Keywords:
Tissue characterisation, Neoplasia, Observer performance, Imaging sequences, Diagnostic procedure, MR-Spectroscopy, MR-Functional imaging, MR, Pelvis, Oncology, Genital / Reproductive system male
Authors:
F. M. Ronza, S. V. Setola, V. Granata, R. Fusco, S. Filice, A. Di Finizio, R. Grassi, A. Rotondo, A. Petrillo; Naples/IT
DOI:
10.1594/ecr2011/C-0886
Methods and Materials
We performed Magnetic Resonance Imaging/Spectroscopy (MRI/MRS) with a 1.5 T scanner in 115 consecutive patients (mean age and SD: 67.8 ± 7.1,
from September 2009 to July 2010) with PSA values from 2.5 to 10 ng/ml,
before they underwent a 12-cores sampling primary TRUS biopsy.
In 20 patients endorectal coil was not used. MRI included high resolution transverse,
coronal,
and sagittal T2-weighted sequences.
Spectroscopy data were acquired by a 3D chemical shift imaging pulse sequence.
Peripheral areas with a low signal on T2-w images and a Choline to Citrate ratio>1 were considered suspicious for cancer (figg.1-3).
Patients were divided into 3 groups of 35 consecutive patients.
For each group we evaluated sensibility (SEN),
specificity (SPE),
negative (NPV) and positive predictive value (PPV) and accuracy (ACC) of the young radiologist,
compared to biopsy as gold standard.
The reader reviewed all exams of each group at the light of biopsies results before starting evaluation of the next group of patients.