The Knee 21 (2014) 1063–1068
Contents lists available at ScienceDirect
The Knee
Quantitative measurement of lower limb mechanical alignment and coronal knee laxity in early flexion David F. Russell a,b,1, Angela H. Deakin b, Quentin A. Fogg c, Frederic Picard b,⁎ a b c
Faculty of Biomedical and Life Sciences, Thomson Building, University of Glasgow, University Avenue, Glasgow G12 8QQ, United Kingdom Golden Jubilee National Hospital, Agamemnon Street, Clydebank, West Dunbartonshire G81 4DY, United Kingdom William Hunter Lecturer in Anatomy, Faculty of Biomedical and Life Sciences, Thomson Building, University of Glasgow, University Avenue, Glasgow G12 8QQ, United Kingdom
a r t i c l e
i n f o
Article history: Received 22 June 2013 Received in revised form 7 July 2014 Accepted 9 July 2014 Keywords: Mechanical alignment Non-invasive Ligamentous laxity Total knee arthroplasty Image-free navigation
a b s t r a c t Background: Non-invasive quantification of lower limb alignment using navigation technology is now possible throughout knee flexion owing to software developments. We report the precision and accuracy of a non-invasive system measuring mechanical alignment of the lower limb including coronal stress testing of the knee. Methods: Twelve cadaveric limbs were tested with a commercial invasive navigation system against the noninvasive system. Coronal mechanical femorotibial (MFT) alignment was measured with no stress, then 15 Nm varus and valgus applied moments. Measurements were recorded at 10° intervals from extension to 90° flexion. At each flexion interval, coefficient of repeatability (CR) tested precision within each system, and limits of agreement (LOA) tested agreement between the two systems. Limits for CR & LOA were set at 3° based on requirements for surgical planning and evaluation. Results: Precision was acceptable throughout flexion in all conditions of stress using the invasive system (CR ≤ 1.9°). Precision was acceptable using the non-invasive system from extension to 50° flexion (CR ≤ 2.4°), beyond which precision was unacceptable (N3.4°). With no coronal stress applied, agreement remained acceptable from extension to 40° (LOA ≤2.4°), and when 15 Nm varus or valgus stress was applied agreement was acceptable from extension to 30° (LOA ≤ 2.9°). Higher angles of knee flexion had a negative impact on precision and accuracy. Conclusion & clinical relevance: The non-invasive system provides reliable quantitative data in-vitro on coronal MFT alignment and laxity in the range relevant to assessment of collateral ligament injury, pre-operative planning of arthroplasty and flexion instability following arthroplasty. In-vivo validation should be performed. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Understanding lower limb alignment in health, disease and in the surgical setting is crucial to lower limb reconstruction, both in sports medicine and knee arthroplasty. The ability to quantify lower limb alignment in both coronal and sagittal planes is important to preoperative decision making, planning and post-operative evaluation. ‘Long-leg radiographs’ or ‘hip-knee-ankle’ radiographs remain the gold-standard using the definition by Moreland et al. [1] for mechanical alignment in the coronal plane, allowing measurement of the mechanical femorotibial (MFT) angle. Long-leg radiographs whilst far superior to short leg radiographs [2] are prone to rotational error [3,4] especially in patients with flexion contracture [4]. Other methods used routinely ⁎ Corresponding author at: Department of Orthopaedics, Golden Jubilee National Hospital. Tel.: +44 141 951 5000. E-mail addresses: [email protected] (D.F. Russell), [email protected] (F. Picard). 1 Tel.: +447779 153 424.
http://dx.doi.org/10.1016/j.knee.2014.07.008 0968-0160/© 2014 Elsevier B.V. All rights reserved.
in a clinical setting to assess alignment may provide a means of appreciating disease progression but are not reliable enough for surgical planning; these include visual assessment with use of a goniometer [5], and other landmark based methods [6–8]. Image-free navigation technology has been thoroughly validated in acquisition of coronal and sagittal mechanical alignments [9,10]. Clinical studies, have demonstrated that image-free navigation gives consistent accurate placement of components and limb alignment [11–13], supporting the intra-operative use of navigation systems [14]. However computer navigation presently depends on invasive placement of trackers meaning that it is limited to the operating theatre. The ability to use this type of system to quantify these parameters in a noninvasive manner with minimal adverse consequence to the patient would allow accurate assessment in a clinic situation. Clarke et al. validated a non-invasive adaptation of image-free navigation for measuring the MFT angle in extension and with applied stress, and for early flexion–extension measurement [15–17]. These early results are promising but do not include validation of the measurement of the MFT angle with the knee beyond 10° of flexion. Initial
1064
D.F. Russell et al. / The Knee 21 (2014) 1063–1068
results from a pilot study on cadavers were repeatable in measuring the MFT angle and knee laxity from extension to 40° [18]; however, the methodology in this study was limited in terms of consistency in limb positioning and forces applied and use of embalmed cadaveric materials. Therefore the primary aim in this study was to determine the reliability, precision and accuracy of a non-invasive adaptation of image-free navigation technology in determining the MFT angle of the lower limb in flexion. A secondary aim was to determine the reliability, precision and accuracy of the non-invasive measurements of maximum extension and flexion.
2. Materials and methods Twelve fresh cadaveric lower limbs were used. A similar experiment setup used to test anteroposterior laxity in the knee has already been reported and was carried out on the same specimens [19]. A single investigator (DFR) performed all tests. A FDA, CE validated image-free navigation system was used for the study (Orthopilot, BBraun Aesculap, Tuttlingen, Germany). The hip, knee and ankle centres were registered during both invasive and non-invasive testing using the same software; algorithms used to register the lower limb were identical to those of validated, commercially available software currently used in image-free computer-assisted high tibial osteotomy surgery which permits noninvasive registration of anatomical landmarks of each joint. Using this software, all points requiring digitisation were located on the skin without the need for incision(s), and the kinematic centres of the hip, knee and ankle were identified through recording a series of prescribed lower-limb movements [20]. Two methods of passive tracker fixation were used: commercially available bone screws and a baseplate secured by fabric strapping (Fig. 1). In order to minimise soft-tissue artefacts from the limb resting on the laboratory table, it was necessary to suspend the limb. A bicortical eyelet screw (length of 20 mm, width of 75 mm, manufacturer part no. N330, B&Q, UK) was inserted into the proximal femur to suspend the thigh from a stand, maintaining a hip flexion of 20° (Fig. 2). To create a foot support, a loop of cord was secured proximal to the metatarsal heads; this maintained knee flexion angle (Fig. 3). In order to apply a standardised varus/valgus moment of 15 Nm, unicortical 7.5 mm eyelet screws were inserted in the medial and lateral sides of the distal tibia, aligned in the coronal plane, and at a set distance from the joint line, depending on the length of the lower limb (Fig. 3). Side supports (Fig. 2) and manual support (Fig. 3) were both employed during coronal stress testing to stabilise the knee. Fifteen Nm varus/valgus stress is similar to that exerted during clinical examination [21–23]. A force transducer (model 251066, Silverline, Somerset, UK, CE certified)
Fig. 2. Photograph of limb set up, side supports to supplement manual support against varus and valgus stress and the femoral pin used to suspend the thigh are labelled. Trackers present in this photograph are placed on mounts secured by bone screws.
was attached to these eyelet screws in order to apply a discrete force in the coronal plane (Fig. 3). To minimise soft tissue creep throughout the experiment a protocol of 24 hip circumductions, 24 full flexion and extensions of the limb & 24 manual varus/valgus stresses was performed prior to testing. Experiment protocol (given below) was first performed using invasively mounted trackers, then repeated using non-invasively mounted trackers. The invasively mounted trackers were secured using bicortical bone screws, one in the anterior distal femur, and one in the anterior proximal tibia as is standard practise during intraoperative image-free navigation (Fig. 2). The non-invasive trackers were secured 8 cm proximal to the proximal pole of patella overlying the distal vastus medialis obliquus muscle, and 4 cm distal to the tibial tuberosity, again on the medial aspect of the lower limb to maximise tracker exposure to the localising camera (Fig. 1). The only difference in the protocol when using the non-invasive method was that the trackers and fabric strapping were removed and replaced between registrations.
