Fig. 39.1
Different variations of the Cleveland Clinic and Helen Hayes methods (From Manal et al. [15], with permission)
Manal et al. [15] evaluated the accuracy of 6 variations of such attachments for measuring axial tibial rotations during gait. They found it to be about ±4° using the most accurate variation (Fig. 39.1b). Given that the range of axial tibial rotation averages approximately 9°, this is a significant error. Ferber et al. [23] later used this variation of the attachment system to evaluate the test-retest reliability of this attachment system to record knee joint kinematics during running. For many kinematic parameters, intra-class correlation coefficients (ICCs) were found to be below 0.75, which is considered to be the limit of acceptability [24].
Other research groups have developed their own, more elaborate attachment systems and used them in recording lower limb gait kinematics [25–27]. The system developed and used by our research group is now called the KneeKGTM (Emovi inc., Montreal, Canada). First described by Sati et al. [28] under the name of exoskeleton, this system includes a femoral and a tibial component (Fig. 39.2). The design of the femoral component was inspired by a fluoroscopic study that showed that the amplitude of skin displacement about the knee varies greatly depending on the exact location [10]. The study identified two locations where the displacement is minimal. The femoral component attaches to these anatomical locations. It is composed of a rigid arch with spring-loaded forms at each extremity, effectively forming a clamp. The medial extremity of this clamp inserts between the vast medialis and the sartorius muscle tendon; the lateral extremity inserts between the femoral biceps and the ilio-tibial band (ITB). Both extremities rest atop the femoral condyles. A medial condyle support pad and a stabilizing plaque that is attached to a Velcro strap, placed around the proximal thigh, prevent rotation of the system about the clamp’s contact points.
Fig. 39.2
The tibial and femoral components of the KneeKGTM attachment system
The accuracy of the KneeKGTM was evaluated by a fluoroscopic study. Radio-opaque beads were fixed to the femoral component, which was installed on three different subjects who performed active flexion/extension under fluoroscopy. Average errors were found to be −0.4° in abduction/adduction and −2.3° in tibial rotation [28]. A subsequent study of five subjects, with a similar methodology, found that the quadratic error is diminished by a factor of 4.3 for abduction/adduction and by a factor of 6.2 for axial tibial rotation when compared with skin mounted markers [29]. Both these studies evaluated the accuracy of the KneeKGTM using non weight-bearing knee flexions.
The intra- and inter-observer reliability of the KneeKGTM were evaluated to insure that when the system was removed and reinstalled, similar results were obtained regardless of the examiner who installed it [30]. In the intra-observer setting, a single observer installed the system and recorded the gait kinematics three times each for 12 different subjects. Reliability was found to be high, with intra-class correlation coefficients (ICC) between 0.88 and 0.94 for knee rotations. The average standard measurement error (SEM) was below 1° for rotations about all three axes. In the inter-observer setting, three observers each installed the system twice on all 12 subjects. Similar to the intra-observer data, the ICCs from the inter-observer measurements were between 0.89 and 0.94, with all SEMs below 1°.
Movement Representation
Euler Angles Versus Helical Axis Definition
After knee movement has been measured precisely and reliably, it is necessary to represent it in a meaningful way. It has been suggested that the unambiguous description of spatial motion is more difficult than its measurement [31]. Because the knee is not a hinge and movement about that joint does not occur in a 2D plane, it is difficult to represent knee kinematics. Typically, 2 methods are used to study 3D kinematics of the knee: Euler angles [32] and helical axes [33].
Even though the knee is not gyroscopic, the Euler angle method is the most widely employed. With this method, it is possible to describe a 3D movement as 3 successive rotations about three different axes defined in space: flexion/extension, abduction/adduction and internal/external tibial rotation. These axes can be fixed or floating, and represented locally or globally. The major advantage of this method is that it is easier to interpret the results clinically with anatomical descriptions of movements. With this method, it is also possible to compute anteroposterior (AP), proximodistal (PD), and mediolateral (ML) translations.
However, the main disadvantage of the Euler angle method is that it is very sensitive to anatomical reference axes definition. Small errors (1–2 mm in the definition of points used to build the coordinate system) cause errors in orientation as well as in kinematic amplitude on the order of 2°. When coordinate systems are built on subjects, errors in landmark definition can be on the order of 30 mm. These large errors make it difficult, even impossible, to compare results [19]. Also, it is not clear if differences in bone geometry affect the kinematic patterns that are generated. For this reason, we do not know the 3D kinematics of the normal knee. Each knee has a kinematic representation associated with a given local coordinate system.
