Validation of functional imaging with pathology for tumor delineation in the prostate

摘要: Abstract Introduction A study was performed to validate magnetic resonance (MR) based prostate tumor delineations with pathology. Material and methods Five patients biopsy proven cancer underwent a T2 weighted (T2w), diffusion MRI (DW-MRI) dynamic contrast-enhanced (DCE-MRI) scan before prostatectomy. Suspicious regions were delineated on all available MR information. After prostatectomy whole-mount hematoxylin–eosin stained (H&E) sections made. Tumor tissue the H&E compared delineations. The registration accuracy between images estimated. Results coverage of 44–89% reached by application margin ∼5 mm yielded 85–100% in patients. Errors created during procedure 2–3 mm, which cannot completely explain limited coverage. Conclusions An accurate processing method presented (registration error 2–3 mm), enables validation Reasonable about 85% larger found when applying ∼5 mm