Fig. 1. Photograph showing the non-invasive method of optical tracker mounting secured with fabric strapping.
D.F. Russell et al. / The Knee 21 (2014) 1063–1068
1065
Table 1 Reliability of measuring the MFT angle using the invasive and non-invasive methods whilst applying no coronal stress, valgus stress and varus stress as given by the intraclass correlation coefficient (ICC). Data are presented as mean (range).
No stress 15 Nm valgus stress 15 Nm varus stress
Fig. 3. Distal lower limb, label ‘A’ depicts lateral distal tibial eyelet screw with the force transducer attached and held in position, the other hand is supporting the knee. Label ‘B’ highlights the medial distal tibial eyelet screw and label ‘C’ demonstrates the foot support.
3. Experiment protocol The lower limb was registered with the software identifying the hip, knee and ankle centres and the axis of rotation of the knee. The mechanical femorotibial (MFT) angle with the limb in maximum extension was recorded. Measurement of maximum extension and flexion angle was recorded. The investigator was not blinded to measurement of maximum flexion or extension angle and force application was not standardised. Once these angles were recorded, the display of the MFT angle was covered to blind the investigator to this variable; it was necessary that the flexion angle was still visible on the monitor throughout in order to achieve the required flexion intervals. Neutral or unstressed MFT angle i.e. the natural coronal alignment of the lower limb was recorded at 10° intervals from full extension to 90° flexion. A second registration was then performed and the MFT angle in extension was recorded. Providing the MFT angle from the second registration was within 2° of the first, the coronal alignment display on the computer monitor was covered once again and the MFT angle recorded using no stress was applied to the leg at 10° intervals. At the end of recording, side supports were fitted to the table at the level of the proximal thigh to provide medial & lateral support (Fig. 2). The MFT angle was measured at each 10° interval from maximum extension to maximum flexion with first 15 Nm varus and then 15 Nm valgus stress applied. These measurements were repeated. This meant that two measurements were taken at each condition of stress and flexion/extension angles. Statistical analysis was carried out to calculate the reliability, precision and accuracy of the measurement of coronal alignment with and without stress and for maximum flexion and extension. Reliability within each method of tracker fixation used in was analysed by calculating intraclass correlation coefficient (ICC) [24]. A coefficient of ≥0.75 demonstrates very good reliability [25,26]. Calculation of ICC was
Invasive method ICC
Non-invasive method ICC
0.976 (0.957–0.988) 0.995 (0.988–1.000) 0.998 (0.992–1.000)
0.917 (0.785–0.988) 0.980 (0.950–0.995) 0.989 (0.959–0.997)
performed using an IBM SPSS® Statistics 17.0 software (IBM Corp., Armonk, NY, USA). The coefficient of repeatability (CR) was calculated to demonstrate precision of measurements within each method of tracker fixation [27]. The CR defines interval within which 95% of differences between two measurements lie, a CR of 3° conveys that 95% of all measurements are within a range of ± 1.5°. This would be substantially better than current clinical methods of measuring the MFT angle and would be acceptable for pre-operative assessment of coronal alignment [16,28]. For the measurement of flexion/extension the acceptable CR was set to be 5°, and ± 2.5° based on the reported accuracy of measurement with goniometers [29]. To assess the accuracy of the non-invasive system 95% limits of agreement (LOA) between the invasive and non-invasive measurements were determined using the corrected standard deviation of the differences [27]. Acceptable LOA were defined as 3° (± 1.5°) for the MFT angle and 5° (±2.5°) for flexion/extension. CR & LOA calculations were performed using Microsoft Excel® (Microsoft Corp, Redmond, Washington, USA). 4. Results Mean age of specimens was 80.5 years (range 65–91 years), five were female. Mean fixed flexion contracture (maximum extension) for the 12 limbs was 5.8° (range 6° hyperextension to 15° fixed flexion). In full extension, mean mechanical femorotibial (MFT) angle was 2.4° valgus (range 3° varus to 6° valgus). The mean varus and valgus laxities with the knee in full extension were 1.8° (range 0° to 4°) and 1.5° (range 0° to 4°) respectively. Mean maximum flexion for the 12 specimens was 132° (range 111° to 155°). For MFT angle with no applied stress and applying 15 Nm of varus or valgus stress the mean and range of intraclass correlation coefficients (ICCs) were acceptable (N0.75) throughout the range of flexion tested indicating good reliability for both the invasive and non-invasive methods (Table 1). The precision given by the coefficient of repeatability (CR) using the invasive system to measure the MFT angle with no stress was acceptable (CR b 3°) throughout flexion (Fig. 4, Table 2). The precision using the non-invasive system was acceptable from extension to 50° (Fig. 4, Table 2). The CRs for both the invasive and non-invasive methods measuring the MFT angle whilst 15 Nm of varus or valgus stress was applied were acceptable (b3°) throughout the range of flexion tested (Table 2). Accuracy of the non-invasive system as given by the limits of agreement (LOA) between the invasive and non-invasive methods was acceptable (LOA b3°) for measuring
Fig. 4. Coefficient of repeatability (CR) for invasive and non-invasive measurement of MFT angle throughout the range of knee flexion with no stress applied to the limb (CR b3° is acceptable).
1066
D.F. Russell et al. / The Knee 21 (2014) 1063–1068
Table 2 Precision of measuring the MFT angle throughout the range of flexion using the invasive and non-invasive methods whilst applying varus and valgus stresses as given by the coefficient of repeatability (CR). Data are presented as mean (range).
No stress 15 Nm valgus stress 15 Nm varus stress
Invasive method CR ( °)
Non-invasive method CR ( °)
1.3 (0.6–1.9) 0.6 (0.0–0.9) 0.3 (0.6–1.8)
3.0 (0.9–6.9) 1.2 (0.0–0.8) 1.1 (0.6–2.6)
the MFT angle from extension to 40° knee flexion when no stress was applied, and from extension to 30° knee flexion when 15 Nm varus or valgus stress was applied (Fig. 5). Regarding maximum flexion and extension, the invasive and non-invasive methods of tracker fixation displayed acceptable reliability of measurement with ICCs N 0.75 (Table 3). Both methods of tracker fixation gave acceptable precision measuring full extension and flexion (CR b 5°) (Table 4). Accuracy of the non-invasive system was on the limit (LOA = 5.1°) for measuring full extension but was acceptable for measuring full flexion (Table 4).
