Influence of thigh cluster configuration on the estimation of hip axial rotation
Introduction
Hip axial rotation during gait is of clinical significance. For example, children with cerebral palsy typically walk with increased hip internal rotation [1]. Hip axial rotation during gait is therefore a common outcome measure for studies evaluating the effects of surgical interventions in cerebral palsy [2], [3], [4]. This necessitates accurate measurement techniques for non-invasively estimating hip axial rotation.
Conceptually, there are two major sources of error associated with the non-invasive estimation of hip axial rotation. The first source concerns the process of defining the relevant anatomical frames (AF). Specifically, the procedure for identifying the femoral AF frontal plane (conventionally defined by the knee joint flexion–extension axis) can be prone to error [5]. As ‘neutral’ hip axial rotation is reliant upon the orientation of the knee flexion–extension axis, errors in defining this axis manifest as offsets in the hip axial rotation kinematic profile. Optimisation procedures designed to minimise this error have recently been proposed [6], [7], [8], [9]. The second source concerns soft tissue artefact (STA). In the context of hip axial rotation, it is STA at the thigh that is most critical. The thigh is surrounded by large amounts of soft tissue and contains few areas of subcutaneous bone. Hence, it is associated with the greatest degree of STA in the lower limb [10], [11], [12]. As the amplitude of hip axial rotation during gait is small [13], errors must be minimised. Quantification of thigh STA and its effect on the estimation of hip axial rotation is therefore necessary to be able to critically interpret results. This is especially important if results are to be used to aid clinical decision-making.
Despite the potential for the non-invasive estimation of hip axial rotation to be erroneous, few studies have investigated this issue. Of those that have, only simple non-daily living motor tasks have been considered. Lamoreux [14] evaluated the performance of several different thigh clusters for one subject using an isolated lower limb longitudinal rotation task. Hip axial rotation kinematic profiles generated from a known reference, obtained in this instance from shank markers, were compared with those generated from markers at different locations on the thigh. Thigh markers were found to be only capable of estimating, at best, up to 70% of the reference amount of movement. Cappozzo et al. [10] also utilised a longitudinal rotation task. The orientation of a femoral AF generated from a technical frame (TF) rigid with the bone (via an external fixator) was compared with that generated from a skin marker TF for two subjects. Depending upon thigh marker location, transverse plane orientation errors varied from 6° to 28° for rotations in the range of 45°. With respect to gait, several studies have compared joint kinematics generated from skin versus bone-mounted markers [12], [15], [16]. However, none of these studies inserted bone pins into the pelvis. Thus, the effect of STA on hip joint kinematics has not been considered. Limited knowledge therefore exists regarding the effect of STA at the thigh on the non-invasive estimation of hip axial rotation.
One way to evaluate the effect of STA at the thigh is to compare the performance of different cluster configurations. This is an attractive approach given the ethical constraints that may be associated with the use of bone pins or fluoroscopy [17], [18]. Furthermore, natural soft tissue movement is not impeded and results are readily transferable to clinical practice. The magnitude of STA at the thigh is known to be location specific [10], [14], [19]. This information can be used to devise thigh clusters comprising of markers overlying areas least susceptible to STA. In support of this approach, previous studies have shown that STA propagation to estimated knee joint kinematics is strongly dependent upon cluster configuration [10], [16], [19], [20], [21]. In the context of the current study, preliminary data from Lamoreux [14] and Cappozzo et al. [10] suggest that cluster configuration is also likely to be critical in minimising errors associated with the non-invasive estimation of hip axial rotation.
The purpose of this study was to compare the relative performance of four alternative thigh cluster configurations. It was hypothesised that: (a) the estimation of hip axial rotation would be sensitive to thigh cluster configuration [H1]; (b) a more favourable solution would be obtained from clusters overlying areas on the thigh that are least susceptible to STA [H2].
Section snippets
Subjects
Twenty able-bodied adults (five males; 15 females) with a mean height of 165.8 (S.D. 7.9) cm, body mass of 60.9 (S.D. 9.7) kg and age of 20.8 (S.D. 4.1) years were voluntarily recruited. Approval was obtained from the Royal Children's Hospital Ethics in Human Research Committee prior to commencement and subjects signed a consent form.
