Purpose: After a seed implant of the prostate, computerized tomography (CT)
is ideal for determining seed distribution but soft tissue anatomy is freq
uently not well visualized. Magnetic resonance (MR) images soft tissue anat
omy well but seed visualization is problematic, We describe a method of fus
ing CT and MR images to exploit the advantages of both of these modalities
when assessing the quality of a prostate seed implant.
Methods and Materials: Eleven consecutive prostate seed implant patients we
re imaged with axial MR and CT scans, MR and CT images were fused in three
dimensions using the Pinnacle 3.0 version of the ADAC treatment planning sy
stem. The urethra and bladder base were used to "line up" MR and CT image s
ets during image fusion, Alignment was accomplished using translation and r
otation in the three ortho-normal planes. Accuracy of image fusion was eval
uated by calculating the maximum deviation in millimeters between the cente
r of the urethra on axial MR versus CT images. Implant quality was determin
ed by comparing dosimetric results to previously set parameters.
Results: Image fusion was performed with a high degree of accuracy, When li
ning up the urethra and base of bladder, the maximum difference in axial po
sition of the urethra between MR and CT averaged 2.5 mm (range 1.3-4.0 mm,
SD 0.9 mm), By projecting CT-derived dose distributions over MR images of s
oft tissue structures, qualitative and quantitative evaluation of implant q
uality is straightforward.
Conclusions: The image-fusion process we describe provides a sophisticated
way of assessing the quality of a prostate seed implant. Commercial softwar
e makes the process time-efficient and available to any clinical practice w
ith a high-quality treatment planning system. While we use MR to image soft
tissue structures, the process could be used with any imaging modality tha
t is able to visualize the prostatic urethra (e.g., ultrasound). (C) 1998 E
lsevier Science Inc.