5. Discussion We aimed to determine the reliability, precision and accuracy of a non-invasive adaptation of image-free navigation technology in measuring the coronal and sagittal mechanical alignment of the lower limb in flexion by direct comparison with a commercially available invasive system. The non-invasive system demonstrated acceptable reliability, precision and accuracy when compared with the invasive system when measuring the coronal MFT angle from extension to 40° flexion with no coronal stress applied and extension to 30° flexion when 15 Nm varus or valgus stress was applied and when measuring maximum extension and flexion. A major limitation of the methodology is the use of cadaveric specimens; tissue quality, muscle tone and soft tissue artefacts will differ from the in-vivo setting. Although the investigator remained blinded to measurement of the coronal MFT angle, it was not possible to remain blinded to measurement of maximum extension and flexion. A further limitation to accuracy testing is that measurements using the invasive and non-invasive system were taken during two separate episodes of testing. This introduces error in accuracy testing but may also make accuracy testing between methods more robust. Standardised force application was not employed to position the limb in flexion, which may account for the small inconsistency in flexion angle reached. Although the precision of the non-invasive system was acceptable when measurements were taken with applied stress, for the unstressed measurements of the MFT angle precision became unacceptable beyond 50° flexion. A potential reason for this may be the inherent laxity of the knee joint in mid-flexion leading to slight variation in coronal alignment between episodes of positioning when no stress is being applied to the
Table 3 Reliability of measuring maximum extension and maximum flexion using the invasive and non-invasive methods as given by the intraclass correlation coefficient (ICC). Data are presented as mean (range).
Maximum extension Maximum flexion
Invasive method ICC
Non-invasive method ICC
0.93 (0.78–0.98) 1.00 (1.00–1.00)
0.94 (0.8–0.98) 1.00 (0.99–1.00)
limb, augmenting the effect of soft tissue artefact variability. However when stress is applied to the limb, the knee reaches the outer envelope of its laxity, this may be a more consistent position thus giving a more consistent alignment between measurement episodes. Soft-tissue artefacts affecting the non-invasive system are most likely the reason for decreasing accuracy of the non-invasive methods when measuring the MFT angle with increased knee flexion and coronal stress (Fig. 5). Application of coronal stress is likely to cause rotatory displacement of the bony anatomy which will be detected directly using invasive tracker mounting, but possibly not so well by skin mounted trackers. Clarke et al. [16] demonstrated high levels of reliability and precision using this device between investigators measuring the MFT angle in extension and when measuring coronal mechanical alignment in extension with no applied stress, varus and valgus stress respectively. The limits of agreement (b 3°) were similar to those found in this study. Measuring full extension, precision is inferior to that reported by Clarke [15]; their cohort exhibited hyperextension unlike the fixed flexion deformity observed in the cadaveric specimens used in this study. Knees in hyperextension may allow more consistent limb positioning as the flexion contracture and lack of muscle tone in the cadaveric specimens are likely to nullify stability inherent to the normal in-vivo extended knee joint as a result of a functioning ‘screw-home’ mechanism supported by active musculature [30]. Nonetheless, the noninvasive method of tracker fixation is superior to visual assessment which has been shown to be very unreliable [5] and to goniometry which has a margin of error of at least ±5° [29]. In knee arthroplasty, it is difficult to communicate findings from examination of the knee using examination techniques which are not standardised and cannot be quantified, far less make reliable recommendations as to how to proceed with management of soft-tissues during total knee replacement [31–33]. Short leg radiographs are inadequate and can even cause clinicians to completely misinterpret alignment [2]. Long leg radiographs are superior in determining lower limb mechanical alignment however they are prone to rotational error [3,34], do not allow dynamic assessment, and expose the patient to radiation. Ideally, clinicians would have access to a relatively inexpensive, reliable, precise and accurate method of assessing dynamic, weight-bearing mechanical
Fig. 5. Limits of agreement (LOA) representing agreement between the two methods of tracker fixation measuring MFT angle throughout flexion in conditions of: no coronal stress, 15Nm valgus stress and 15Nm varus stress (LOA b3° is acceptable).
D.F. Russell et al. / The Knee 21 (2014) 1063–1068 Table 4 Precision and accuracy of measuring maximum extension and maximum flexion using the invasive and non-invasive methods as given by the coefficient of repeatability (CR) and the limits of agreement (LOA).
Maximum extension Maximum flexion
Invasive method CR (°)
Non-invasive method CR (°)
Limits of agreement (°)
3.3 1.5
2.6 2.4
5.1 3.4
alignment and knee joint laxity in extension and early flexion before and after total knee arthroplasty. Prior to surgery, this would allow a detailed assessment of kinematics and planning of tissue release & resection. Following surgery, such a device would be ideal for audit and research as described above, as well as assessing problematic knees such as those describing flexion instability which accounts for a significant proportion of revision surgery [35,36]. In evaluation of soft-tissue/sports injuries, collateral ligament laxity testing is currently limited in terms of using devices to quantify coronal laxity in the examination of joint space opening [37,38]. This device could allow the quantification or grading of collateral ligament injury in a consistent manner and is within the range relevant to laxity testing such as that used in soft-tissue management algorithms for total knee replacement [39–41]. This study would support that it is possible to use optical tracking of skin-mounted markers as a basis for a reliable knee assessment system in early knee flexion. It is accepted that the current prototype could be further developed to maximise efficiency in clinical use, however it has been used on patients and does allow more accurate dynamic assessment than that available to clinicians at the moment. With the relentless improvement in imaging technology it would not be unreasonable to envisage an optical system using small inexpensive cameras providing a permanent setup in orthopaedic clinic rooms making dynamic alignment assessment via optical tracking routine. 6. Conclusion The non-invasive method demonstrated satisfactory reliability, precision and accuracy for measuring lower limb alignment in the early part of knee flexion. This has applications for sports medicine and knee arthroplasty; and in pre-operative decision-making, planning and post-operative evaluation. Declaration of interest statement No financial support was given by any commercial body. The first author received a bursary for research to aid career development from the West of Scotland Orthopaedic Research Society. The authors received material and software support from BBraun Aesculap, Tuttlingen, Germany. Mr. Frederic Picard has separate licences and patents with BBraun. The Department of Orthopaedics at the Golden Jubilee National Hospital receives funding for research from a number of orthopaedic manufacturers including BBraun Aesculap, however no direct funding was received for this particular study. Acknowledgements We give special thanks to the technical and administrative staff at the Laboratory of Human Anatomy, University of Glasgow. We also thank Mr. Phil Cleary & Mr. Iain Freer for the excellent support provided in supplying the equipment to facilitate this study. References [1] Moreland JR, Bassett LW, Hanker GJ. Radiographic analysis of the axial alignment of the lower extremity. J Bone Joint Surg Am 1987;69:745–9.
