E2202. Utility of Coronal Zoomed Diffusion Weighted MR Imaging for Evaluating the Prostatic Lesions
Authors
Shigeaki Umeoka;
Japanese Red Cross Society, Wakayama Medical Center
Koji Tokunaga;
Japanese Red Cross Society, Wakayama Medical Center
Tomoaki Otani;
Japanese Red Cross Society, Wakayama Medical Center
Chikara Maeda;
Japanese Red Cross Society, Wakayama Medical Center
Objective:
Coronal diffusion weighted MR imaging (DWI) is a suitable imaging for detecting prostatic lesions especially located in the periphery of the apex/base or for evaluating craniocaudal tumor extension. However, image degrading due to image distortion and artifact frequently becomes a major problem in coronal DWI. This study aims to prospectively evaluate whether newly developed zoomed DWI technique, which is known to reduce image distortion, could improve image quality of the coronal prostatic DWI, in comparison with conventional coronal DWI.
Materials and Methods:
A total of 35 patients with clinically suspicious prostatic cancer were enrolled in this study. All patients underwent coronal zoomed DWI (z-DWI) and conventional DWI (c-DWI) using a 3.0T MR scanner (Skyra fit, Siemens Healthcare) with a 60-channel body coil. Scan parameters for z-DWI were: field-of-view (FOV) 200 x 90.9 mm, matrix 110 x 50, resolution 0.91 x 0.91 x 3.0 mm, number of slices 35, TR 6000 msec, TE 65 msec, b factor 0 and 800s/mm2, scanned time 2 minutes 24 sec. FOV, matrix and scanned time of c-DWI were: 200 x 200 mm, 110 x 77, and 2 minutes 44 sec. Other parameters of c-DWI were almost same with those of z-DWI. Computed high b value DWI (b = 2000 s/mm2) were also generated based on both obtained c-DWI and z-DWI. Obtained images (c-DWIs and z-DWI) were evaluated using a five-point scale by two independent radiologists as to delineation of the contour and conspicuity of the internal structure of the prostate and/or prostatic lesions. In cases of prostatic lesions in the peripheral zone (n=15), the observers also evaluated which DWI (1. z-DWI b = 800, 2.c-DWI b = 800, computed z-DWI b = 2000, and computed c-DWI b=2000) could afford the strongest contrast of the prostatic lesions. Discrepancy was solved by the third reviewer. Agreement between the two reviewers was evaluated with Cohen kappa statistic as to 5-point scale evaluation (contour and internal structure).
Results:
Agreement between two readers was moderate (k = 0.45; contour of the prostate, k = 0.54 internal structure). In 20 of 35 cases (57%), z-DWI could provide clearer delineation of the prostate, compared with c-DWI. In 28 of 35 cases (80%), z-DWI depicted internal structure of both normal prostate and prostate lesions, more conspicuously. The scores of both contour and internal structures of z-DWI were statistically higher than those of c-DWI (p = 7.7 x 10-6; contour, p = 3.2 x 10-7; internal structure: Wilcoxon signed-rank test). In all cases of peripheral prostate lesions, computed z-DWI (b = 2000) afforded the strongest contrast of the lesions.
Conclusion:
Coronal z-DWI could be a promising tool to improve image quality and to facilitate the appropriate evaluation of the prostatic lesions. Computed DWI generated from coronal z-DWI can provide sufficient contrast of the prostatic lesions.
