propagation of measurement errors to foot kinematics analysis

1
Joint Kinematics Presentation O-164 S167 PROPAGATION OF MEASUREMENT ERRORS TO FOOT KINEMATICS ANALYSIS Hossein Rouhani (1), Julien Favre (1), Xavier Crevoisier (2), Brigitte M. Jolles (2), Kamiar Aminian (1) 1. Ecole Polytechnique Fédérale de Lausanne (EPFL-LMAM), Switzerland 2. Centre Hospitalier Universitaire Vaudois & University of Lausanne, Switzerland Introduction With the precision of recently developed motion capture systems, a trend to study foot as a multi segment limb has emerged [Arndt, 2007; Simon, 2006]. However, the classic methods of body movement analysis have not shown to be enough reliable for tiny movements of foot joints. The precision of measurement system and markers (or sensors) misplacement induce errors in kinematic results. The propagation of these errors in foot kinematics could also depend on the mathematical tools used to express foot kinematics. This study aimed to appraise such error propagations with the use of different methods of foot kinematics expression. Method 21 tiny markers (3 mm diameter) were mounted on main anatomical landmarks of foot and 4 other markers (9 mm diameter) on knee. Six young healthy subjects (age: 27±3) were asked to walk on a forceplate (Kistler, CH) surrounded by 6 infra-red cameras (VICON, UK). The stance phase of one gait cycle was captured in 9~12 trials per subject. Foot and ankle complex was considered as 6 rigid segments: 1) Shank, 2) Hindfoot, 3) Mid-foot, 4) Medial forefoot, 5) Lateral forefoot and 6) Phalanges (Fig. 1). A local coordinate system was assigned to each segment [Leardini, 2007]. 3D angles between each two segments were calculated based on three prominent methods: a) helical angle, b) joint coordinate system and c) projection of lines representing segments on three anatomical planes. Two sources for error were considered: i) 1 st Err.: The precision of measurement system as a 3D error changing at every samples (dispersion: 0.2mm) ii) 2 nd Err.: The precision of landmark localization as a 3D constant error at each trial (dispersion: 6mm). The difference between the original and corrupted angles were calculated (as RMS values) and averaged over all trials of all subjects. Results Values of propagated errors of joints are given in table 1 for three foot joints, namely: hind-mid foot (2-3), mid-medial forefoot (3-4) and mid-lateral forefoot (3-5). Figure 1- Local coordinate system for foot segments Method Anatomical plane Foot Joint 1 st Err. (deg) 2 nd Err. (deg) Helical angle 3D 2-3 0.9 2.0 3-4 0.7 1.7 3-5 0.5 1.6 Joint coordinate system Sagittal 2-3 0.9 2.7 3-4 0.7 2.0 3-5 0.6 1.4 Coronal 2-3 0.4 0.6 3-4 0.6 1.1 3-5 0.6 1.6 Transverse 2-3 0.4 0.6 3-4 0.5 1.6 3-5 0.4 0.9 Projection of segment lines on anatomical planes Sagittal 2-3 0.9 1.8 3-4 0.9 1.9 3-5 0.5 0.9 Coronal 2-3 0.4 0.8 3-4 0.5 0.9 3-5 0.4 0.8 Transverse 2-3 1.2 1.8 3-4 1.1 2.2 3-5 0.7 1.5 Table 1- Propagation of the measurement errors (RMS values in degree) to different joint angles Discussion This study showed that the three angle calculation methods had comparable robustness against device error and marker misplacement. The propagated errors were enough small to allow their use for multi segment foot’s kinematics. Clinical studies with foot pathology are required to confirm these results. References Arndt et al, Biomechanics, 40: 2672-2678, 2007 Leardini et al, Gait & Posture, 25: 453-462, 2007 Simon et al, Gait & Posture, 23: 411-424, 2006 16th ESB Congress, Oral Presentations, Tuesday 8 July 2008 Journal of Biomechanics 41(S1)