1067
[2] van Raaij TM, Brouwer RW, Reijman M, Bierma-Zeinstra SM, Verhaar JA. Conventional knee films hamper accurate knee alignment determination in patients with varus osteoarthritis of the knee. Knee 2009;16:109–11. [3] Krackow KA, Pepe CL, Galloway EJ. A mathematical analysis of the effect of flexion and rotation on apparent varus/valgus alignment at the knee. Orthopedics 1990; 13:861–8. [4] Yaffe MA, Koo SS, Stulberg SD. Radiographic and navigation measurements of TKA limb alignment do not correlate. Clin Orthop Relat Res 2008;466:2736–44. [5] Shetty GM, Mullaji A, Lingaraju AP, Bhayde S. How accurate are orthopaedic surgeons in visually estimating lower limb alignment? Acta Orthop Belg 2011;77:638–43. [6] McDaniel G, Mitchell KL, Charles C, Kraus VB. A comparison of five approaches to measurement of anatomic knee alignment from radiographs. Osteoarthritis Cartilage 2010;18:273. [7] Hinman RS, May RL, Crossley KM. Is there an alternative to the full-leg radiograph for determining knee joint alignment in osteoarthritis? Arthritis Rheum 2006; 55:306–13. [8] Navali AM, Bahari LAS, Nazari B. A comparative assessment of alternatives to the full-leg radiograph for determining knee joint alignment. Sports Med Arthrosc Rehabil Ther Technol 2012;4:40. [9] Picard F. Computer assisted orthopaedics — the image free concept. Berlin: Pro BUSINESS GmbH; 2007. [10] Pearle AD, Solomon DJ, Wanich T, Moreau-Gaudry A, Granchi CC, Wickiewicz TL, et al. Reliability of navigated knee stability examination: a cadaveric evaluation. Am J Sports Med 2007;35:1315–20. [11] Kim YH, Park JW, Kim JS. Computer-navigated versus conventional total knee arthroplasty: a prospective randomized trial. J Bone Joint Surg Am 2012; 21:2017–24. [12] Matziolis G, Krocker D, Weiss U, Tohtz S, Perka C. A prospective, randomized study of computer-assisted and conventional total knee arthroplasty. Threedimensional evaluation of implant alignment and rotation. J Bone Joint Surg Am 2007;89:236–43. [13] Brin YS, Nikolaou VS, Joseph L, Zukor DJ, Antoniou J. Imageless computer assisted versus conventional total knee replacement. A Bayesian meta-analysis of 23 comparative studies. Int Orthop 2011;35:331–9. [14] DiGioia AM, Mor AB. Accuracy and validation for surgical navigation systems. AAOS: American Academy of Orthopaedic Surgeons; 2005. [15] Clarke JV. The non-invasive measurement of knee kinematics in normal, osteoarthritic and prosthetic knees. Strathprints: The University of Strathclyde Institutional Repository. University of Strathclyde; 2012. [16] Clarke JV, Riches PE, Picard F, Deakin AH. Non-invasive computer-assisted measurement of knee alignment. Comput Aided Surg 2012;17:29–39. [17] Clarke JV, Wilson WT, Wearing SC, Picard F, Riches PE, Deakin AH. Standardising the clinical assessment of coronal knee laxity. Proc Inst Mech Eng H 2012; 226:699–708. [18] Russell DF, Deakin AH, Fogg QA, Picard F. Non-invasive quantification of lower limb mechanical alignment in flexion. Comput Aided Surg 2014:1–7[Online open access, URL http://informahealthcare.com/doi/pdf/10.3109/10929088.2014.885566]. [19] Russell D, Deakin AH, Fogg QA, Picard F. Non-invasive quantification of lower limb mechanical alignment in flexion. Comput Aided Surg 2014. http://dx.doi.org/10.3109/ 10929088.2014.885566[submitted for publication]. [20] Saragaglia D, Mercier N, Colle PE. Computer-assisted osteotomies for genu varum deformity: which osteotomy for which varus? Int Orthop 2010;34:185–90. [21] Grood ES, Noyes FR, Butler DL, Suntay WJ. Ligamentous and capsular restraints preventing straight medial and lateral laxity in intact human cadaver knees. J Bone Joint Surg Am 1981;63:1257–69. [22] Stahelin T, Kessler O, Pfirrmann C, Jacob HA, Romero J. Fluoroscopically assisted stress radiography for varus–valgus stability assessment in flexion after total knee arthroplasty. J Arthroplasty 2003;18:513–5. [23] Wilson WT, Deakin AH, Picard F, Riches PE, Clarke JV. Standardising the assessment of coronal knee laxity. J Bone Joint Surg 2013;95(30):9. [24] Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull 1979;86:420–8. [25] Portney LG, Watkins MP. Foundations of clinical research: applications to practice. Norwalk, Conn.: Appleton & Lange; 1993 [26] Fleiss JL. Statistical methods for rates and proportions. New York: Wiley; 1981. [27] Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–10. [28] Picard F, Deakin AH, Clarke IV, Dillon JM, Kinninmonth AW. A quantitative method of effective soft tissue management for varus knees in total knee replacement surgery using navigational techniques. Proc Inst Mech Eng H 2007;221:763–72. [29] Edwards JZ, Greene KA, Davis RS, Kovacik MW, Noe DA, Askew MJ. Measuring flexion in knee arthroplasty patients. J Arthroplasty 2004;19:369–72. [30] Croce RV, Miller JP. Coactivation patterns of the medial and lateral hamstrings based on joint position and movement velocity during isokinetic movements. Electromyogr Clin Neurophysiol 2006;46:113–22. [31] Engh GA. The difficult knee: severe varus and valgus. Clin Orthop Relat Res 2003:58–63. [32] Ries MD, Haas SB, Windsor RE. Soft-tissue balance in revision total knee arthroplasty. J Bone Joint Surg Am 2003;85-A(Suppl. 1):S38–42. [33] Krackow KA. The technique of total knee arthroplasty. St. Louis: C.V. Mosby; 1990. [34] Mahaluxmivala J, Bankes MJ, Nicolai P, Aldam CH, Allen PW. The effect of surgeon experience on component positioning in 673 press fit condylar posterior cruciatesacrificing total knee arthroplasties. J Arthroplasty 2001;16:635–40. [35] Fowler PJ. The classification and early diagnosis of knee joint instability. Clin Orthop Relat Res 1980:15–21. [36] Fehring TK, Odum S, Griffin WL, Mason JB, Nadaud M. Early failures in total knee arthroplasty. Clin Orthop Relat Res 2001:315–8.
1068
D.F. Russell et al. / The Knee 21 (2014) 1063–1068
[37] Gwathmey Jr FW, Tompkins MA, Gaskin CM, Miller MD. Can stress radiography of the knee help characterize posterolateral corner injury? Clin Orthop Relat Res 2012;470:768–73. [38] LaPrade RF, Spiridonov SI, Coobs BR, Ruckert PR, Griffith CJ. Fibular collateral ligament anatomical reconstructions: a prospective outcomes study. Am J Sports Med 2010;38:2005–11. [39] Jenny JY. Coronal plane knee laxity measurement: is computer-assisted navigation useful? Orthop Traumatol Surg Res 2010;96:583–8.
[40] Lehnen K, Giesinger K, Warschkow R, Porter M, Koch E, Kuster MS. Clinical outcome using a ligament referencing technique in CAS versus conventional technique. Knee Surg Sports Traumatol Arthrosc 2011;19:887–92. [41] Picard F, Deakin AH, Clarke JV, Dillon JM, Gregori A. Using navigation intraoperative measurements narrows range of outcomes in TKA. Clin Orthop Relat Res 2007; 463:50–7.
