AJR Women's Imaging Online
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lee, I. S.
Right arrow Articles by Kang, H. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lee, I. S.
Right arrow Articles by Kang, H. S.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?
DOI:10.2214/AJR.05.1191
AJR 2006; 186:1778-1782
© American Roentgen Ray Society

Original Research

Reliability Analysis of 16-MDCT in Preoperative Evaluation of Total Knee Arthroplasty and Comparison With Intraoperative Measurements

In Sook Lee1,2,3, Jung-Ah Choi1,2, Tae Kyun Kim4, Ilkyu Han4, Joon Woo Lee1,2 and Heung Sik Kang1,2

1 Department of Radiology, Seoul National University Bundang Hospital, 300 Gumi-dong, Bundang-gu, Seongnam, Gyeongi-do 463-707, Korea.
2 Department of Radiology and Institute of Radiation Medicine, Seoul National University College of Medicine, Seoul, Korea.
4 Department of Orthopedic Surgery, Seoul National University Bundang Hospital, Seoul, Korea.

Received July 11, 2005; accepted after revision September 4, 2005.

 
Address correspondence to J.-A. Choi (jacrad{at}radiol.snu.ac.kr).

3 Present address: Department of Radiology, Pusan National University College of Medicine, Pusan, Korea.


Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
OBJECTIVE. The aim of our study was to determine how consistent measurements of the distal femoral condyle on 16-MDCT are with intraoperative measurements and the reliability of 16-MDCT for the preoperative planning of total knee arthroplasty.

MATERIALS AND METHODS. Between August 2003 and March 2004, 33 consecutive patients (two men and 31 women; age range, 53-89 years; mean age, 71 years) presenting with osteoarthritis underwent 16-MDCT (Mx 8000 IDT) of a total of 53 knees before total knee arthroplasty. The prospective analysis included measurements of transepicondylar distance, maximum anteroposterior dimension of medial and lateral femoral condyles, and trochlear width on a Rapidia workstation. To increase reliability, we repeated the measurements on a CT workstation after 2 months and compared them with the previous values. The values measured on the CT workstation were compared with the measurements obtained in the intraoperative field. Statistical analysis was performed using kappa statistics. A p value of less than 0.05 was considered statistically significant.

RESULTS. The mean values of transepicondylar distance, maximum anteroposterior dimension of medial and lateral femoral condyles, and trochlear width were 75, 57, 58, and 38 mm at first measurement; 76, 58, 59, and 39 mm at second measurement on the CT workstation; and 79, 57, 60, and 42 mm at intraoperative measurement, respectively. At reliability analysis between the first measurements on the CT workstation and the intraoperative measurements, kappa values were 0.84 for the transepicondylar distance, 0.81 for the maximum anteroposterior dimension of the medial femoral condyle, 0.89 for the maximum anteroposterior dimension of the lateral femoral condyle, and 0.62 for the trochlear width. The kappa values for the second measurements were 0.86, 0.77, 0.84, and 0.61, respectively. Intraoperative measurements and measured values on the CT workstation showed excellent and almost perfect agreement, and intraobserver agreement was almost perfect.

CONCLUSION. Femoral sizing on a CT workstation had excellent or almost perfect correlation with intraoperative measurements. Therefore, reliable preoperative planning for total knee arthroplasty may be done with 16-MDCT and an advanced workstation.

Keywords: anatomy • arthroplasty • CT technique • MDCT • knee • musculoskeletal imaging


Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
For successful total knee arthroplasty, accurate implant alignment and ligament balance are essential. The dimension of the distal femur and proper rotational alignment of the femoral component are especially critical for the outcome of total knee arthroplasty [1, 2]. The importance of correct sizing of components for total knee arthroplasty for optimal function and long-term results has been stressed in many reports [3, 4].

Usually, orthopedic surgeons obtain conventional radiographs of standing knees or legs before total knee arthroplasty for preoperative planning. But in some cases there are limitations in the correct sizing of the femoral component and selecting the proper implants using only anteroposterior and lateral conventional radiographs. Also, quantifying the shape of bones that constitute the osteoarthritic knee is difficult because the knee is often deformed; and depending on the alignment of knee joint and the patient's position, the size is measured differently [5]. To our knowledge, the proper sizing of the femoral component for total knee arthroplasty with 16-MDCT has not been the subject of a previous article. With the introduction of 16-MDCT and advanced workstations, 3D measurement has become feasible, and so the proper size of the femoral component can theoretically be measured, irrespective of alignment deformity or the patient's position.

