Value of image fusion in the staging of prostatic carcinoma

Abstract

We assessed the value of image fusion in the staging of prostatic cancer in a series of 32 patients who underwent preoperative evaluation with transrectal colour-Doppler ultrasonography (TRUS) and magnetic resonance imaging (MRI). Colour-Doppler TRUS exams were performed using a 7.5-MHz biplanar probe. MRI exams were done with a scanner operating at 1.5 Tesla (T) using an endorectal coil. All patients underwent radical prostatectomy within 2 weeks from the imaging assessment. Whole-mount sections were prepared from the surgical specimens and were subsequently digitised by using a high-resolution scanner. The Digital Imaging and Communications in Medicine (DICOM) TRUS and MR images as well as the digitised pathological images were transferred to a graphic workstation to perform image fusion. Image fusion was technically possible in 25/32 cases in which axial TRUS images were available. The following fusion images were obtained: TRUS + pathological sections; MRI + pathological sections; TRUS + MRI + pathological sections. The final pathological staging concerning the T status was: four pT2b, fourteen pT2c, three pT3a and four pT3b. The three types of image fusion led to the following results: TRUS + pathological sections, correct staging in 20/25 cases (accuracy 80%); MRI + pathological sections, correct staging in 22/25 cases (accuracy 88%); TRUS + MRI + pathological sections, correct staging in 23/25 cases (accuracy 92%). Our study suggests that by using image fusion between colour-Doppler TRUS and endorectal MRI, it is possible to improve the accuracy of pathological staging in patients who are candidates for radical prostatectomy.