Multimodal and three-dimensional imaging of prostate cancer

Accurate characterization of prostate cancer is crucial for treatment planning and patient management. Non-invasive SPECT imaging using a radiolabeled monoclonal antibody, In-111-labeled capromab pendetide, offers advantage over existing means for prostate cancer diagnosis and staging. However, there are difficulties associated with the interpretation of these SPECT images. In this study, we developed a 3D surface-volume hybrid rendering method that utilizes multi-modality image data to facilitate diagnosis of prostate cancer. SPECT and CT or MRI (or both) images were aligned either manually or automatically. 3D hybrid rendering was implemented to blend prostate tumor distribution from SPECT in pelvis with anatomic structures from CT/MRI. Feature extraction technique was also implemented within the hybrid rendering for tumor uptake enhancement. Autoradiographic imaging and histological evaluation were performed to correlate with the in-vivo SPECT images. Warping registration of histological sections was carried out to compensate the deformation of histology slices during fixation to help the alignment between histology and in-vivo images. Overall, the rendered volumetric evaluation of prostate cancer has the potential to greatly increase the confidence in the reading of radiolabeled monoclonal antibody scans, especially in patients where there is a high suspicion of prostate tumor metastasis. (c) 2005 Elsevier Ltd. All rights reserved.