The purpose of our study was to determine how consistent measurements of the distal femoral condyle on 16-MDCT are with intraoperative measurements and the reliability of 16-MDCT for the preoperative planning of total knee arthroplasty.


Figure 1
View larger version (64K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1A —Methods of femoral sizing on a CT workstation. All measurements were done using multiplanar reconstruction and 3D volume-rendered images. Dot cursors for measuring on 3D volume-rendered image were placed in identical portions on corresponding axial, coronal, and sagittal images. Transepicondylar distance was measured as distance between most medial and most lateral prominences of epicondyles.

 


Figure 2
View larger version (61K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1B —Methods of femoral sizing on a CT workstation. All measurements were done using multiplanar reconstruction and 3D volume-rendered images. Dot cursors for measuring on 3D volume-rendered image were placed in identical portions on corresponding axial, coronal, and sagittal images. Maximum anteroposterior dimension of medial femoral condyle was measured at most anterior and most posterior projections.

 


Figure 3
View larger version (62K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1C —Methods of femoral sizing on a CT workstation. All measurements were done using multiplanar reconstruction and 3D volume-rendered images. Dot cursors for measuring on 3D volume-rendered image were placed in identical portions on corresponding axial, coronal, and sagittal images. Maximum anteroposterior dimension of lateral femoral condyle was measured at most anterior and most posterior projections.

 


Figure 4
View larger version (58K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1D —Methods of femoral sizing on a CT workstation. All measurements were done using multiplanar reconstruction and 3D volume-rendered images. Dot cursors for measuring on 3D volume-rendered image were placed in identical portions on corresponding axial, coronal, and sagittal images. Trochlear width was determined as distance between most anterior projections of medial and lateral condyles.

 

Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Between August 2003 and March 2004, 33 consecutive patients (two men and 31 women; age range, 53-89 years; mean age, 71 years) presenting with osteoarthritis underwent 16-MDCT (Mx 8000 IDT, Philips Medical Systems) of a total 53 knees before total knee arthroplasty. The MDCT protocol was as follows: slice thickness, 2 mm; interslice gap, 1 mm; field of view, 305-426 x 305-426; and matrix size, 512 x 512. All images were retrieved to the Rapidia 2.8 workstation (Infinitt), and then only portions of interest of the distal femoral condyle were reformatted with 3D reconstruction. CT was routinely performed before total knee arthroplasty at our institution according to the surgeon's need. Institutional review board approval was not required for this study.

Prospective and independent analysis was done by two musculoskeletal radiologists who had discussed the parameters of interest and methods of measuring the parameters among themselves, and with the relevant orthopedic surgeon. The analysis included measurements on a workstation of the transepicondylar distance, the maximum anteroposterior dimension of the medial and lateral femoral condyles, and the trochlear width, which were deemed by the orthopedic surgeon to be the most important measurements in determining proper implant size. Transepicondylar distance (Fig. 1A) was defined as the distance between the most medial and most lateral prominences of the epicondyles. Maximum anteroposterior dimensions of medial (Fig. 1B) and lateral (Fig. 1C) femoral condyles were measured at the most anterior and most posterior projections. Trochlear width (Fig. 1D) was determined as the distance between the most anterior projections of the medial and lateral condyles. These dimensions were measured using multiplanar reconstruction and 3D volume-rendered images. For accuracy, dot cursors for measuring on the 3D volume-rendered image were placed in identical portions on corresponding axial, coronal, and sagittal images. Two musculoskeletal radiologists measured independently at first, and then two sets of values were compared and the degree of agreement was assessed. However, the measurements by only one radiologist were compared with the intraoperative measurements and included in the results. To increase reliability, one radiologist repeated the measurements on the CT workstation after 2 months and compared them with the previous values. All measurements were recorded in millimeters and were rounded to one decimal place.

The intraoperative measurements were done at surgery using a caliper (special caliper for neurosurgery, AA845R, 125 mm, 5-inch length, Aesculap AG & Co.), by one orthopedic surgeon who was also blinded to the preoperative measurements done by the two radiologists. The values measured on the CT workstation were compared with the measurements obtained in the intraoperative field. The two sets of values were compared for each dimension measured.

