

3.3. Experiment 3: determination of the model’s minimal detectable

metatarsal head (2MTH), apex 2nd toe (A2), apex 1st toe (A1), hallux trihedron (HT),

1st metatarsal head (1MTH), 1st metatarsal shaft (1MTS), medial cuneiform (MC) difference

and navicular tuberosity (NT). Palpation guidelines are provided in Additional File 1.

The mean joint angles for the population are presented in Fig. 2.

The SEM and MDD90 values are also presented for each degree of

freedom of the three joints (Table 2). The MDD90 threshold values

2.3. Experiment 2: accuracy tests

ranged from 2.178 to 9.368 when estimating tibiocalcaneal joint

Twenty participants with a mean age of 21.2 years (95% CI = 19.5–22.8 years),

kinematics, from 1.038 to 9.298 when estimating tarsometatarsal

height of 1.74 m (95% CI = 1.70–1.79 m), Euro shoe size 41.4 (95% CI = 40.2–42.6)

joint kinematics and from 1.758 to 9.128 when estimating

and body mass of 69.6 kg (95% CI = 63.0–76.1 kg) were recruited. The marker

set was applied to the right foot by the experienced rater and a medial–lateral metatarsophalangeal joint kinematics.

Please cite this article in press as: Bishop C, et al. The reliability, accuracy and minimal detectable difference of a multi-segment

kinematic model of the foot–shoe complex. Gait Posture (2012), http://dx.doi.org/10.1016/j.gaitpost.2012.09.020

G Model

GAIPOS-3720; No. of Pages 6

4 C. Bishop et al. / Gait & Posture xxx (2012) xxx–xxx

Table 1

Reliability and accuracy of marker placement on the leg and shoe. All distance measures are provided in [mm]. Reliability is presented as the mean-difference (MD) in marker

placement both within and between raters. 95% confidence intervals of the MD and ICC statistic are also presented. Accuracy is presented as the MD in marker placement on

the shoe relative to the underlying anatomical landmark along the medial–lateral, anterior–posterior and dorsal–plantar axes. 95% confidence intervals of the MD are

reported as a measure of variability.

Marker Experiment 1 – reliability of marker placement Experiment 2 – marker placement accuracy

Intra-rater 1 Intra-rater 2 Inter-rater Medial–lateral Anterior–posterior Dorsal–plantar

MD 95% CI ICC MD 95% CI ICC MD 95% CI ICC Mean 95% CI Mean 95% CI Mean 95% CI

LFE 7.6 1.0–14.3 0.95 4.4 1.6–7.2 0.95 3.3 0.6–5.9 0.97 – – – – – –

MFE 2.8 1.6–4.0 0.99 12.6 5.9–19.4 0.76 6.9 2.7–11.1 0.94 – – – – – –

LM 2.5 0.1–4.9 0.70 3.4 1.8–5.0 0.58 2.1 0.7–3.5 0.83 – – – – – –

MM 1.2 0.4–2.0 0.98 2.4 0.8–3.9 0.87 2.4 1.3–3.5 0.95 – – – – – –

SP 2.3 0.9–3.6 0.91 2.9 1.3–4.6 0.89 3.7 1.9–5.5 0.84 6.5 4.1–9.0 3.2 1.6–4.7 À0.1 À2.9 to 2.8

NT 1.9 0.3–3.5 0.95 4.5 1.0–8.1 0.59 4.6 1.2–8.1 0.75 4.2 2.4–5.9 2.4 0.9–4.0 0.2 À2.9 to 3.2

5MTP 3.3 0.9–5.7 0.89 3.3 1.4–5.2 0.86 3.6 1.4–5.7 0.87 5.7 3.5–7.9 1.0 À0.7 to 2.7 0.0 À2.8 to 2.8

1MTP 3.7 1.3–6.2 0.92 4.0 2.0–5.9 0.92 3.4 1.7–5.1 0.94 2.9 1.8–4.0 1.5 4.0–2.4 À0.2 À1.7 to 1.3

2MTP 3.5 0.6–6.4 0.93 7.9 2.9–13.0 0.38 13.1 7.9–18.4 0.91 0.3 À1.8 to 2.4 4.1 1.4–6.9 5.6 3.5–7.8

Apex 2 4.1 À1.8 to 10.1 0.81 2.5 0.5–4.5 0.97 3.2 1.6–4.9 0.98 6.7 3.7–9.6 6.6 4.1–9.1 4.5 2.1–6.9

Apex 1 2.1 0.6–3.6 0.99 4.1 1.5–6.8 0.93 2.8 1.5–4.0 0.98 À1.2 À2.8 to 0.4 3.6 1.7–5.5 3.8 1.8–5.7

(–) Data were not extracted in an individual plane on a chosen X-ray image (see Experiment 2).

