RISK FACTORS FOR KNEE INJURIES IN SKIIN

Get Complete Project Material File(s) Now! »

Chapter 6: Validation of the kinematic measurements

This study validates the measurement equipment utilized for the studies and its accuracy. Kinematic data often serve as the input for computer simulations. Thus, a high level of measurement accuracy is required in order to assess technique or injury mechanisms in snow sports in certain situations or following equipment interventions. However, some imprecision must be accepted as measurement errors are inevitable. For instance, a standard resolution of 768 by 576 and an image of recording of 10 m leads to a pixel length of 1.3 cm (Blumenbach, 2005). Accordingly, the joints of a calf of 50 cm length show a distance of 38 pixels to each other. Digitisation of both end points result in a theoretical standard deviation of √2*1.3 cm=1.8 cm, which corresponds with a relative measurement error of 4% of the segment length (Arndt, Brüggemann, Virmavirta, & Komi, 1995). Hence, adjustments for kinematic data acquisition need to be optimal, which is highly challenging under ski field conditions. Thus methodological validation is required to ensure the best possible accuracy of any results obtained. Little work has investigated the validation of kinematic parameters in the field of snow sports (Darlow, 2007; Klous, Schwameder, & Mueller, 2006a; Schwirtz, Boesl, Hartung, & Huber, 2006). Therefore, the current sub-study focused on the validation of a mobile high speed video system (MHSVS) and motion capture procedures for field studies in snow sports.

Methods

This study consisted of two parts, 1) a laboratory-based validation of system precision and 2) the investigation of a possible difference in precision between skin-based markers (condition SB) versus markers attached to a racing suit (condition RS). This testing session took place in the Biomechanics laboratory of The University of Auckland, New Zealand. Two participants, free of neuromusculoskeletal injury, wore a racing suit with 19 reflective markers with a diameter of 40 mm attached on the bony landmarks (Figure 38). The MHSVS (Simi Motion, Germany) was set up with four cameras recording at 100 Hz and a resolution of 656 by 490 pixels with an angle of greater than 60 degrees of any two optical axes and a distance of five meters from the participant. A Vicon-MX (Vicon Motion Systems Inc., UK) system was set up for comparison with 8 infrared cameras recording at the same frequency of 100 Hz. Both camera systems were calibrated using the same calibration cube. The touch down of a dropped marker was regarded as the reference frame for synchronisation of the systems. For part I, static standing trials were collected with the RS before specific movement trials. The analysis protocol included the identification of inter-marker distances (Figure 38). After the first upright static trial (UST), participants remained in a position with a 90° and 45° knee angle for an additional two static recordings. For the dynamic trials (DT), participants performed 10 squat jumps with slight twists, rotating the body position 45° to the sides during each jump in line with marks on the floor. This simulated a realistic range of movement and movement speed that complies with a field testing situation in freestyle skiing (Mustonen, 2007). For comparison of the two systems the trajectories of all points were measured by both systems and compared. Therefore the automatic tracking function was used which pinpoints the middle of the marker automatically for digitisation. This included knee joint centre position (KJCPos) and ankle joint centre position (AJCPos).
Following this, both the postures with pre-determined knee angles and the UST were recorded again in order to allow for evaluation of marker movement during specific body movements and sustained changes in marker distances at rest. Each participant repeated this procedure twice within the session. After three trials both participants repeated the same procedure in the SB-condition for part II in order to identify possible differences in marker distances during both static and dynamic trials (condition RS vs. SB). Descriptive statistics were employed to show the between-trial variability within the sessions for each inter-marker distance by using the averages and standard deviations from the Microsoft Excel Software. The Bland Altman plots were used to illustrate the comparison between the systems.

Results

The KJCPos deviated about 3 mm in a positive and 4 mm in a negative direction with respect to the vertical axis (Figure 39). The AJCPos showed a greater range of deviation with up to 18 mm.The measured differences between the markers in the RS condition varied 28 mm between the knee and the trochanter and 20 mm between the knee and the ankle during the squatting task. The difference between joint markers such as the knee or the ankle and non joint markers such as the tibia or the thigh remained with 4-6 mm variability relatively constant (Figure 40). The measured differences between the markers in the SB condition varied about 47 mm between the knee and the trochanter and 35 mm between the knee and the ankle during the squatting movement (Table 6). Joint markers versus non-joint markers showed a variability of about 21 mm. The distance between the markers in the UST and the pre-determined 45° and 90° knee angle position varied slightly from the first to the second trial with a 1-5 mm difference but not from the second to the third trial (1 to max. 3 mm variability). The marker distances in the SB condition varied constantly between 2-5 mm during all static trials depending on the marker combination. All data were processed by two persons and the mean values served as the illustrated results (Table 6)

