Create an optimization model of the internal structure of the knee joint to quantify the correlation between external knee adduction moment (M[add]) during gait with the medial-to-lateral ratio of compartment loading (MLR). Patients were examined the week before, and six months after, surgical knee joint realigment with a high tibial osteotomy (HTO).

Thirty patients (six females, twenty-four males; age = 50.0 ± 9.4 yrs.; BMI = 30.0±2.8) with clinically diagnosed OA primarily affecting the medial compartment of the knee underwent a medial opening wedge HTO. Walking gait analysis was performed immediately pre-surgery and at six months post-surgery using optical motion analysis (eight Eagle camera EvaRT system, Motion Analysis Corp, Santa Rosa, CA, USA) and floor-mounted force plate (OR6, AMTI, Watertown, MA, USA). External joint kinetics were calculated using inverse dynamics. Kinematic and force plate data served as input for the internal knee joint model. The anatomical geometry was generic but scaled to patient height and knee alignment. Included were four ligaments (ACL, PCL, LCL, MCL), two contact surfaces (medial and lateral) and eleven muscles (quadriceps, hamstrings, gracilis, sartorius, popliteus and gatrocnemius). A loading solution was found to satisfy mechanical equilibrium and minimise the sum of squares of all structural loads. Output was the ratio of medial-to-lateral compartment compression (MLR). Paired t-tests compared M[add] pre-op versus post-op and MLR pre-op versus post-op. A Pearson R2 coefficient of determination was calculated correlating M[add] to MLR for the pre-operative condition.

Peak M[add] decreased from 2.53 ± 1.32 to 1.63 ± 0.81 [%body weight*ht] (p< 0.001). The peak MLR decreased from 2.63 ± 1.08 to 1.52 ± 0.56 [unit-less] (p< 0.001). There was a moderate correlation between M[add] and MLR with the Pearson R2=0.457 (p=0.014).

These results suggest that adduction moment is an acceptable proxy for quantifying the internal compressive loading in the knee. Even without considering muscle loading and possible co-contraction of antagonists, adduction moment explains nearly half of the variance in the internal loading of the knee joint compartments. However, further research is required with a larger sample size to increase confidence in this proxy measure in a clinical setting.

Optical motion analysis (MA) is a useful tool for evaluating musculoskeletal function in health and disease. MA is particularly useful in quantifying joint kinematic and kinetic abnormalities accompanying osteoarthritis. However, current practice does not allow the joints of the foot to be measured since the foot is treated as a single rigid segment. To develop a multi-segment kinematic model of the foot for use in a clinical motion analysis laboratory. Apply the model to a healthy population during normal walking and gait intentionally disrupted by a high arch orthotic.

The foot was defined as five rigid segments: hindfoot (calcaneus), midfoot (tarsus), medial forefoot (first metatarsal), lateral forefoot (fifth metatarsal) and the hallux (both phalanges). Each of these segments were tracked individually using custom-built marker triads attached to the skin. Thirty healthy subjects (eleven male, nineteen female; mean age 27.7 years, range 19–53) were examined using MA (eight Eagle camera, EvaRt system, Motion Analysis Corp., Santa Rosa, CA, USA) during normal walking and gait disrupted with a high arch orthotic taped to the plantar surface. All trials were performed barefoot. The special foot marker system was applied to the right foot with the remaining markers in the Helen Hayes configuration. Three motions are reported. The hallux-medial forefoot angulation (HA) is reported in the sagittal plane (plantar-dorsiflexion). The hindfoot-midfoot angulation (HFA) is also reported in the sagittal plane (plantar-dorsiflexion). The height-to-length ratio of the medial-longitudinal arch (MLA) is reported, normalised to zero in quiet standing. Paired t-tests compared the normal and disrupted gait conditions. All angles were compared at the instant of foot flat.

HA was not significantly changed between normal and disrupted conditions: from 8.5° ± 6.4° to 8.6° ± 7.4° (p=0.88). The HFA plantar-flexion significantly increased from 0.5 ° ± 3.3° (normal) to 2.9° ± 4.4° (disrupted; p< 0.01); mean difference = +2.5° (95% CI: 0.81 to 4.1°). The MLA was significantly increased (arch raised) from 0.004 ± 0.018 (normal) to 0.017 ± 0.021 (disrupted; p< 0.01); mean increase = +0.012 (95% CI: 0.00421 to 0.021).