The helical axes method [33] uses the 3D position of each bone to describe the movement of the knee between 2 moments in time as a unique rotation and a unique translation about a finite rotation axis. Therefore, when employing this method to describe knee kinematics, we need to define the time period during which we want to express the rotation and translation of one bone with respect to the other. The main advantage of this method is that it is independent of an anatomical coordinate system definition. However, the use of helical axes to describe joint movements is not well understood by the clinical community [32]. Also, the method is sensitive to noise in the measurement and to the time period used for computation of the finite rotation axis.
We chose to follow the recommendation of the International Society of Biomechanics (ISB) by representing knee movements using Euler angles, which allow for better clinical interpretation. We therefore needed to define anatomical coordinate systems associated with the femur and tibia.
Anatomical Coordinate Systems
Anatomical landmarks need to be identified in order to establish the bone-embedded coordinate systems. These landmarks can be defined by a pointer or by fixing markers directly over them. This technique generates errors of many mm or even a few cm. To diminish imprecision when building coordinate systems, we have developed a method that uses fewer anatomical landmarks to define the reference coordinate system than other methods in the literature [34].
This coordinate system is called the Functional Postural (FP) method and was first described by Hagemeister et al. [35] The joint centers are defined as follows.
Ankle joint center (AJC):
The midpoint between the medial and lateral malleoli.
Hip joint center (HJC):
The center of the femoral head. It is identified from a recording of hip circumduction using a pivot algorithm, as proposed by Siston and Delph [36].
Knee joint center (KJC):
The midpoint between the medial and lateral epicondyles, projected onto the mean axis of knee flexion/extension
The PD axes of the tibia and femur are respectively defined by the vectors joining the KJC to the AJC and the KJC to the HJC. The subject is then placed in a reference guide with his frontal plane aligned with that of the guide. The sagittal plane is then defined during a movement of slight flexion/extension, alternating between approximately 10° of flexion and maximum extension. It is defined as the plane whose normal vector is the cross product of the normal vector of the frontal plane with the vector joining the HJC and the AJC. The neutral position of the knee joint is defined at the moment when the PD axes of the tibia and femur are aligned in the sagittal plane. In this position, the AP axes of the tibia and femur are defined as perpendicular to the normal vector of the sagittal plane and their PD axes. Finally, the ML axes of each of the bones are the axes that complete the orthonormal sets (Fig. 39.3).
Fig. 39.3
Coordinate axes construction using femoral condyles, malleoli (gray dots) and a functional method for definition of the centre of the femoral head and the centre of the knee (black dots)
Inter- and Intra-tester Variability
To test inter- and intra-tester variability for the calibration procedure, the following protocol was performed. The attachment system was first installed on each subject, and three testers performed the above-described calibration procedure on four subjects. Each tester repeated the procedure 5 times. Then, the subjects walked on a treadmill at a comfortable speed for 3 min, and finally, 30 gait cycles were recorded. The mean of these 30 cycles was used to compute the kinematic parameters (flexion/extension, abduction/adduction and internal/external rotation of the tibia as well as A-P translation as a function of percentage of the gait cycle) in association with the 15 calibration procedures. The resulting 15 curves were compared for each subject using an adjusted coefficient of multiple determinations. Figure 39.4 presents an example of the kinematic parameters calculated with 5 calibration procedures performed by 1 tester on 1 subject.
Fig. 39.4
Example of kinematic parameters calculated with 5 calibration procedures performed by 1 tester on 1 subject. Abduction is negative, internal rotation is negative, anterior translation of the femur is positive. Each curve corresponds to one trial
The results show that the calibration method allows the measurement of 3D knee kinematics with good reproducibility. Mean errors generated by the calibration procedure are 1.1° in flexion/extension, 1.1° in abduction/adduction, 0.8° in internal/external tibial rotation, and 2.6 mm in A-P translation.
Gait Analysis
Following an ACL injury, gait analysis has been shown to provide beneficial information for assessing knee stability [37] and functional impairments [38] under dynamic conditions. Numerous studies have already reported that patients with an ACL deficiency present altered 3D knee kinematics [9, 38–43] and 3D knee joint moments [38, 44–46]during gait. In these studies, 3D biomechanical patterns of the knee are presented over a full gait cycle, which is divided into a stance phase and in a swing phase. These phases represent 60 % and 40 % of the total gait cycle (GC), respectively. During the stance phase, sub-phases occur as follow: the initial contact (1–2 % of the GC), the loading phase (1–10 % of the GC), the mid stance phase (10–30 % of the GC), the terminal stance phase (30–50 % of the GC) and the pre-swing phase (50–60 % of the GC) [47].