Statistical analysis was done using intraclass correlation coefficients and kappa statistics, which were calculated with commercially available software (SPSS for Windows [Microsoft], version 10.0). According to Landis and Koch [6], on the basis of kappa values agreement was classified as almost perfect ({kappa} = 0.81-1.00), excellent ({kappa} = 0.61-0.80), good or moderate ({kappa} = 0.41-0.60), fair ({kappa} = 0.21-0.40), slight ({kappa} = 0.00-0.20), or poor ({kappa} = 0.00 or negative). A p value of less than 0.05 was considered to be statistically significant.


Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
The kappa values for interobserver agreement between the two musculoskeletal radiologists at first measurement were as follows: 0.96 for the transepicondylar distance, 0.94 for the maximum anteroposterior dimension of the medial femoral condyle, 0.95 for the maximum anteroposterior dimension of the lateral femoral condyle, and 0.86 for the trochlear width. Because the results showed almost perfect interobserver agreement, only the values measured by one radiologist were used for comparison with the intraoperative values. The mean values of transepicondylar distance, maximum anteroposterior dimension of medial and lateral femoral condyles, and trochlear width were measured as 75, 57, 58, and 38 mm at first measurement on the CT workstation; as 76, 58, 59, and 39 mm at second measurement on the CT workstation; and as 79, 57, 60, and 42 mm at intraoperative measurement, respectively (Table 1). The intraoperatively measured values tended to be larger than both measurements on the CT workstation for transepicondylar distance, maximum anteroposterior dimension of the lateral femoral condyle, and trochlear width. The individual values for maximum anteroposterior dimensions of the medial condyle on the CT workstation at first measurement were more variable than those at intraoperative measurement; however, the mean values were the same (Table 1).


View this table:
[in this window]
[in a new window]

  		TABLE 1: Mean Values Measured on CT Workstation and Intraoperatively

 

At reliability analysis between the first measurements on the CT workstation and the intraoperative measurements, single-measure intraclass correlation coefficients were 0.84 for the transepicondylar distance, 0.81 for the maximum anteroposterior dimension of the medial femoral condyle, 0.89 for the maximum anteroposterior dimension of the lateral femoral condyle, and 0.62 for the trochlear width. The values for the second measurement were 0.86, 0.77, 0.84, and 0.61, respectively (Table 2). Transepicondylar distance and maximum anteroposterior dimension of the lateral femoral condyle showed almost perfect agreement with intraoperative measurements at both of the two measurements. Maximum anteroposterior dimension of the medial femoral condyle showed almost perfect consistency at first measurement and excellent consistency at second measurement. Trochlear width showed relatively lower consistency compared with intraoperative measurements than other values at both of the two measurements; however, the overall consistency was still excellent.


View this table:
[in this window]
[in a new window]

  		TABLE 2: Agreement Between Measurements on CT Workstation and Intraoperative Measurements Assessed Using Kappa Statistics

 

Also, the intraclass correlation coefficient method was used to evaluate intraobserver agreement. The kappa values were 0.96 for the transepicondylar distance, 0.92 for the maximum anteroposterior dimension of the medial femoral condyle, 0.95 for the maximum anteroposterior dimension of the lateral femoral condyle, and 0.88 for trochlear width. The degree of intraobserver agreement was almost perfect for all measurements.


Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Proper implant sizing can help to avoid complications and maximize outcome, and appropriate femoral sizing is important to obtain accurate ligament balancing in flexion and extension [7, 8]. Selecting the correct size of femoral component for a particular patient can be problematic in some cases. Problems may arise when the distal femoral anteroposterior dimension does not precisely fit any of the sizes for standard distal femoral components, and in cases in which the anteroposterior dimensions fit precisely but the mediolateral dimension of the distal femur may be relatively broad or narrow [2]. Medial or lateral overhang on the femur or tibia can result in soft-tissue irritation and affect balancing efforts. Undersizing can lead to difficulties in soft-tissue balancing and an increased flexion gap or anterior notching, or leave cancellous bone exposed, which could be a source of increased bleeding or periprosthetic fracture [7, 9].

If surgeons are aware of the correct size of the femoral component preoperatively, they may be able to use anatomically better-sized implants for each patient. In other words, if sizing of the femoral component on imaging is nearly consistent with real femoral sizing before total knee arthroplasty, general preoperative planning would be possible, and these results could allow the surgeon to determine the properly sized prosthesis and the prosthesis manufacturers to provide appropriately sized implants preoperatively. Also, preoperatively understanding the morphology of the distal femoral condyle is important. However, exact quantification of the dimensions of the osteoarthritic knee is often difficult because the knee is often severely deformed [10].