4. Discussion Based on the accuracy and reliability results, we propose a series

of thresholds that describe the ability of the model to describe a true

The first aim of this study was to establish a marker set of the difference. Although the calculation of the minimal detectable

foot–shoe complex and investigate its accuracy and reliability. difference conducted in this study is informed by pilot research, the

The inter-rater reliability of marker placement on the shoe SEM values calculated are consistent with previous kinematic SEM90

ranged from good to excellent, which indicates that markers can values reported in the literature [10]. Although MDD90 values <58

be applied reliably on the shoe between raters when using well- have been identified in some instances, they directly correspond to

developed palpation guidelines. While the reliability of marker joints that have small ranges of motion. When considering a

placement has been demonstrated, it is seemingly worthless if difference as a percentage of total joint ROM, it is likely that a large

the bone-embedded anatomical frame is defined incorrectly [7]. percentage difference will still need to be identified to identify a

The results of this study indicate that the accuracy of shoe- significant difference. Therefore, it must be questioned whether a

mounted marker placement (regardless of plane) was 6.7 mm foot–shoe complex model can detect small difference in joint

for all markers studied; more detail can be found in Bishop et al. kinematics (i.e. <58) as any such difference is likely to be subject to a

[7]. The variability in accuracy values on the toe box segment can combination of material/tissue artefact and measurement noise.

be partially explained by the shoe toe box, where the surface of Despite this model being the first known model of the foot–

the shoe often does not articulate with the dorsal surface of the shoe complex, it is not without limitations. Firstly, we must stress

foot. that the model was developed for a basic structured shoe. Although

Fig. 2. Joint kinematics. The flexion–extension (A), abduction–adduction (B) and inversion–eversion (C) motion at each of the tibiocalcaneal, tarsometatarsal and

metatarsophalangeal joints are presented. Individual gait events (IC – initial contact, 15–15% of stance, 50–50% of stance, PHD – peak toe box dorsiflexion and TO – toe off) are

presented on the abscissa of each diagram.

Please cite this article in press as: Bishop C, et al. The reliability, accuracy and minimal detectable difference of a multi-segment

kinematic model of the foot–shoe complex. Gait Posture (2012), http://dx.doi.org/10.1016/j.gaitpost.2012.09.020

G Model

GAIPOS-3720; No. of Pages 6

C. Bishop et al. / Gait & Posture xxx (2012) xxx–xxx 5

Table 2

Model’s minimal detectable difference. The standard error of measurement (SEM) and minimal detectable difference (MDD90) are reported for each rotation about the three

joints at specified gait events during stance phase. The SEM and MDD90 are also reported for total joint range of motion (ROM) for each degree of freedom. All values are

reported in degrees [8].

Outcome Measure Rotation Talocalcaneal joint Tarsometatarsal joint Metatarsophalangeal

