British Journal of Radiology (2005) 78, 634-636
© 2005 British Institute of Radiology
doi: 10.1259/bjr/28412864
The use of CT-MR image registration to define target volumes in pelvic radiotherapy in the presence of bilateral hip replacements
N Charnley, FRCR
1
A Morgan, BSc MSc
2
E Thomas, FRCR
1
S Wilson
2
S Bacon, MPhys, MSc
2
D Wilson, BSc, MSc
2 and
D Bottomley, FRCR
1
1 Radiotherapy Department and 2 Medical Physics Department, Cookridge Hospital, Cookridge, Leeds LS16 6QB, UK
Correspondence: A Morgan
 |
Abstract
|
|---|
Increasing numbers of patients with hip replacements are presenting for pelvic radiotherapy, which is usually planned using CT images. Image artefacts caused by the presence of metallic implants tend not to be severe for single hip replacements and allow for adequate definition of target volumes. When bilateral hip replacements are present, the image artefacts can render CT images useless for target definition, particularly for tumours of the prostate and bladder. MR images are not susceptible to such severe artefacts. This note describes a small series of patients with bilateral hip replacements on whom CT-MR image registration has been used to successfully define adequate target volumes.
|
Introduction
|
|---|
There are an increasing number of elderly patients with bilateral hip replacements who require radiotherapy planning of the pelvis. This presents technical difficulties, as efforts must be made to avoid irradiating these implants in order to prevent shielding of the proposed planning target volume (PTV) [1]. The definition of the clinical target volume (CTV) on CT scans, particularly of the bladder and prostate, may itself be problematic as the interference produced by two artificial hips may severely compromise the image quality as demonstrated in Figure 1
.
View larger version (114K):
[in this window]
[in a new window]
|
Figure 1. Pelvic CT image of a patient with bilateral hip prostheses, demonstrating severe image artefacts.
|
|
The use of MRI in such cases may provide a satisfactory solution. We report on the use of CT-MR image registration to facilitate target definition in four patients with prostate or bladder tumours who have had bilateral hip replacements.
|
Method
|
|---|
Pelvic CT studies were obtained for each patient for radiotherapy treatment planning purposes according to local protocols. Images were acquired on a GE Lightspeed CT scanner (General Electric Medical Systems, Milwaukee, WI) fitted with a flat-topped carbon fibre couch. In all cases, the image matrix size was 512 x 512. The CT slice thickness and spacing was 5 mm. Following observation of the artefacts demonstrated in Figure 1
, patients were referred for an MR pelvic study. T2 weighted images were acquired on a 1.5 T Siemens Symphony MR scanner (Siemens, Erlangen, Germany). Again the matrix size used was 512 x 512 and the slice thickness and separation were 5 mm.
All CT and MR images were transferred to a TMS radiotherapy treatment planning system (Nucletron, Veenendal, The Netherlands) for image registration and subsequent planning. The image registration software on TMS requires operator identification of at least three anatomical landmarks in each image set, though more than 10 such points were used in each case. All registration work was performed by a dosimetrist and a clinical oncologist. The software used did not provide a figure of merit on the quality of the registration, nor allow for the visualization of a fused image but once the minimum number of points has been entered and the registration is active, the operator may enter test points to assess the accuracy achieved. These test points may then be converted to registration points if acceptable.
|
Results
|
|---|
In all cases, it was observed that an accuracy of better than 5 mm was obtained over the clinically relevant volume. After registration, a CTV was marked on the MR study and transferred to the CT dataset for expansion to a PTV using our standard margins and subsequent conformal 3D treatment planning as shown in Figure 2. Treatment plans were produced without density correction.
|
Discussion
|
|---|
There are several potential methods of dealing with the problem of poorly visualized target volumes for radiotherapy treatment planning. These include approximation of the target volume from its predicted position using anatomical landmarks. Another method would be to plan the pelvis based on traditional methods such as orthogonal radiographic films. However, neither of these methods would allow for conformal treatment of an accurate CTV.