Contents lists available at ScienceDirect
The Knee
Quantitative measurement of lower limb mechanical alignment and coronal knee laxity in early flexion David F. Russell a,b,1, Angela H. Deakin b, Quentin A. Fogg c, Frederic Picard b,⁎ a b c
Faculty of Biomedical and Life Sciences, Thomson Building, University of Glasgow, University Avenue, Glasgow G12 8QQ, United Kingdom Golden Jubilee National Hospital, Agamemnon Street, Clydebank, West Dunbartonshire G81 4DY, United Kingdom William Hunter Lecturer in Anatomy, Faculty of Biomedical and Life Sciences, Thomson Building, University of Glasgow, University Avenue, Glasgow G12 8QQ, United Kingdom
a r t i c l e
i n f o
Article history: Received 22 June 2013 Received in revised form 7 July 2014 Accepted 9 July 2014 Keywords: Mechanical alignment Non-invasive Ligamentous laxity Total knee arthroplasty Image-free navigation
a b s t r a c t Background: Non-invasive quantification of lower limb alignment using navigation technology is now possible throughout knee flexion owing to software developments. We report the precision and accuracy of a non-invasive system measuring mechanical alignment of the lower limb including coronal stress testing of the knee. Methods: Twelve cadaveric limbs were tested with a commercial invasive navigation system against the noninvasive system. Coronal mechanical femorotibial (MFT) alignment was measured with no stress, then 15 Nm varus and valgus applied moments. Measurements were recorded at 10° intervals from extension to 90° flexion. At each flexion interval, coefficient of repeatability (CR) tested precision within each system, and limits of agreement (LOA) tested agreement between the two systems. Limits for CR & LOA were set at 3° based on requirements for surgical planning and evaluation. Results: Precision was acceptable throughout flexion in all conditions of stress using the invasive system (CR ≤ 1.9°). Precision was acceptable using the non-invasive system from extension to 50° flexion (CR ≤ 2.4°), beyond which precision was unacceptable (N3.4°). With no coronal stress applied, agreement remained acceptable from extension to 40° (LOA ≤2.4°), and when 15 Nm varus or valgus stress was applied agreement was acceptable from extension to 30° (LOA ≤ 2.9°). Higher angles of knee flexion had a negative impact on precision and accuracy. Conclusion & clinical relevance: The non-invasive system provides reliable quantitative data in-vitro on coronal MFT alignment and laxity in the range relevant to assessment of collateral ligament injury, pre-operative planning of arthroplasty and flexion instability following arthroplasty. In-vivo validation should be performed. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Understanding lower limb alignment in health, disease and in the surgical setting is crucial to lower limb reconstruction, both in sports medicine and knee arthroplasty. The ability to quantify lower limb alignment in both coronal and sagittal planes is important to preoperative decision making, planning and post-operative evaluation. ‘Long-leg radiographs’ or ‘hip-knee-ankle’ radiographs remain the gold-standard using the definition by Moreland et al. [1] for mechanical alignment in the coronal plane, allowing measurement of the mechanical femorotibial (MFT) angle. Long-leg radiographs whilst far superior to short leg radiographs [2] are prone to rotational error [3,4] especially in patients with flexion contracture [4]. Other methods used routinely ⁎ Corresponding author at: Department of Orthopaedics, Golden Jubilee National Hospital. Tel.: +44 141 951 5000. E-mail addresses: [email protected] (D.F. Russell), [email protected] (F. Picard). 1 Tel.: +447779 153 424.
http://dx.doi.org/10.1016/j.knee.2014.07.008 0968-0160/© 2014 Elsevier B.V. All rights reserved.
in a clinical setting to assess alignment may provide a means of appreciating disease progression but are not reliable enough for surgical planning; these include visual assessment with use of a goniometer [5], and other landmark based methods [6–8]. Image-free navigation technology has been thoroughly validated in acquisition of coronal and sagittal mechanical alignments [9,10]. Clinical studies, have demonstrated that image-free navigation gives consistent accurate placement of components and limb alignment [11–13], supporting the intra-operative use of navigation systems [14]. However computer navigation presently depends on invasive placement of trackers meaning that it is limited to the operating theatre. The ability to use this type of system to quantify these parameters in a noninvasive manner with minimal adverse consequence to the patient would allow accurate assessment in a clinic situation. Clarke et al. validated a non-invasive adaptation of image-free navigation for measuring the MFT angle in extension and with applied stress, and for early flexion–extension measurement [15–17]. These early results are promising but do not include validation of the measurement of the MFT angle with the knee beyond 10° of flexion. Initial
1064
D.F. Russell et al. / The Knee 21 (2014) 1063–1068
results from a pilot study on cadavers were repeatable in measuring the MFT angle and knee laxity from extension to 40° [18]; however, the methodology in this study was limited in terms of consistency in limb positioning and forces applied and use of embalmed cadaveric materials. Therefore the primary aim in this study was to determine the reliability, precision and accuracy of a non-invasive adaptation of image-free navigation technology in determining the MFT angle of the lower limb in flexion. A secondary aim was to determine the reliability, precision and accuracy of the non-invasive measurements of maximum extension and flexion.
2. Materials and methods Twelve fresh cadaveric lower limbs were used. A similar experiment setup used to test anteroposterior laxity in the knee has already been reported and was carried out on the same specimens [19]. A single investigator (DFR) performed all tests. A FDA, CE validated image-free navigation system was used for the study (Orthopilot, BBraun Aesculap, Tuttlingen, Germany). The hip, knee and ankle centres were registered during both invasive and non-invasive testing using the same software; algorithms used to register the lower limb were identical to those of validated, commercially available software currently used in image-free computer-assisted high tibial osteotomy surgery which permits noninvasive registration of anatomical landmarks of each joint. Using this software, all points requiring digitisation were located on the skin without the need for incision(s), and the kinematic centres of the hip, knee and ankle were identified through recording a series of prescribed lower-limb movements [20]. Two methods of passive tracker fixation were used: commercially available bone screws and a baseplate secured by fabric strapping (Fig. 1). In order to minimise soft-tissue artefacts from the limb resting on the laboratory table, it was necessary to suspend the limb. A bicortical eyelet screw (length of 20 mm, width of 75 mm, manufacturer part no. N330, B&Q, UK) was inserted into the proximal femur to suspend the thigh from a stand, maintaining a hip flexion of 20° (Fig. 2). To create a foot support, a loop of cord was secured proximal to the metatarsal heads; this maintained knee flexion angle (Fig. 3). In order to apply a standardised varus/valgus moment of 15 Nm, unicortical 7.5 mm eyelet screws were inserted in the medial and lateral sides of the distal tibia, aligned in the coronal plane, and at a set distance from the joint line, depending on the length of the lower limb (Fig. 3). Side supports (Fig. 2) and manual support (Fig. 3) were both employed during coronal stress testing to stabilise the knee. Fifteen Nm varus/valgus stress is similar to that exerted during clinical examination [21–23]. A force transducer (model 251066, Silverline, Somerset, UK, CE certified)
Fig. 2. Photograph of limb set up, side supports to supplement manual support against varus and valgus stress and the femoral pin used to suspend the thigh are labelled. Trackers present in this photograph are placed on mounts secured by bone screws.
was attached to these eyelet screws in order to apply a discrete force in the coronal plane (Fig. 3). To minimise soft tissue creep throughout the experiment a protocol of 24 hip circumductions, 24 full flexion and extensions of the limb & 24 manual varus/valgus stresses was performed prior to testing. Experiment protocol (given below) was first performed using invasively mounted trackers, then repeated using non-invasively mounted trackers. The invasively mounted trackers were secured using bicortical bone screws, one in the anterior distal femur, and one in the anterior proximal tibia as is standard practise during intraoperative image-free navigation (Fig. 2). The non-invasive trackers were secured 8 cm proximal to the proximal pole of patella overlying the distal vastus medialis obliquus muscle, and 4 cm distal to the tibial tuberosity, again on the medial aspect of the lower limb to maximise tracker exposure to the localising camera (Fig. 1). The only difference in the protocol when using the non-invasive method was that the trackers and fabric strapping were removed and replaced between registrations.