Discussion

Results demonstrate a high precision of the Simi Motion system and a reasonably acceptable marker movement on the race suit compared to skin based markers. The Vicon system is deemed to be the ‗Gold Standard‘ with a precision of approximately 2 mm (Dempsey et al., 2007). The Simi Motion Software‘s accuracy depends on several parameters that need to be systematically analysed. The differences in resolution of both systems should be considered as an influencing factor.
Klous and associates (2007; 2006b; 2006), for their methodological experiments, used a five camera set up for kinematic data collection at 50 Hz, which was accurate with error margins of one to two cm on a measuring image of 20 m. Square tapes of 2.5 cm by 2.5 cm were used as markers. Briggs et al (2003)reported mean absolute error values of 7 mm for calculated distance measured by a Simi Motion system. These results comply with the current study based on values gained in the laboratory environment.
Skin markers are deemed to represent the skeletal knee joint motion regarding flexion and extension. However, according to Reinschmidt et al.(1997) internal-external rotation and adduction-abduction display larger errors using skin markers during running. Alexander & Andriacchi (2001) and Capozzo et al (1996) showed a skin based marker movement of up to 40 mm during performance. However, skin based markers are considered the most reliable, and are used as the reference in this study. Several studies have used bone pin markers and derived marker trajectories to verify soft tissue movement during walking. It has been noted that skin movement artifacts vary across subjects (Benoit et al., 2007). Holden (1997) measured errors for maximum translation of about 13 mm and rotational errors of up to 8° around the calf‘s longitudinal axis. These changes in kinematics caused kinetic changes of 9 Nm in knee joint abduction-adduction moment and up to 39 N for medio-lateral knee forces. Other studies reported maximum rotational errors of 4° at the calf which elicited errors of 4 Nm in the abduction-adducton moment (Manal, McClay, Richards, Galinat, & Stanhope, 2002). The utilisation of bone pin markers, however, is not feasible for snow sports. Due to safety issues, skin markers are impossible to use in the ski field. Hence, markers attached to clothing are necessary and the resulting limitations due to marker movement must be tolerated. In the current study, the race suit based markers show the most obvious marker movement between two joint markers such as the knee and the trochanter during the squatting task. However, the variability of marker distances between the second and third static trial including pre-determined squat positions are negligible. This suggests after the first movements the racing suit fits more appropriately around the body. This information shows that all markers should be readjusted or attached following a warm up to improve precision. The marker movement of the skin based and race suit based markers was identical following the first dynamic trial for all static trials. The RS condition seems to show less marker movement than the SB condition. However, as can be assumed the RS condition is likely to have greater marker movement. The markers on the skin are assumed to move relatively accurately with the skin except the known reported errors due to wobble effects. In contrast, the race suit is malleable and moves with the skin. The SB condition showed 15-20 mm greater marker movement compared to the RS condition. Therefore a reduced marker movement in the RS condition can be deceptive, because it should be considered that the RS markers move slightly with the race suit even though the distance does not vary as much as in the SB condition. Thus it is expected that, opposed to common belief, less marker movement as shown for the RS condition needs to be interpreted as an indicator for slightly less precision in the measurements. However, considering the need for attire in snow sport field tests and the relative similarity of the results, the utilisation of markers on a race suit appeared to be applicable and recommendable for outdoor applications in snow sports.

READ  Adjustment of Children and Adolescents in Stepfamilies

Chapter 7: Model development and verifications

Study II and III of the current project describe a comprehensive set of equipment and methods including high speed cameras and calibration techniques as presented within study I. This chapter contains an outline of the kinetic measurement, a ski boot stiffness test and an introduction of a specific computer modelling approach. Furthermore, previous validations of the original model that the current model is based on will be briefly presented as well as a sensitivity analysis of the current modelling application.