Instrumentation
Kinematic data were acquired using a three-dimensional (3D) motion analysis system (VICON 512, Oxford Metrics, Oxford, England) with six high-resolution M1 cameras
Estimating hip axial rotation during gait
Hip axial rotation kinematic profiles during gait were sensitive to thigh cluster configuration (Fig. 2). The overall similarity between profiles was quite poor, with a mean CMD value of 0.274 (Table 3). The paired comparisons demonstrated that the hip axial rotation kinematic profile for thigh cluster A was dissimilar to the profiles for all other thigh clusters (CMD < 0.2). In contrast, the hip axial rotation kinematic profiles for thigh clusters C and D were virtually identical. Thus, the
Estimating hip axial rotation during gait
The estimated hip axial rotation kinematic profile during gait was sensitive to thigh cluster configuration. The kinematic profile from thigh cluster A was distinctly different to the profiles from all other clusters. In contrast, the kinematic profiles from thigh clusters C and D were virtually identical (Fig. 2 and Table 3). The estimated hip axial rotation kinematic profile during gait is therefore more dependent on thigh cluster location than specific construction effects (e.g.,
Acknowledgment
This project was financially supported by a Health Professional Research Training Fellowship from the Australian National Health and Medical Research Council (Grant ID: 237153).
References (37)
- et al.
Defining the knee joint flexion–extension axis for purposes of quantitative gait analysis: an evaluation of methods
Gait Posture
(2006) - et al.
A new approach to determine the hip rotation profile from clinical gait analysis data
Hum Mov Sci
(1999) - et al.
Repeatability of gait data using a functional hip joint centre and a mean helical knee axis
J Biomech
(2003) A correction for axis misalignment in the joint angle curves representing knee movement in gait analysis
J Biomech
(2005)- et al.
A new method for estimating joint parameters from motion data
J Biomech
(2005) - et al.
Position and orientation in space of bones during movement: experimental artefacts
Clin Biomech
(1996) - et al.
Tibiofemoral and tibiocalcaneal motion during walking: external vs. skeletal markers
Gait Posture
(1997) - et al.
Effect of skin movement artifact on knee kinematics during gait and cutting motions measured in vivo
Gait Posture
(2006) - et al.
Validity and comparisons of tibiofemoral orientations and displacement using a femoral tracking device during early to mid stance of walking
Gait Posture
(2004) - et al.
Human movement analysis using stereophotogrammetry. Part 3. Soft tissue artifact assessment and compensation
Gait Posture
(2005)
Quantification of soft tissue artefact in motion analysis by combining 3D fluoroscopy and stereophotogrammetry: a study on two subjects
Clin Biomech
Skin movement artefact assessment and compensation in the estimation of knee joint kinematics
J Biomech
A gait analysis data collection and reduction technique
Hum Mov Sci
Bone position estimation from skin marker coordinates using global optimisation with joint constraints
J Biomech
Tibio-femoral movement in the living knee. A study of weight bearing and non-weight bearing knee kinematics using ‘interventional’ MRI
J Biomech
Imaging knee position using MRI, RSA/CT and 3D digitisation
J Biomech
The movement of the normal tibio-femoral joint
J Biomech
Three-dimensional kinematics of the human knee during walking
J Biomech
Cited by (39)
Reductions in rearfoot eversion posture due to proximal muscle strengthening are dependent on foot-ankle varus alignment
2024, Journal of Bodywork and Movement TherapiesA new marker cluster anchored to the iliotibial band improves tracking of hip and thigh axial rotations
2023, Journal of BiomechanicsMuscle actions on crossed and non-crossed joints during upright standing and gait: A comprehensive description based on induced acceleration analysis
2022, Journal of BiomechanicsCitation Excerpt :Thirty-two of the markers were placed on anatomical landmarks (adapted from Wu et al. (2002), Langley et al. (2019), Dorn et al. (2012b)) to allow determining the model dimensions and position, according to an initial static trial with all the markers. In addition, semi-rigid clusters of tracking markers, each with four markers (a total of 16 markers), were attached to the thighs and shanks (Manal et al., 2000; Schache et al., 2008), and some of the markers on anatomical landmarks were also used for motion tracking (Dorn et al., 2012b). The task trials (gait and upright standing) were conducted with the tracking markers only.
The effects of small and large varus alignment of the foot-ankle complex on lower limb kinematics and kinetics during walking: A cross-sectional study
2020, Musculoskeletal Science and PracticeCitation Excerpt :The kinematic and kinetic data collected during walking was processed with the Visual3D software (C-Motion Inc, Rockville, United States). The biomechanical model of forefoot (Leardini et al., 2007), rearfoot (Leardini et al., 2007), foot (C-Motion Visual3D Documentation, 2015), shank (Bruening et al., 2012; Manal et al., 2000), thigh (Schache et al., 2008; Wu et al., 2002) and pelvis (Wu et al., 2002) were implemented. The marker trajectories and GRF data were filtered using fourth-order Butterworth low-pass filters with 6 Hz and 10 Hz cut-off frequencies, respectively (Robertson and Dowling, 2003).