Fig. 1. Photograph showing the non-invasive method of optical tracker mounting secured with fabric strapping.
D.F. Russell et al. / The Knee 21 (2014) 1063–1068
1065
Table 1 Reliability of measuring the MFT angle using the invasive and non-invasive methods whilst applying no coronal stress, valgus stress and varus stress as given by the intraclass correlation coefficient (ICC). Data are presented as mean (range).
No stress 15 Nm valgus stress 15 Nm varus stress
Fig. 3. Distal lower limb, label ‘A’ depicts lateral distal tibial eyelet screw with the force transducer attached and held in position, the other hand is supporting the knee. Label ‘B’ highlights the medial distal tibial eyelet screw and label ‘C’ demonstrates the foot support.
3. Experiment protocol The lower limb was registered with the software identifying the hip, knee and ankle centres and the axis of rotation of the knee. The mechanical femorotibial (MFT) angle with the limb in maximum extension was recorded. Measurement of maximum extension and flexion angle was recorded. The investigator was not blinded to measurement of maximum flexion or extension angle and force application was not standardised. Once these angles were recorded, the display of the MFT angle was covered to blind the investigator to this variable; it was necessary that the flexion angle was still visible on the monitor throughout in order to achieve the required flexion intervals. Neutral or unstressed MFT angle i.e. the natural coronal alignment of the lower limb was recorded at 10° intervals from full extension to 90° flexion. A second registration was then performed and the MFT angle in extension was recorded. Providing the MFT angle from the second registration was within 2° of the first, the coronal alignment display on the computer monitor was covered once again and the MFT angle recorded using no stress was applied to the leg at 10° intervals. At the end of recording, side supports were fitted to the table at the level of the proximal thigh to provide medial & lateral support (Fig. 2). The MFT angle was measured at each 10° interval from maximum extension to maximum flexion with first 15 Nm varus and then 15 Nm valgus stress applied. These measurements were repeated. This meant that two measurements were taken at each condition of stress and flexion/extension angles. Statistical analysis was carried out to calculate the reliability, precision and accuracy of the measurement of coronal alignment with and without stress and for maximum flexion and extension. Reliability within each method of tracker fixation used in was analysed by calculating intraclass correlation coefficient (ICC) [24]. A coefficient of ≥0.75 demonstrates very good reliability [25,26]. Calculation of ICC was
Invasive method ICC
Non-invasive method ICC
0.976 (0.957–0.988) 0.995 (0.988–1.000) 0.998 (0.992–1.000)
0.917 (0.785–0.988) 0.980 (0.950–0.995) 0.989 (0.959–0.997)
performed using an IBM SPSS® Statistics 17.0 software (IBM Corp., Armonk, NY, USA). The coefficient of repeatability (CR) was calculated to demonstrate precision of measurements within each method of tracker fixation [27]. The CR defines interval within which 95% of differences between two measurements lie, a CR of 3° conveys that 95% of all measurements are within a range of ± 1.5°. This would be substantially better than current clinical methods of measuring the MFT angle and would be acceptable for pre-operative assessment of coronal alignment [16,28]. For the measurement of flexion/extension the acceptable CR was set to be 5°, and ± 2.5° based on the reported accuracy of measurement with goniometers [29]. To assess the accuracy of the non-invasive system 95% limits of agreement (LOA) between the invasive and non-invasive measurements were determined using the corrected standard deviation of the differences [27]. Acceptable LOA were defined as 3° (± 1.5°) for the MFT angle and 5° (±2.5°) for flexion/extension. CR & LOA calculations were performed using Microsoft Excel® (Microsoft Corp, Redmond, Washington, USA). 4. Results Mean age of specimens was 80.5 years (range 65–91 years), five were female. Mean fixed flexion contracture (maximum extension) for the 12 limbs was 5.8° (range 6° hyperextension to 15° fixed flexion). In full extension, mean mechanical femorotibial (MFT) angle was 2.4° valgus (range 3° varus to 6° valgus). The mean varus and valgus laxities with the knee in full extension were 1.8° (range 0° to 4°) and 1.5° (range 0° to 4°) respectively. Mean maximum flexion for the 12 specimens was 132° (range 111° to 155°). For MFT angle with no applied stress and applying 15 Nm of varus or valgus stress the mean and range of intraclass correlation coefficients (ICCs) were acceptable (N0.75) throughout the range of flexion tested indicating good reliability for both the invasive and non-invasive methods (Table 1). The precision given by the coefficient of repeatability (CR) using the invasive system to measure the MFT angle with no stress was acceptable (CR b 3°) throughout flexion (Fig. 4, Table 2). The precision using the non-invasive system was acceptable from extension to 50° (Fig. 4, Table 2). The CRs for both the invasive and non-invasive methods measuring the MFT angle whilst 15 Nm of varus or valgus stress was applied were acceptable (b3°) throughout the range of flexion tested (Table 2). Accuracy of the non-invasive system as given by the limits of agreement (LOA) between the invasive and non-invasive methods was acceptable (LOA b3°) for measuring
Fig. 4. Coefficient of repeatability (CR) for invasive and non-invasive measurement of MFT angle throughout the range of knee flexion with no stress applied to the limb (CR b3° is acceptable).
1066
D.F. Russell et al. / The Knee 21 (2014) 1063–1068
Table 2 Precision of measuring the MFT angle throughout the range of flexion using the invasive and non-invasive methods whilst applying varus and valgus stresses as given by the coefficient of repeatability (CR). Data are presented as mean (range).
No stress 15 Nm valgus stress 15 Nm varus stress
Invasive method CR ( °)
Non-invasive method CR ( °)
1.3 (0.6–1.9) 0.6 (0.0–0.9) 0.3 (0.6–1.8)
3.0 (0.9–6.9) 1.2 (0.0–0.8) 1.1 (0.6–2.6)
the MFT angle from extension to 40° knee flexion when no stress was applied, and from extension to 30° knee flexion when 15 Nm varus or valgus stress was applied (Fig. 5). Regarding maximum flexion and extension, the invasive and non-invasive methods of tracker fixation displayed acceptable reliability of measurement with ICCs N 0.75 (Table 3). Both methods of tracker fixation gave acceptable precision measuring full extension and flexion (CR b 5°) (Table 4). Accuracy of the non-invasive system was on the limit (LOA = 5.1°) for measuring full extension but was acceptable for measuring full flexion (Table 4).