Kinetic measurement

7.1.1 Ski force plate design
The six-component dynamometer for measuring ground reaction forces is made of aluminium, weighs 2 kg and is 36 mm high × 62 mm wide. It consists of three different parts, specifically the boot sole adapter, the load cell and the binding adapter (Figure 41).
The ski boot is attached to the adapter with a buckle. The load cell is situated between the other two parts and contains six one-dimensional force sensors and a mechanical uncoupling device. Transducer elements are designed as shear beams instrumented with strain gauges. Due to the specifically defined positions of the sensors, it is possible to calculate all moments (Mx, My and Mz) and pinpoint the centre of force in the x/y-plane, using specific equations (Equation 1). C is the calibration factor for the conversion of the recorded signal into the acting forces. The factors of the first three rows are N/mV and convert the sensor signals into the forces Fx, Fy and Fz. Rows four to six are Nm/mV and convert the sensor signals into the moments Mx, My and Mz. The device has an external power supply for the gauges and a data logger (miniature computer) is connected to the device in order to transfer collected data for further processing. The binding adapter, as the third part of the system, is adjustable in length from 290 to 390 mm and clicks the whole system on the binding (Kiefmann et al., 2006).
The coordinate system of the dynamometers was defined x-direction; anteroposterior (alongside the ski), y-direction; medio-lateral (across the ski) and the z- axis vertical relative to the ski. The midpoint of the sensor assembly was the origin 7.1.2 Ski force plate accuracy
A laboratory-based validation study has been conducted by German collaborators to assess the accuracy of the calibration matrix. Two different performance testing rounds were conducted, each followed by design improvements and modifications. The first test was a static load assessment to provide information about the capability of the device as well as support for the design and data processing refinements prior to on-snow field testing. For the second testing session dynamic loads of different magnitudes and directions were applied to examine the device‘s accuracy under relevant conditions to those encountered in the field tests (Figure 42). An algorithm was programmed using MATLAB software to enable the calculation of the calibration factors from raw data. This procedure aimed at using the calibration matrix to derive the sensor signals to the actual loads. Figure 42 displays one of the components measured by the ski force plate in comparison to a floor mounted Kistler force plate as a reference in the laboratory environment. The lateral force Fy and the longitudinal moment Mz showed minor inconsistencies (Kiefmann et al., 2006) The overall accuracy was ± 4 N and ± 0.1 Nm, which is within the acceptable range for force plate errors of GRF < 2º as indicated by Lewis (2007).

CHAPTER 1: INTRODUCTION
1.1 RELEVANCE OF THE PROJECT
1.2 BACKGROUND ON SKIING1.3 SUMMARY
CHAPTER 2: SKIING INJURIES AND RISK FACTORS
2.1 EPIDEMIOLOGY
2.2 RISK FACTORS FOR KNEE INJURIES IN SKIIN
2.3 INJURY PREVENTION STRATEGIES
2.4 SUMMARY
CHAPTER 3: BIOMECHANICAL INVESTIGATIONS OF SKIING TECHNIQUE
3.1 KINEMATIC ANALYSIS
3.2 KINETIC ANALYSIS
3.3 COMPUTER SIMULATION – A PROMISING APPROACH TO AID INJURY PREVENTION IN SKIING?
3.4 SUMMARY
CHAPTER 4: AIMS AND OUTLINE
4.1 GENERAL RESEARCH INTENTION
4.2 AIMS
4.3 METHODOLOGICAL OVERVIEW
CHAPTER 5: STUDY ONE: SHORT TERM ADAPTATION EFFECT OF A PROTOTYPE FORCE PLATE ON KINEMATICS AND PERCEPTION IN MOGUL SKIING
5.1 INTRODUCTION AND PURPOSE
5.2 HYPOTHESES
5.3 METHODS
5.4 RESULTS
5.5 DISCUSSION
5.6 CONCLUSION
CHAPTER 6: VALIDATION OF THE KINEMATIC MEASUREMENTS
6.1 METHODS
6.2 RESULTS
6.3 DISCUSSION
CHAPTER 7: MODEL DEVELOPMENT AND VERIFICATIONS
7.1 KINETIC MEASUREMENT
7.2 SKI BOOT STIFFNESS TESTS
7.3 ANYBODY MODELING SYSTEM
CHAPTER 8: STUDY TWO: ON MOUNTAIN DATA COLLECTION – VERIFICATION OF A TESTING APPROACH INCLUDING AN ASSESSMENT OF A SKI BOOT MODIFICATION ON BIOMECHANICAL PARAMETERS IN FREESTYLE SKIING
8.1 INTRODUCTION AND PURPOSE
8.2 HYPOTHESES
8.3 METHODS
8.4 RESULTS
8.5 DISCUSSION
8.6 CONCLUSION
CHAPTER 9: STUDY THREE: ON MOUNTAIN DATA COLLECTION ON MOGUL SKIING TECHNIQUE – IMPROVED SKI BOOT INTERVENTION
9.1 INTRODUCTION AND PURPOSE
9.2 HYPOTHESES
9.3 METHODS
9.4 RESULTS
9.5 DISCUSSION
9.6 CONCLUSION
CHAPTER 10: GENERAL DISCUSSION
10.1 COMPARISON WITH MOGUL SKIING LITERATURE
10.2 COMPARISON WITH GENERAL SKIING LITERATURE
10.3 COMPARISON WITH OTHER MOVEMENTS
10.4 IMPLICATIONS FOR INJURY PREVENTION STRATEGIES
10.5 COMPUTER MODELLING
10.6 LIMITATIONS OF THE CURRENT PROJECT
CHAPTER 11: GENERAL CONCLUSION
11.1 CONTRIBUTION OF THE CURRENT PROJECT
11.2 FUTURE DIRECTIONS
EPILOGUE
GET THE COMPLETE PROJECT

Related Posts