joint

SEM MDD90 SEM MDD90 SEM MDD90

Angle @ initial contact Dorsi/plantarflexion 1.6 3.8 3.0 7.1 1.2 2.8

Ab/adduction 2.2 5.2 2.4 5.7 2.0 4.7

In/eversion 2.9 6.8 0.4 1.0 1.3 3.0

Angle @ 15% stance Dorsi/plantarflexion 1.3 3.0 1.8 4.3 3.2 7.4

Ab/adduction 3.2 7.5 2.5 5.9 2.1 4.8

In/eversion 3.4 7.9 2.2 5.1 1.5 3.4

Angle @ 50% stance Dorsi/plantarflexion 2.9 6.7 4.0 9.3 3.9 9.1

Ab/adduction 2.9 6.7 2.5 5.9 2.1 5.0

In/eversion 3.9 9.0 1.4 3.3 2.9 6.9

Angle @ peak toe box Dorsi/plantarflexion 3.4 7.8 1.0 2.4 2.7 6.4

dorsiflexion Ab/adduction 3.5 8.1 2.9 6.7 1.8 4.2

In/eversion 2.7 6.2 1.7 3.9 2.8 6.5

Angle @ toe-off Dorsi/plantarflexion 2.2 5.2 3.0 7.0 1.3 3.1

Ab/adduction 3.9 9.1 2.6 6.1 1.4 3.3

In/eversion 4.0 9.3 1.1 2.6 2.4 5.6

ROM Dorsi/plantarflexion 4.0 9.4 2.9 6.7 0.8 1.8

Ab/adduction 0.9 2.2 1.1 2.5 1.2 2.8

In/eversion 3.1 7.3 0.8 1.9 0.8 1.9

future work will investigate this issue in more detail, it is likely that University of South Australia for providing access to the radiology

more structured shoes will have a greater impact on marker suite and expertise required. ErgoLab was supported by AutoCRC

placement accuracy if anatomical landmarks are palpated through and the South Australian Government. We would also like to thank

a rigid heel counter. It is acknowledged that the kinematics of the Miss Hayley Uden and Mr Josh Cavanagh for their assistance in

foot inside the shoe are the major influences on the joint motion marker placement and Simpleware for the provision of a three-

measured, yet this motion will still be directly affected by the month software licence.

physical constraints of the shoe. We have not suggested that the

kinematics described in this study are either those of the foot or the Conflict of interest

shoe, yet rather those of the foot–shoe complex. Therefore this

model must be interpreted as being representative of the None declared.

interaction between the heel, midfoot and toe box during stance

phase of walking gait. In saying this, the kinematics reported in this

paper are likely to differ if a different shoe with a different

Appendix A. Supplementary data

structural design and/or material properties was used. Despite this

limitation, the methods proposed here allow for the translation of

the marker set and model to different shoe designs. This is Supplementary data associated with this article can be found, in

especially apparent when there is not the need and/or facilities to the online version, at http://dx.doi.org/10.1016/j.gaitpost.2012.

cut holes in shoes to quantify in-shoe foot kinematics. By no means 09.020.

does this negate the need for the understanding of in-shoe foot

function, and future research utilising this model must begin to References

describe how the foot and shoe surface interact together during

functional tasks that require shoes to be worn. [1] Kidder SM, Abuzzahab Jr FS, Harris GF, Johnson JE. A system for the analysis of

foot and ankle kinematics during gait. IEEE Transactions on Rehabilitation

Engineering 1996;4(1):25–32.

5. Conclusion

[2] Carson MC, Harrington ME, Thompson N, O’Connor JJ, Theologis TN. Kine-

matic analysis of a multi-segment foot model for research and clinical

applications: a repeatability analysis. Journal of Biomechanics 2001;34(10):

This is the first kinematic model purpose-built for quantifying

1299–307.

the kinematics of the foot–shoe complex. We have demonstrated

[3] McCulloch MU, Brunt D, Vander Linden D. The effect of foot orthotics and gait

accurate and reliable anatomical landmark palpation comparable velocity on lower limb kinematics and temporal events of stance. Journal of

Orthopaedic and Sports Physical Therapy 1993;17(1):2–10.

to models designed for barefoot investigations. Future applications

[4] Stacoff A, Reinschmidt C, Nigg BM, Van den Bogert AJ, Lundberg A, Denoth J,

of this model must use the thresholds proposed as an indicator of

et al. Effects of shoe sole construction on skeletal motion during running.

measurable change or differences when investigating either Medicine and Science in Sports and Exercise 2001;33(2):311–9.

between or within group effects. Any angular displacements [5] Shultz R, Jenkyn T. Determining the maximum diameter for holes in the shoe

without compromising shoe integrity when using a multi-segment foot model.

below these thresholds should be discounted as type I error.

Medical Engineering and Physics 2012;34(1):118–22.

[6] Van Sint Jan S. Color atlas of skeletal landmark definitions: guidelines for

Acknowledgments reproducible manual and virtual palpations. Edinburugh: Churchill Living-

stone-Elsevier; 2007.