Gait Biomechanics
Kinematics
To identify knee biomechanical deficiencies following an injury, a good understanding of “normal” patterns is required, since they serve as a reference to compare the pathological patterns [48]. Numerous studies have published “normal” knee biomechanical patterns during gait. Typically, the knee arcs of motion during walking in the sagittal, frontal and transverse planes are approximately 70°, 5° and 9°, respectively [47]. However, a qualitative review of these studies unveils a lack of correspondence in normal 3D knee kinematic patterns. These discrepancies are generally found in the frontal and transverse plane. The variation in methodologies used to record 3D joint kinematics between studies is potentially responsible for these differences [49].
Kinetics
During gait, the knee joint is continuously submitted to important moments and forces that influence the joint kinematics in all three planes of movement. Since joint kinetics cannot be measured in vivo, we must first record the joint kinematics and the ground reaction forces to compute these forces and moments [50] using calculations of inverse dynamics. The internal joint forces and moments applied by the joint’s soft tissue structures (muscles and ligaments) must constantly counteract the external forces and moments acting upon the knee. For example, during gait, the limb must alternately counterbalance moments that tend to extend and flex the knee joint. Quantifying knee joint kinetics allows a better understanding of functional adaptations following an injury such as an ACL tear [51].
ACL-Deficient Gait
The scientific literature relating to gait adaptations in ACL-deficient patients is abundant. However, a consensus on which gait compensatory mechanism is adopted by these patients remains to be established [52]. In fact, it was suggested that different biomechanical adaptations could be adopted within an ACLD population [43] and that gait adaptation changes over time [38]. To date, two main gait compensatory mechanisms have been proposed: (1) the quadriceps avoidance gait and (2) the hamstring facilitation strategy. Furthermore, only a restricted number of studies have looked at knee biomechanical patterns associated with the role of the ACL (i.e. anteroposterior translation and internal/external axial rotation). The following section will present a brief review of theses adaptations and biomechanical deficiencies and relate them to the results of our studies.
We conducted a 3D knee biomechanical assessment of 29 chronic ACLD patients and 15 healthy participants during treadmill walking. The 3D knee biomechanics were recorded using the KneeKGTM and a VICON optoelectronic system (Oxford Metrics, Oxford, UK). Joint centers and coordinate systems were defined using the FP method and the biomechanical patterns were computed using the ISB convention [32].
Quadriceps Avoidance Gait
Quadriceps avoidance gait is defined by the absence of the external knee flexor moment during the mid stance phase of the gait cycle [45]. Since this external joint moment tends to flex the knee while the body is progressing forward, an eccentric contraction of the quadriceps is typically required to counterbalance this moment. Berchuck et coll. 1990 [45] suggested that ACLD patients adopt this compensatory mechanism by reducing the quadriceps contraction. The rationale behind this hypothesis is that ACLD patients tend to avoid the anterior traction of the proximal tibia provoked by the quadriceps contraction, which could lead to AP instability. One study did show a decrease in quadriceps muscle activity during stance phase [53] and several other studies [38, 51, 54, 55] identified a decrease in the external knee flexor moment in ACLD patient.
However, not all of the scientific community endorses this compensatory mechanism theory. In fact, recently published studies refute the very presence of these changes [42, 52, 56]. In our study, ACLD patients did not exhibit a quadriceps avoidance gait pattern (Fig. 39.5). This is supported by several studies demonstrating no EMG changes in the quadriceps muscle activity [52, 57–59]. In contrast with the quadriceps avoidance gait theory, some authors suggest that a reduction of the external knee flexor moment could be related to other biomechanical adaptations. Indeed, some believe that this gait adaptation could be associated with higher knee flexion angles [38, 45, 55]. Others suggest it could be linked to an increased external hip flexor moment [43, 55, 60], which would decrease the tension in the quadriceps. These controversial results underline the lack of consensus concerning this knee biomechanical adaptation [61].
Fig. 39.5
Sagittal plane knee joint moments over a full gait cycle
Roberts et al. [52] suggested that the presence of the quadriceps avoidance gait pattern could be linked to methodological considerations since most of the studies reporting this strategy used a simpler linked segment model to compute knee joint moments. In our study, joint moments were computed using inverse dynamics with the wrench notation and quaternion algebra [62].