5. Discussion We aimed to determine the reliability, precision and accuracy of a non-invasive adaptation of image-free navigation technology in measuring the coronal and sagittal mechanical alignment of the lower limb in flexion by direct comparison with a commercially available invasive system. The non-invasive system demonstrated acceptable reliability, precision and accuracy when compared with the invasive system when measuring the coronal MFT angle from extension to 40° flexion with no coronal stress applied and extension to 30° flexion when 15 Nm varus or valgus stress was applied and when measuring maximum extension and flexion. A major limitation of the methodology is the use of cadaveric specimens; tissue quality, muscle tone and soft tissue artefacts will differ from the in-vivo setting. Although the investigator remained blinded to measurement of the coronal MFT angle, it was not possible to remain blinded to measurement of maximum extension and flexion. A further limitation to accuracy testing is that measurements using the invasive and non-invasive system were taken during two separate episodes of testing. This introduces error in accuracy testing but may also make accuracy testing between methods more robust. Standardised force application was not employed to position the limb in flexion, which may account for the small inconsistency in flexion angle reached. Although the precision of the non-invasive system was acceptable when measurements were taken with applied stress, for the unstressed measurements of the MFT angle precision became unacceptable beyond 50° flexion. A potential reason for this may be the inherent laxity of the knee joint in mid-flexion leading to slight variation in coronal alignment between episodes of positioning when no stress is being applied to the
Table 3 Reliability of measuring maximum extension and maximum flexion using the invasive and non-invasive methods as given by the intraclass correlation coefficient (ICC). Data are presented as mean (range).
Maximum extension Maximum flexion
Invasive method ICC
Non-invasive method ICC
0.93 (0.78–0.98) 1.00 (1.00–1.00)
0.94 (0.8–0.98) 1.00 (0.99–1.00)
limb, augmenting the effect of soft tissue artefact variability. However when stress is applied to the limb, the knee reaches the outer envelope of its laxity, this may be a more consistent position thus giving a more consistent alignment between measurement episodes. Soft-tissue artefacts affecting the non-invasive system are most likely the reason for decreasing accuracy of the non-invasive methods when measuring the MFT angle with increased knee flexion and coronal stress (Fig. 5). Application of coronal stress is likely to cause rotatory displacement of the bony anatomy which will be detected directly using invasive tracker mounting, but possibly not so well by skin mounted trackers. Clarke et al. [16] demonstrated high levels of reliability and precision using this device between investigators measuring the MFT angle in extension and when measuring coronal mechanical alignment in extension with no applied stress, varus and valgus stress respectively. The limits of agreement (b 3°) were similar to those found in this study. Measuring full extension, precision is inferior to that reported by Clarke [15]; their cohort exhibited hyperextension unlike the fixed flexion deformity observed in the cadaveric specimens used in this study. Knees in hyperextension may allow more consistent limb positioning as the flexion contracture and lack of muscle tone in the cadaveric specimens are likely to nullify stability inherent to the normal in-vivo extended knee joint as a result of a functioning ‘screw-home’ mechanism supported by active musculature [30]. Nonetheless, the noninvasive method of tracker fixation is superior to visual assessment which has been shown to be very unreliable [5] and to goniometry which has a margin of error of at least ±5° [29]. In knee arthroplasty, it is difficult to communicate findings from examination of the knee using examination techniques which are not standardised and cannot be quantified, far less make reliable recommendations as to how to proceed with management of soft-tissues during total knee replacement [31–33]. Short leg radiographs are inadequate and can even cause clinicians to completely misinterpret alignment [2]. Long leg radiographs are superior in determining lower limb mechanical alignment however they are prone to rotational error [3,34], do not allow dynamic assessment, and expose the patient to radiation. Ideally, clinicians would have access to a relatively inexpensive, reliable, precise and accurate method of assessing dynamic, weight-bearing mechanical
Fig. 5. Limits of agreement (LOA) representing agreement between the two methods of tracker fixation measuring MFT angle throughout flexion in conditions of: no coronal stress, 15Nm valgus stress and 15Nm varus stress (LOA b3° is acceptable).
D.F. Russell et al. / The Knee 21 (2014) 1063–1068 Table 4 Precision and accuracy of measuring maximum extension and maximum flexion using the invasive and non-invasive methods as given by the coefficient of repeatability (CR) and the limits of agreement (LOA).
Maximum extension Maximum flexion
Invasive method CR (°)
Non-invasive method CR (°)
Limits of agreement (°)
3.3 1.5
2.6 2.4
5.1 3.4
alignment and knee joint laxity in extension and early flexion before and after total knee arthroplasty. Prior to surgery, this would allow a detailed assessment of kinematics and planning of tissue release & resection. Following surgery, such a device would be ideal for audit and research as described above, as well as assessing problematic knees such as those describing flexion instability which accounts for a significant proportion of revision surgery [35,36]. In evaluation of soft-tissue/sports injuries, collateral ligament laxity testing is currently limited in terms of using devices to quantify coronal laxity in the examination of joint space opening [37,38]. This device could allow the quantification or grading of collateral ligament injury in a consistent manner and is within the range relevant to laxity testing such as that used in soft-tissue management algorithms for total knee replacement [39–41]. This study would support that it is possible to use optical tracking of skin-mounted markers as a basis for a reliable knee assessment system in early knee flexion. It is accepted that the current prototype could be further developed to maximise efficiency in clinical use, however it has been used on patients and does allow more accurate dynamic assessment than that available to clinicians at the moment. With the relentless improvement in imaging technology it would not be unreasonable to envisage an optical system using small inexpensive cameras providing a permanent setup in orthopaedic clinic rooms making dynamic alignment assessment via optical tracking routine. 6. Conclusion The non-invasive method demonstrated satisfactory reliability, precision and accuracy for measuring lower limb alignment in the early part of knee flexion. This has applications for sports medicine and knee arthroplasty; and in pre-operative decision-making, planning and post-operative evaluation. Declaration of interest statement No financial support was given by any commercial body. The first author received a bursary for research to aid career development from the West of Scotland Orthopaedic Research Society. The authors received material and software support from BBraun Aesculap, Tuttlingen, Germany. Mr. Frederic Picard has separate licences and patents with BBraun. The Department of Orthopaedics at the Golden Jubilee National Hospital receives funding for research from a number of orthopaedic manufacturers including BBraun Aesculap, however no direct funding was received for this particular study. Acknowledgements We give special thanks to the technical and administrative staff at the Laboratory of Human Anatomy, University of Glasgow. We also thank Mr. Phil Cleary & Mr. Iain Freer for the excellent support provided in supplying the equipment to facilitate this study. References [1] Moreland JR, Bassett LW, Hanker GJ. Radiographic analysis of the axial alignment of the lower extremity. J Bone Joint Surg Am 1987;69:745–9.