[7] Bishop C, Thewlis D, Uden H, Ogilvie D, Paul G. A radiological method to

We would like to acknowledge ASICS Oceania for supplying the

determine the accuracy of motion capture marker placement on palpable

footwear used in this study and the School of Health Sciences, anatomical landmarks through a shoe. Footwear Science 2011;3(3):169–77.

Please cite this article in press as: Bishop C, et al. The reliability, accuracy and minimal detectable difference of a multi-segment

kinematic model of the foot–shoe complex. Gait Posture (2012), http://dx.doi.org/10.1016/j.gaitpost.2012.09.020

G Model

GAIPOS-3720; No. of Pages 6

6 C. Bishop et al. / Gait & Posture xxx (2012) xxx–xxx

[8] Leardini A, Benedetti MG, Berti L, Bettinelli D, Nativo R, Giannini S. Rear-foot, [16] Murley G, Menz H, Landorf K. A protocol for classifying normal- and flat-

mid-foot and fore-foot motion during the stance phase of gait. Gait and Posture arched foot posture for research studies using clinical and radiographic

2007;25(3):453–62. measurements. Journal of Foot and Ankle Research 2009;2(1):p22.

[9] Long JT, Eastwood DC, Graf AR, Smith PA, Harris GF. Repeatability and sources [17] WMA. World Medical Association Declaration of Helsinki: ethical principles

of variability in multi-center assessment of segmental foot kinematics in for medical research involving human subjects; 2008, Available from: http://

normal adults. Gait and Posture 2010;31(1):32–6. www.wma.net/en/30publications/10policies/b3/17c.pdf.

[10] Wright CJ, Arnold BL, Coffey TG, Pidcoe PE. Repeatability of the modified [18] Lu TW, O’Connor JJ. Bone position estimation from skin marker co-ordinates

Oxford foot model during gait in healthy adults. Gait and Posture 2011;33(1): using global optimisation with joint constraints. Journal of Biomechanics

108–12. 1999;32(2):129–34.

[11] Caravaggi P, Benedetti MG, Berti L, Leardini A. Repeatability of a multi- [19] Wu G, Siegler S, Allard P, Kirtley C, Leardini A, Rosenbaum D, et al. ISB

segment foot protocol in adult subjects. Gait and Posture 2011;33(1):133–5. recommendation on definitions of joint coordinate system of various joints

[12] Scott SH, Winter DA. Talocrural and talocalcaneal joint kinematics and kinetics for the reporting of human joint motion – Part I: Ankle, hip, and spine. Journal

during the stance phase of walking. Journal of Biomechanics 1991;24(8):743–52. of Biomechanics 2002;35(4):543–8.

[13] Collins TD, Ghoussayni SN, Ewins DJ, Kent JA. A six degrees-of-freedom marker [20] Landis JR, Koch GG. The measurement of observer agreement for categorical

set for gait analysis: repeatability and comparison with a modified Helen data. Biometrics 1977;33(1):159–74.

Hayes set. Gait and Posture 2009;30(2):173–80. [21] Winter D. Biomechanics and motor control of human movement. New Jersey:

[14] Deschamps K, Staes F, Roosen P, Nobels F, Desloovere K, Bruyninckx H, et al. John Wiley & Sons, Inc.; 2005.

Body of evidence supporting the clinical use of 3D multisegment foot models: [22] Zeni Jr JA, Richards JG, Higginson JS. Two simple methods for determining gait

a systematic review. Gait and Posture 2011;33(3):338–49. events during treadmill and overground walking using kinematic data. Gait

[15] Cappozzo A, Catani F, Della Croce U, Leardini A. Position and orientation in and Posture 2008;27(4):710–4.

space of bones during movement: anatomical frame definition and determi- [23] Portney LG, Watkins MP. Foundations of clinical research: applications to

nation. Clinical Biomechanics 1995;10(4):171–8. practice. Upper Saddle River, NJ: Pearson/Prentice Hall; 2009, xix, 892 pp..

Please cite this article in press as: Bishop C, et al. The reliability, accuracy and minimal detectable difference of a multi-segment

kinematic model of the foot–shoe complex. Gait Posture (2012), http://dx.doi.org/10.1016/j.gaitpost.2012.09.020