Upload: hossein-rouhani

Post on 02-Jul-2016

212 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: PROPAGATION OF MEASUREMENT ERRORS TO FOOT KINEMATICS ANALYSIS

Joint Kinematics Presentation O-164 S167

PROPAGATION OF MEASUREMENT ERRORS TO FOOT KINEMATICS ANALYSIS

Hossein Rouhani (1), Julien Favre (1), Xavier Crevoisier (2), Brigitte M. Jolles (2), Kamiar Aminian (1)

1. Ecole Polytechnique Fédérale de Lausanne (EPFL-LMAM), Switzerland

2. Centre Hospitalier Universitaire Vaudois & University of Lausanne, Switzerland

Introduction With the precision of recently developed motion capture systems, a trend to study foot as a multi segment limb has emerged [Arndt, 2007; Simon, 2006]. However, the classic methods of body movement analysis have not shown to be enough reliable for tiny movements of foot joints. The precision of measurement system and markers (or sensors) misplacement induce errors in kinematic results. The propagation of these errors in foot kinematics could also depend on the mathematical tools used to express foot kinematics. This study aimed to appraise such error propagations with the use of different methods of foot kinematics expression. Method 21 tiny markers (3 mm diameter) were mounted on main anatomical landmarks of foot and 4 other markers (9 mm diameter) on knee. Six young healthy subjects (age: 27±3) were asked to walk on a forceplate (Kistler, CH) surrounded by 6 infra-red cameras (VICON, UK). The stance phase of one gait cycle was captured in 9~12 trials per subject. Foot and ankle complex was considered as 6 rigid segments: 1) Shank, 2) Hindfoot, 3) Mid-foot, 4) Medial forefoot, 5) Lateral forefoot and 6) Phalanges (Fig. 1). A local coordinate system was assigned to each segment [Leardini, 2007]. 3D angles between each two segments were calculated based on three prominent methods: a) helical angle, b) joint coordinate system and c) projection of lines representing segments on three anatomical planes. Two sources for error were considered: i) 1st Err.: The precision of measurement system as a 3D error changing at every samples (dispersion: 0.2mm) ii) 2nd Err.: The precision of landmark localization as a 3D constant error at each trial (dispersion: 6mm). The difference between the original and corrupted angles were calculated (as RMS values) and averaged over all trials of all subjects. Results Values of propagated errors of joints are given in table 1 for three foot joints, namely: hind-mid foot (2-3), mid-medial forefoot (3-4) and mid-lateral forefoot (3-5).

Figure 1- Local coordinate system for foot segments

Method Anatomical plane

Foot Joint

1st Err. (deg)

2nd Err. (deg)

Helical angle

3D 2-3 0.9 2.0 3-4 0.7 1.7 3-5 0.5 1.6

Joint coordinate

system

Sagittal 2-3 0.9 2.7 3-4 0.7 2.0 3-5 0.6 1.4

Coronal 2-3 0.4 0.6 3-4 0.6 1.1 3-5 0.6 1.6

Transverse 2-3 0.4 0.6 3-4 0.5 1.6 3-5 0.4 0.9

Projection of segment

lines on anatomical

planes

Sagittal 2-3 0.9 1.8 3-4 0.9 1.9 3-5 0.5 0.9

Coronal 2-3 0.4 0.8 3-4 0.5 0.9 3-5 0.4 0.8

Transverse 2-3 1.2 1.8 3-4 1.1 2.2 3-5 0.7 1.5

Table 1- Propagation of the measurement errors (RMS values in degree) to different joint angles

Discussion This study showed that the three angle calculation methods had comparable robustness against device error and marker misplacement. The propagated errors were enough small to allow their use for multi segment foot’s kinematics. Clinical studies with foot pathology are required to confirm these results. References Arndt et al, Biomechanics, 40: 2672-2678, 2007 Leardini et al, Gait & Posture, 25: 453-462, 2007 Simon et al, Gait & Posture, 23: 411-424, 2006

16th ESB Congress, Oral Presentations, Tuesday 8 July 2008 Journal of Biomechanics 41(S1)