Hamstring Facilitation Strategy
In the past few years, several biomechanical studies have found that ACLD patients adopt a hamstring facilitation strategy. The rationale behind this compensatory mechanism theory is that higher hamstring muscle activity will act as an agonist to the injured ACL by generating a posterior traction of the proximal tibia. Numerous studies have shown a significant increase hamstring muscles activity [52, 53, 57, 59]. Furthermore, many studies have identified biomechanical changes that could be linked to this adaptation strategy. First, ACLD patients have been shown to walk with a higher knee flexion angle during the terminal stance phase [53, 58, 63–65]. Interestingly, the knee flexion angle during the stance phase was positively correlated with the duration of hamstring activity [57]. Additionally, ACLD patients have been found to display higher hamstring activity in the swing to stance transition [66]. The findings of our study are in agreement with these kinematic compensations. Indeed, Fig. 39.6 shows the knee flexion/extension pattern over a full gait cycle for the ACLD group and the control group. The ACLD group walked with significantly higher knee flexion angles at initial ground contact, during the terminal stance phase and at the end of the swing phase. The asterisks (*) show where statistical differences (P < 0.05) were identified.
Fig. 39.6
Sagittal plane knee kinematic patterns during gait for ACL-deficient patients and healthy participants
Biomechanical Deficiencies Associated with the Role of the ACL
Studies quantifying the AP translation and internal-external axial rotation of the knee are scarce. This is mainly due to the high level of error associated with the measurement of these small amplitude translations and rotations. Nevertheless, an increased anterior translation of the tibia in the transition between a non-weightbearing and weightbearing condition was reported in ACLD patients [67]. Furthermore, two studies using an electrogoniometer to quantify the kinematics of ACLD knees during gait showed an increase in anterior tibial translation [9, 68]. However, given the error associated with the measurement of proximal AP translation of the tibia [14], these results should be considered with caution.
Although only a few papers were published on the impact of an ACL tear on the knee axial rotation, contradictions emerge between studies. Some authors identified an increase in external tibial rotation with regards to the femur during gait [9, 52]. Others found a shift towards internal tibial rotation [40, 41]. Our study is in agreement with the latter. Whereas previous papers showed that this shift occurs throughout the gait cycle, our results only identified statistical differences in the swing to stance transition and during the beginning of the loading phase (Fig. 39.7).
Fig. 39.7
Transverse knee kinematic patterns during gait for ACL-deficient patients and healthy participants
The identified decrease in external tibial rotation in ACLD patients at the end of the swing phase, where the knee reaches near full extension, was also reported in previous studies [40, 68]. These results support previous findings that the screw home mechanism is altered after an ACL injury [69]. This altered axial tibial rotation could also be explained by the lack of knee extension at the end of the swing phase.
Three-dimensional biomechanical evaluations have allowed a better understanding of the impact of an ACL injury on the function of the knee joint. The large number of studies on this subject underlines the importance of such evaluations as complements to orthopaedic physical assessments in helping to improve current treatments for ACL injuries.
Analysis of the Pivot Shift Phenomenon
During a clinical evaluation, an ACL-deficient knee will generally present an increase in joint laxity, especially in the AP axis. This uniaxial laxity is relatively easy to evaluate using clinical examinations. The most sensitive of these examinations is the Lachman test, whereby the clinician applies an anterior force to the tibia while holding the femur in place. The amount of anterior displacement is very useful in diagnosing an ACL rupture [70–73] but it shows no correlation to subjective criteria of knee joint function [74–78]. The pivot shift test is a dynamic clinical test that reproduces the 3D rotational instability felt by patients who describe feeling that their knee is giving way. The pivot shift grade correlates with reduced patient satisfaction, partial giving way, full giving way, difficulty cutting, difficulty twisting, activity limitation, reduced overall knee function, reduced sports participation, and Lysholm score [74]. Subjects with higher-grade pivot shift tests had less satisfaction, more limitations and lower knee function.
A meta-analysis found that the Lachman test is more sensitive but less specific than the pivot shift test [70]. In other words, a positive pivot shift test indicates an ACL rupture and a negative Lachman test rules it out.
Although it is obviously important to diagnose an ACL rupture, it is often the level of knee joint function that is of most interest and that clinicians aim to restore. Both tests are important parts of a clinical evaluation but serve different purposes. The pivot shift test’s specificity makes it a valuable complement to the Lachman test in establishing a diagnosis, but more importantly, the pivot shift test is the only test which can be used to assess the level of knee joint function and predict long term outcome. As such, many studies conclude that the objective of reconstructive surgery should be to eliminate the presence of a pivot shift and not only to diminish AP laxity [39, 79–81].