A multi-segment kinematic model of the foot has been successfully implemented in an optical motion analysis laboratory. The model was sensitive to an intentional disruption of normal foot kinematics during walking in a healthy population.

Purpose: To develop a computerized inverse dynamic 3D model of the upper limb, focussing on the elbow.

Methods: Anatomic bony landmarks were identified in one cadaveric arm using an electromagnetic tracking device (Flock of Birds, Ascension Technologies, VT). The articular surfaces of the radiohumeral and ulnohumeral joints were digitized, thereby identifying the areas over which the contact forces could act. Attachment sites of the medial collateral (MCL) and lateral collateral (LCL) ligaments and the major muscles (BRA=brachialis, BIC=biceps, BRD=brachioradialis, TRI=triceps) were also digitized to create line-of-action vectors. These data were fed into custom-written software (MATLAB®, The MathWorks Inc., MA) that simulated flexion with gravity as external loading, and calculated the forces exerted by the joint structures. As an indeterminate system, computerized mathematical optimization solved for the internal loads using a cost function that minimized the sum of forces squared.

Results: Model outputs were comparable with results from previous muscle activity and cadaveric studies. Force ratios among the elbow’s prime movers at 30 degrees of flexion agreed quite closely with previous findings (Funk et al, 1987), with percent differences of 11% (BRA), −5% (BIC), −6% (BRD), and −1% (TRI). Overall, the brachialis force was the highest throughout flexion, being the prime mover, while extensor (triceps) activity remained quiet through mid-range. The magnitude of the radiohumeral contact force showed a decreasing pattern through the arc of flexion, similar to the trend found experimentally by others (Morrey et al, 1988). The results also demonstrated stabilizing forces supplied by the MCL, but not the LCL.

Conclusions: Current understanding of upper extremity loading is very limited. Creating an accurate computerized model of the elbow joint, would reduce the need for experimental testing with cadavers, which are always of limited availability. While stability of the elbow has been experimentally investigated, this model will be able to quantify the forces within the stabilizing structures. By establishing a normal baseline of these forces, surgical procedures and joint replacement designs can be validated. Thus, this model can provide a significant contribution to upper extremity biomechanics research and clinical treatments.

The peak external knee adduction moment during walking gait has been proposed to be a clinically useful measure of dynamic knee joint load in patients with knee osteoarthritis. However, there is limited information about the reliability of this measure, or its ability to detect change. The test-retest reliability and sensitivity to change of peak knee adduction moments were evaluated in thirty patients with varus gonarthrosis. Indices of relative and absolute reliability were excellent (intra-class correlation coefficient = 0.85, standard error of measurement = 0.36 % BW*Ht), and the sensitivity to change following high tibial osteotomy was high (standardized response mean = 1.2).

To estimate the test-retest reliability, measurement error and sensitivity to change of the peak knee adduction moment during gait.

Thirty patients (44”11 yrs, 1.7”0.09 m, 87”20 kg, twenty males, ten females) with varus gonarthrosis underwent gait analyses on two pre-operative test occasions within one week, and on a third test occasion six months after medial opening wedge high tibial osteotomy. Three-dimensional kinematic and kinetic gait data were collected during self-paced walking and used to calculate the peak knee adduction moment.

An intraclass correlation coefficient of 0.85 (95%CI: 0.71, 0.93) indicated excellent relative reliability, and a standard error of measurement of 0.36 %BW*Ht (95%CI: 0.29, 0.49) indicated low measurement error. The peak knee adduction moment after surgery (1.66”0.72 %BW*Ht) was significantly (p< 0.001) lower than before surgery (2.58”0.72 %BW*Ht). A standardized response mean of 1.2 (95%CI: 0.77, 1.6) indicated the size of this change was large.

Based on 95% confidence levels, these results suggest the error in an individual’s peak knee adduction moment at one point in time is 0.70 % BW*Ht, the minimal detectable change in an individual’s peak adduction moment is 1.0 %BW*Ht, and it is sensitive to change following treatment.

The peak knee adduction moment during gait has appropriate reliability for use in studies evaluating the effect of treatments intended to decrease the load on the knee. When considering measurement error, the knee adduction moment is also appropriate for clinical use in evaluating change in individual patients.

Funding: CIHR, Arthrex Inc.