AUTOMATIC 3-DIMENSIONAL CORRELATION OF CT-CT, CT-MRI, AND CT-SPECT USING CHAMFER MATCHING

Image correlation is often required to utilize the complementary infor mation in CT, MRI, and SPECT. A practical method for automatic image c orrelation in three-dimensions (3D) based on chamfer matching is descr ibed. The method starts with automatic extraction of contour points in one modality and automatic segmentation of the corresponding feature in the other modality. A distance transform is applied to the segmente d volume and a cost function is defined that operates between the cont our points and the distance transform. Matching is performed by iterat ively optimizing the cost function for 3D translation, rotation, and s caling of the contour points. The complete matching process including segmentation requires no user interaction and takes about 100 s on an HP715/50 workstation. Perturbation tests on clinical data with cost fu nctions based on mean, rms, and maximum distances in combination with two general purpose optimization procedures have been performed. The p erformance of the methods has been quantified in terms of accuracy, ca pture range, and reliability. The best results on clinical data are ob tained with the cost function based on the mean distance and the simpl ex optimization method. The accuracy is 0.3 mm for CT-Cr, 1.0 mm for C T-MRI, and 0.7 mm for CT-SPECT correlation of the head. The accuracy i s usually at subpixel level but is limited by global geometric distort ions, e.g., for CT-MRI correlation. Both for CT-CT and CT-MRI correlat ion the capture range is about 6 cm, which is higher than normal diffe rences in patient setup found on the scanners (less than 4 cm). This m eans that the correlation procedure seldom fails (better than 98% reli ability) and user interaction is unnecessary. For CT-SPECT matching th e capture range is about 3 cm (80% reliability), and must be further i mproved. The method has already been introduced in clinical practice.