Clinical Examination
The pivot shift test is performed with the patient supine and the examined leg lifted off the examining table by a clinician. A gentle valgus force is applied to the knee and the knee is flexed in a controlled manner with slight internal rotation of the tibia. In the ACL-deficient knee, as flexion occurs, the tibia translates anteriorly and rotates internally. The joint is subluxed at this point. As the knee is flexed past 30°, soft tissues and joint geometry cause the joint to reduce [82]. This is the pivot shift.
The clinician attributes the grade of the pivot shift relying on his interpretation and experience as being 0 (absent), 1 (glide), 2 (clunk) or 3 (gross) [83]. The nature of this grading scale renders it poorly repeatable [84]. Indeed, it has been shown that different clinicians frequently attribute different grades to a same patient [85]. No objective method for evaluating the pivot shift test currently exists, despite several attempts in the literature [86–91]. In the absence of an objective pivot shift measurement tool, it is difficult for less experienced clinicians to attribute a grade with a sufficiently high level of confidence for it to be used in determining the course of treatment.
Recording the Pivot Shift
The KneeKGTM attachment system, presented in Sect. 2.1.3, was designed for use on a standing subject. Its femoral component, which rests atop the femoral condyles, falls out of place when a subject is placed in a supine position. For this reason, an adapted version of this system was developed to record the kinematics of the pivot shift. The rigid arc of the femoral component is replaced by an elastic Velcro strap (Fig. 39.8a). This strap allows for inward pressure to be applied and it prevents the attachment from falling out of place when the subject is supine. The purpose of the rigid plates is to improve the attachment’s stability and to allow for fixation of the motion capture sensors.
Fig. 39.8
The femoral (a) and tibial (b) components of the attachment system developed for data acquisition with the subject in a supine position
The tibial component is composed of a rigid plate that is held over the tibia with an elastic Velcro strap, immediately distal to the tibial tuberosity (Fig. 39.8b). It is short in length to allow a clinician to manipulate the lower limb without displacing it. A preliminary study conducted with three subjects wearing this attachment system showed it to be as reliable as the KneeKGTM for gait analysis.
The FP method, which is used to establish the anatomical axes, was also adapted to for the pivot shift test. The joint rotations were passively applied on the supine subject by the examiner.
Kinematics of the Pivot Shift
The reduction phase of the pivot shift occurs when the subluxed tibia returns to its normal position. This reduction has been described as a combination of posterior translation and external tibial rotation. With the development of methodologies to record the 3D kinematics of the pivot shift, many studies have investigated these features and how they relate to the grade of the pivot shift established by a clinician. One of the objectives of such studies is to develop an objective grading scale where, for example, some combination of both axial tibial rotation and posterior translation would be taken as a pivot shift score and the grade would be established from these scores.
Recent studies have found that the 3D kinematics of the pivot shift vary too much between subjects for such a simple score to be used. Although the amplitude of posterior translation correlates to the clinical grade, values are very different for two subjects of a same grade. The axial tibial rotation is even more variable: some subjects present significant external rotations during the reduction phase while others show none at all [88], leading authors to question its relevance in measuring the pivot shift.
In a recent study, we recorded the knee joint kinematics of 127 pivot shift tests and investigated the correlation with the clinical grade established by the clinician who performed the test [30, 92, 93]. Figure 39.9 presents a typical recording of a grade 3 pivot shift. A spike in linear velocity and acceleration is clearly visible during the reduction phase.
Fig. 39.9
Kinematics of a typical grade 3 pivot shift (Adapted from Labbe et al. [30], with permission.)
To further investigate the kinematic features of the pivot shift that are related to its clinical grade, we conducted a principal component analysis (PCA). The objective of PCA is to reduce the number of features while retaining as much of the variation present in the original dataset as possible. To do so, it transforms a large number of features into a smaller number of uncorrelated features, called principal components (PCs). The first PC accounts for the largest amount of variability in the data and each of the subsequent PC accounts for the largest amount of the remaining variability.
Related posts:
Biomaterials Used in Orthopedics
Clinical Application of Glass Ceramics
Knee Ligamentoplasty: Prosthetic Ligament or Ligament Allograft?
Efficacy and Safety of an Absorbable Cervical Cage with and Without Plating: A Multicenter Case Study
Pedicle Screw Fixation in Thoracic or Thoracolumbar Burst Fractures
Biomechanics of Osteosynthesis by Screwed Plates
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