CT-MR registration has been investigated to delineate volumes for conformal radiotherapy of prostate cancer. Rasch et al [2] and Sannazzari et al [3] showed that CT over-estimates the CTV compared with MR by approximately 30% and that MR may spare organs at risk.
The use of CT-MR registration in patients with localized prostate cancer in which target definition is difficult has been described by Lau et al [4]. They applied CT-MR registration to two patients who had significantly altered pelvic anatomy due to surgery and in whom CT based target definition proved difficult. This allowed for accurate definition of the CTV and better visualization of normal structures.
In this small study, it was anticipated that there may have been some small differences in patient position due to the fact that the CT studies were performed on a flat topped couch and the MR studies were carried out on a standard MR couch with associated firm mattress. Any differences in patient position were found not to adversely affect the quality of the registration. It could be argued, in view of the fact that the CT plans produced were not density corrected, that a treatment plan could be based on the MR study alone. However, when examining Figure 2 as an example, it is clear that there are noticeable differences in the external patient contour between CT and MR due to the couch top shape and it seems appropriate to base the treatment plan on the CT study, where the couch top will be the same as the treatment unit.
We have also considered the effect of artificial hips on MR image quality. The presence of metallic implants in a patient may create significant image distortion. Specific details of the implants used in these patients were not available to us. However, a qualitative study using a selection of current prostheses imaged within a simple grid phantom filled with water tubes demonstrated distortions not greater than 0.2 cm at distances more than 3 cm from the implant.
|
Conclusion
|
|---|
We have demonstrated a method which allows conformal planning of a PTV using CT-MR image registration. This enabled clear delineation of an obscured CTV. A registration accuracy of better than 5 mm has been achieved. It is not proposed that this should become a standard procedure for PTV definition, but may be used pragmatically where the CTV is difficult to determine, as in the case of a patient with a bilateral hip replacement.
|
Acknowledgments
|
|---|
The authors gratefully acknowledge Dr Jonathan Thompson of DePuy International Ltd for providing a selection of hip prostheses for work associated with this report.
|
Footnotes
|
|---|
Current address for N Charnley, Manchester Molecular Imaging Centre, 27 Palatine Road, Withington, Manchester M20 3LJ, UK. 
Received for publication November 5, 2004.
Revision received December 22, 2004.
Accepted for publication January 25, 2005.
|
References
|
|---|
- Reft C, Alecu R, Das IJ, Gerbi B, Keall P, Lief E, et al. Dosimetric considerations for patients with hip prostheses undergoing pelvic irradiation. Report of the AAPM Committee Task Group 63. Med Phys 2003;30:1162–82.[CrossRef][Medline]
- Rasch C, Barillot I, Remeijer P, Touw A, van Herk M, Lebesque JV. Definition of the prostate in CT and MRI: a multi-observer study. Int J Radiat Oncol Biol Phys 1999;43:57–66.[CrossRef][Medline]
- Sannazzari GL, Ragona R, Ruo Redda MG, Giglioli FR, Isolato G, Guarneri A. CT–MRI image fusion for delineation of volumes in three-dimensional conformal radiation therapy in the treatment of localized prostate cancer. Br J Radiol 2002;75:603–7.[Abstract/Free Full Text]
- Lau HY, Kagawa K, Lee WR, Hunt MA, Shaer AH, Hanks GE. Short communication: CT-MRI image fusion for 3D conformal prostate radiotherapy: use in patients with altered pelvic anatomy. Br J Radiol 1996;69:1165–70.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
M Aubin, O Morin, J Chen, A Gillis, B Pickett, J F Aubry, C Akazawa, J Speight, M Roach III, and J Pouliot
The use of megavoltage cone-beam CT to complement CT for target definition in pelvic radiotherapy in the presence of hip replacement
Br. J. Radiol.,
November 1, 2006;
79(947):
918 - 921.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