Interactive 3D reconstruction models are being increasingly used to simulate surgical procedures and to design custom prostheses for reconstructive surgery [11], in which important surgical decisions are based on spatial measurements of the 3D reconstruction, which may dramatically affect the treatment approach or functional outcome of the surgical procedure. Despite a number of reports advocating 3D reconstruction for surgical planning, no previous study has validated the accuracy of spatial measurements based on 3D reconstruction [12]. In a previous study [13], the accuracy of dimensional measurements of knee anatomy-based, planar oblique slice reconstructions averaged 92-93% compared with direct measurements of the anatomic specimen. That preliminary investigation provides a validation of 3D reconstruction in the medical literature.

To our knowledge, no studies have reported the 3D measurements of the femoral condyle for total knee arthroplasty using 16-MDCT. In this study, we attempted to determine the actual size of the femoral component for total knee arthroplasty with 16-MDCT. Previous studies showed a difference in sizing of the distal femur for total knee arthroplasty according to the patient's position and rotational alignment on conventional radiography, MRI, and conventional CT. Using data from 16-MDCT and an advanced workstation, we could measure the proper size of the femoral component on 3D reconstruction, multiplanar reconstruction coronal and sagittal images, and axial images, irrespective of rotational alignment. The sizing of the femoral component for total knee arthroplasty on the CT workstation showed excellent and almost perfect agreement compared with intraoperative measurements.

Seedhom et al. [14], in a radiographic cadaveric study that evaluated sizing for knee prostheses, supported the idea that the required femoral component size should be based on the mediolateral dimension of the femoral condyle. Although this would avoid mediolateral undersizing or overhang, the effect of flexion-extension balancing techniques could distort the original measurements. Considering that study, proper measurement of the transepicondylar distance and determination of the transepicondylar axis would be fundamental in total knee arthroplasty. Although our results show almost perfect correlation between intraoperative measurement and preoperative measurement of transepicondylar distance, intraoperative values tended to be larger; in some cases, the difference was as large as 10 mm. The reason for this discrepancy may be the inclusion of soft tissue in intraoperative measurements, whereas on CT images, measurements were done from exact bone landmarks—that is, the epicondyles.

Femoral sizing becomes particularly important in patients with a very small or very large femoral component, such as in severe osteoarthritis with deformity resulting from juvenile rheumatoid arthritis or in trauma, when preparation of a properly sized prosthesis is more difficult. In such cases, accurate preoperative measurement and planning using MDCT are essential.

In addition, isotropically reformatted 3D images can help orthopedic surgeons predict the 3D morphology of the knees and further help in preoperative planning. Morphologic and geometric data, which are essential for successful total knee arthroplasty, can be obtained simultaneously by 16-MDCT. Subsequently, these data can be applied to computer-assisted surgery, which is expanding its field; in particular, in robotic surgery actual bone resection for total knee arthroplasty is performed on the basis of information determined during preoperative planning [15]. Although attempts at computer-assisted surgery using navigation and bone morphing, instead of imaging, are made, navigation systems have the tendency to oversize the femoral component [16, 17]. Therefore, CT should still be considered the gold standard because of its exact measurement and reproducibility of data, as shown in this study.

Our study has some limitations. First, for the sizing of the distal femoral component, we compared only intraoperative measurements and measurements on a CT workstation with 16-MDCT; we did not compare measurements on conventional radiographs. However, the purpose of this study was to determine the reliability of measurements on 16-MDCT as compared with intraoperative measurements, and not as compared with those on conventional radiographs. Second, some differences were seen between intraoperative measurements and measurements on the CT workstation. In patients with severely deformed knees and large and multiple osteophytes, the degree of difference tended to be larger; therefore, more careful measurements and averaging of repeated measured values would be needed in such cases. However, some discrepancy may simply be due to the inclusion of soft tissue in the intraoperative measurements, rather than from the inaccuracy inherent to measurement on CT. Third, because intraoperative measurements were performed by only one surgeon and only once, intra- and interobserver agreements for intraoperative measurements were not assessed.

