The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B5. Beijing 2008 
804 
Figure 6. Error in dependence of the initial misalignment (left) 
and computation time (right); Approach A is the 
reference algorithm, approach B is the approach 
contributed in this paper 
5. CONCLUSIONS 
We introduced an optimized approach for the pose correction of 
patients in 6 degrees of freedom. First we relied on the whole 
image domains to compute a 5 DOF estimation of the alignment 
based on two 2D registrations and a low number of DRR 
renderings. Then we improved accuracy and found the 6th 
degree of freedom by optimizing the correlation of certain 
image regions. By reduction of the data used in the X-ray 
reconstruction process and by exclusion of image regions that 
do not improve the matching result, we were able to reduce 
pose estimation time and to preserve the accuracy of the 
matching process. This can be beneficial especially for real time 
matching applications, e.g. for alignment surveillance with 
fluoroscopic imaging devices during the treatment. To achieve 
even higher performance we suggest to combine our approach 
with other techniques, for example the progressive lightfield 
rendering introduced in (Rohlfmga et al., 2004). 
For images that contain large areas of image contents that do 
not occur in either the CT series or the X-ray image(s) our 
approach successfully increased reliability of the pose detection 
outcome. However, the method contributed here still has to 
undergo additional testing, especially with respect to 
radiometric differences of the image modalities and for the 
variety of possible imaged body parts (e.g. for thorax or 
abdomen datasets). 
Further investigations should be made, if our approach 
combined with a multi-resolutional matching could lead to 
further improvements. 
After all, we showed that the solution to achieve lower pose 
detection times with similar accuracies can be to reduce the 
image registration to the relevant sub-sets instead of improving 
the rendering or matching technique itself. As we still use the 
mutual information measure for registration of these sub-sets, 
we still benefit from the stability of this measure. 
REFERENCES 
Birkfellner, W; Wirthl, J.; Burgstaller, W.; Baumann, B.; 
Staedele, H.; Hammer, B.; Gellrich, N. C.; Jacob, A. L.; 
Regazzonil, P.; Messmer, P., 2003. A faster method for 3D/2D 
medical image registration - a simulation study. Phys. Med. 
Biol., 48, pp. 2665-2679. 
Information based Set-up Error Estimator for Radiotherapy, 
Medical Physics, 28(12), pp. 2507-2517. 
Olson, C. F., 2001. Image registration by aligning entropies. 
Proceedings of IEEE Computer Society Conference on 
Computer Vision and Pattern Recognition, 2, pp. 331-336. 
Pluim, J.; Maintz, J.; Viergever, M., 2003. Mutual information 
based registration of medical images: A survey. IEEE 
transactions on medical imaging, 22(8), pp. 986-1004. 
Press, W. H.; Teukolsky, S. A.; Vetterling, W. T.; Flannery, B. 
P., 1992. Numerical Recipes in C 2, Cambridge University 
Press, pp. 408-411. 
Rohlfmga, T; Russakoff, D. B.; Denzler, J.; Mori, K.; Maurer, 
C. R., 2004. Progressive attenuation fields: Fast 2D-3D image 
registration without precomputation. In Proc. of Medical Image 
Computing and Computer-Assisted Intervention, 3216, pp. 631 - 
638. 
Selby, B. P.; Sakas, G.; Walter, S.; Groch, W.-D.; Stilla, U., 
2008. Patient Alignment Estimation in Six Degrees of Freedom 
Using a CT-scan and a Single X-ray Image. Proceedings of 
Bildverarbeitung fur die Medizin 2008, pp. 127-132. 
Selby, B. P.; Sakas, G.; Walter, S.; Groch, W.-D.; Stilla U., 
2007. Detection of pose changes for spatial objects from 
projective images. In: Stilla, U.; Meyer, H.; Rottensteiner, F.; 
Heipke, C.; Hinz, S. (eds), PIA07 Photogrammetric Image 
Analysis 2007. International Archives of Photogrammetry and 
Remote Sensing, 36(3), pp. 105-110. 
Thilmann, C.; Nill, S.; Tücking, T.; Höss, A.; Hesse, B.; 
Dietrich, L.; Bendl, R.; Rhein, B.; Häring, P.; Thieke, C.; 
Oelfke, U.; Debus, J.; Huber, P., 2005. Correction of patient 
positioning errors based on in-line cone beam CTs: clinical 
implementation and first experiences. International Journal of 
Radiation Oncology Biology Physics, 63(1), pp. 550-551. 
Verhey, L. J.; Goitein, M.; McNulty, P.; Munzenrider, J. E.; 
Suit, H. D., 1982. Precise positioning of patients for radiation 
therapy. International Journal of Radiation Oncology Biology 
Physics, 8(2), 289-294. 
Yue, N. J.; Knisely, J. P. S.; Song, H.; Nath, R., 2006. A 
method to implement full six-degree target shift corrections for 
rigid body in image-guided radiotherapy. Medical Physics, 
33(1), pp. 21-31. 
Jeongtae, K.; Fessier, J. A.; Lamy, K. L.; Baltery, J. M.; 
Hakeny, R. K. T., 2001. A Feasibility Study on Mutual