1067
[2] van Raaij TM, Brouwer RW, Reijman M, Bierma-Zeinstra SM, Verhaar JA. Conventional knee films hamper accurate knee alignment determination in patients with varus osteoarthritis of the knee. Knee 2009;16:109–11. [3] Krackow KA, Pepe CL, Galloway EJ. A mathematical analysis of the effect of flexion and rotation on apparent varus/valgus alignment at the knee. Orthopedics 1990; 13:861–8. [4] Yaffe MA, Koo SS, Stulberg SD. Radiographic and navigation measurements of TKA limb alignment do not correlate. Clin Orthop Relat Res 2008;466:2736–44. [5] Shetty GM, Mullaji A, Lingaraju AP, Bhayde S. How accurate are orthopaedic surgeons in visually estimating lower limb alignment? Acta Orthop Belg 2011;77:638–43. [6] McDaniel G, Mitchell KL, Charles C, Kraus VB. A comparison of five approaches to measurement of anatomic knee alignment from radiographs. Osteoarthritis Cartilage 2010;18:273. [7] Hinman RS, May RL, Crossley KM. Is there an alternative to the full-leg radiograph for determining knee joint alignment in osteoarthritis? Arthritis Rheum 2006; 55:306–13. [8] Navali AM, Bahari LAS, Nazari B. A comparative assessment of alternatives to the full-leg radiograph for determining knee joint alignment. Sports Med Arthrosc Rehabil Ther Technol 2012;4:40. [9] Picard F. Computer assisted orthopaedics — the image free concept. Berlin: Pro BUSINESS GmbH; 2007. [10] Pearle AD, Solomon DJ, Wanich T, Moreau-Gaudry A, Granchi CC, Wickiewicz TL, et al. Reliability of navigated knee stability examination: a cadaveric evaluation. Am J Sports Med 2007;35:1315–20. [11] Kim YH, Park JW, Kim JS. Computer-navigated versus conventional total knee arthroplasty: a prospective randomized trial. J Bone Joint Surg Am 2012; 21:2017–24. [12] Matziolis G, Krocker D, Weiss U, Tohtz S, Perka C. A prospective, randomized study of computer-assisted and conventional total knee arthroplasty. Threedimensional evaluation of implant alignment and rotation. J Bone Joint Surg Am 2007;89:236–43. [13] Brin YS, Nikolaou VS, Joseph L, Zukor DJ, Antoniou J. Imageless computer assisted versus conventional total knee replacement. A Bayesian meta-analysis of 23 comparative studies. Int Orthop 2011;35:331–9. [14] DiGioia AM, Mor AB. Accuracy and validation for surgical navigation systems. AAOS: American Academy of Orthopaedic Surgeons; 2005. [15] Clarke JV. The non-invasive measurement of knee kinematics in normal, osteoarthritic and prosthetic knees. Strathprints: The University of Strathclyde Institutional Repository. University of Strathclyde; 2012. [16] Clarke JV, Riches PE, Picard F, Deakin AH. Non-invasive computer-assisted measurement of knee alignment. Comput Aided Surg 2012;17:29–39. [17] Clarke JV, Wilson WT, Wearing SC, Picard F, Riches PE, Deakin AH. Standardising the clinical assessment of coronal knee laxity. Proc Inst Mech Eng H 2012; 226:699–708. [18] Russell DF, Deakin AH, Fogg QA, Picard F. Non-invasive quantification of lower limb mechanical alignment in flexion. Comput Aided Surg 2014:1–7[Online open access, URL http://informahealthcare.com/doi/pdf/10.3109/10929088.2014.885566]. [19] Russell D, Deakin AH, Fogg QA, Picard F. Non-invasive quantification of lower limb mechanical alignment in flexion. Comput Aided Surg 2014. http://dx.doi.org/10.3109/ 10929088.2014.885566[submitted for publication]. [20] Saragaglia D, Mercier N, Colle PE. Computer-assisted osteotomies for genu varum deformity: which osteotomy for which varus? Int Orthop 2010;34:185–90. [21] Grood ES, Noyes FR, Butler DL, Suntay WJ. Ligamentous and capsular restraints preventing straight medial and lateral laxity in intact human cadaver knees. J Bone Joint Surg Am 1981;63:1257–69. [22] Stahelin T, Kessler O, Pfirrmann C, Jacob HA, Romero J. Fluoroscopically assisted stress radiography for varus–valgus stability assessment in flexion after total knee arthroplasty. J Arthroplasty 2003;18:513–5. [23] Wilson WT, Deakin AH, Picard F, Riches PE, Clarke JV. Standardising the assessment of coronal knee laxity. J Bone Joint Surg 2013;95(30):9. [24] Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull 1979;86:420–8. [25] Portney LG, Watkins MP. Foundations of clinical research: applications to practice. Norwalk, Conn.: Appleton & Lange; 1993 [26] Fleiss JL. Statistical methods for rates and proportions. New York: Wiley; 1981. [27] Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–10. [28] Picard F, Deakin AH, Clarke IV, Dillon JM, Kinninmonth AW. A quantitative method of effective soft tissue management for varus knees in total knee replacement surgery using navigational techniques. Proc Inst Mech Eng H 2007;221:763–72. [29] Edwards JZ, Greene KA, Davis RS, Kovacik MW, Noe DA, Askew MJ. Measuring flexion in knee arthroplasty patients. J Arthroplasty 2004;19:369–72. [30] Croce RV, Miller JP. Coactivation patterns of the medial and lateral hamstrings based on joint position and movement velocity during isokinetic movements. Electromyogr Clin Neurophysiol 2006;46:113–22. [31] Engh GA. The difficult knee: severe varus and valgus. Clin Orthop Relat Res 2003:58–63. [32] Ries MD, Haas SB, Windsor RE. Soft-tissue balance in revision total knee arthroplasty. J Bone Joint Surg Am 2003;85-A(Suppl. 1):S38–42. [33] Krackow KA. The technique of total knee arthroplasty. St. Louis: C.V. Mosby; 1990. [34] Mahaluxmivala J, Bankes MJ, Nicolai P, Aldam CH, Allen PW. The effect of surgeon experience on component positioning in 673 press fit condylar posterior cruciatesacrificing total knee arthroplasties. J Arthroplasty 2001;16:635–40. [35] Fowler PJ. The classification and early diagnosis of knee joint instability. Clin Orthop Relat Res 1980:15–21. [36] Fehring TK, Odum S, Griffin WL, Mason JB, Nadaud M. Early failures in total knee arthroplasty. Clin Orthop Relat Res 2001:315–8.
1068
D.F. Russell et al. / The Knee 21 (2014) 1063–1068
[37] Gwathmey Jr FW, Tompkins MA, Gaskin CM, Miller MD. Can stress radiography of the knee help characterize posterolateral corner injury? Clin Orthop Relat Res 2012;470:768–73. [38] LaPrade RF, Spiridonov SI, Coobs BR, Ruckert PR, Griffith CJ. Fibular collateral ligament anatomical reconstructions: a prospective outcomes study. Am J Sports Med 2010;38:2005–11. [39] Jenny JY. Coronal plane knee laxity measurement: is computer-assisted navigation useful? Orthop Traumatol Surg Res 2010;96:583–8.
[40] Lehnen K, Giesinger K, Warschkow R, Porter M, Koch E, Kuster MS. Clinical outcome using a ligament referencing technique in CAS versus conventional technique. Knee Surg Sports Traumatol Arthrosc 2011;19:887–92. [41] Picard F, Deakin AH, Clarke JV, Dillon JM, Gregori A. Using navigation intraoperative measurements narrows range of outcomes in TKA. Clin Orthop Relat Res 2007; 463:50–7.