In conclusion, our study showed that femoral sizing on a CT workstation had almost perfect or excellent correlation with intraoperative measurements; therefore, reliable preoperative planning for total knee arthroplasty can be done with 16-MDCT and an advanced workstation, and the simulation of femoral resection may be feasible. Reliable preoperative 3D measurements with 3D reconstructions using 16-MDCT can be applicable in several instances, especially in the field of computer-assisted surgery, for which its use needs further investigation.


Acknowledgments
 
We offer our special thanks to Yeon Gwi Kang for her help in organizing data for the manuscript.


References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Jeffery RS, Morris RW, Denham RA. Coronal alignment after total knee replacement. J Bone Joint Surg Br1991; 73:709 -714[Medline]
  2. Chin KR, Dalury DF, Zurakowski D, Scott RD. Intraoperative measurements of male and female distal femurs during primary total knee arthroplasty. J Knee Surg 2002;15 : 213-217[Medline]
  3. Insall JN, Easley ME. Surgical techniques and instrumentation in total knee arthroplasty. In: Insall JN, Scott WN, eds. Surgery of the knee. Philadelphia, PA: Churchill Livingstone,2001 : 1553-1620
  4. Police P, Lotke PA, Lonner JH. Principles of instrumentation and component alignment. In: Callaghan JJ, Rosenberg AA, Rubash HE, Simonian PT, Wickiewicz TL, eds. The adult knee. Philadelphia, PA: Lippincott Williams & Wilkins, 2003:1085 -1093
  5. Kellgren JH, Lawrence JS. Radiological assessment of osteo-arthritis. Ann Rheum Dis 1957;16 : 494-502[Free Full Text]
  6. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1997;33 : 159-174
  7. Hitt K, Shurman JR 2nd, Greene K, et al. Anthropometric measurements of the human knee: correlation to the sizing of current knee arthroplasty system. J Bone Joint Surg Am2003; 85-A[suppl 4]:115 -122[Abstract/Free Full Text]
  8. Urabe K, Miura H, Kuwano T, et al. Comparison between the shape of resected femoral sections and femoral prostheses used in total knee arthroplasty in Japanese patients: simulation using three-dimensional computed tomography. J Knee Surg 2003;16 : 27-33[Medline]
  9. McAlinden MG, Masri BA, Garbuz DS, Greidanus NV, Duncan CP. Periprosthetic fractures after total knee arthroplasty. In: Callaghan JJ, Rosenberg AA, Rubash HE, Simonian PT, Wickiewicz TL, eds. The adult knee. Philadelphia, PA: Lippincott Williams & Wilkins,2003 : 1359-1375
  10. Matsuda S, Miura H, Nagamine R, et al. Anatomical analysis of the femoral condyle in normal and osteoarthritic knees. J Orthop Res 2004; 22:104 -109[CrossRef][Medline]
  11. Vannier MW, Marsh JL, Warren JO. Three-dimensional CT reconstruction images for craniofacial surgical planning and evaluation. Radiology 1984;150 : 179-184[Abstract/Free Full Text]
  12. Totty WG, Vannier MW. Complex musculoskeletal anatomy: analysis using three-dimensional surface reconstruction. Radiology 1984;150 : 173-177[Abstract/Free Full Text]
  13. Smith DK, Berquist TH, An KN, et al. Validation of three-dimensional reconstructions of knee anatomy: CT vs MR imaging. J Comput Assist Tomogr 1989;13 : 294-301[Medline]
  14. Seedhom BB, Longton EB, Wright V, Dowson D. Dimensions of the knee: radiographic and autopsy study of sizes required by a knee prosthesis. Ann Rheum Dis 1972;31 : 54-58[Free Full Text]
  15. Mai S, Lorke C, Siebert W. Clinical results with the robot-assisted CASPAR system and the search evolution prosthesis. In: Stiehl JB, Konerman WH, Haaker RG, eds. Navigation and robotics in total joint and spine surgery. Berlin, Germany: Springer, 2004:354 -361
  16. Matsumoto T, Tsumura N, Kurosaka M, et al. Prosthetic alignment and sizing in computer-assisted total knee arthroplasty. Int Orthop 2004; 28:282 -285[CrossRef][Medline]
  17. Stindel E, Briard JL, Merloz P, et al. Bone morphing: 3D morphological data for total knee arthroplasty. Comput Aided Surg 2002; 7:156 -168[CrossRef][Medline]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?



This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lee, I. S.
Right arrow Articles by Kang, H. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lee, I. S.
Right arrow Articles by Kang, H. S.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS