J. Exp. Clin. Cancer Res., 24, 4, 2005 MR Spectroscopy of Prostate Cancer. Initial Clinical Experience E. Squillaci, G. Manenti, S. Mancino, M. Carlani, M. Di Roma, V. Colangelo, G. Simonetti Dept. of Diagnostic Imaging and Interventional Radiology, University "Tor Vergata", Rome, Italy Aim of the study was to evaluate the effectiveness of proton MR Spectroscopic (MRS) imaging in the detection and localization of prostate cancer, prospectively compared with histopathologic findings. Magnetic Resonance (MR) and MRS imaging were performed in 65 patients with high levels of prostate-specific antigen (PSA) and suspicious areas at the transrectal ultrasound (TRUS). At MR areas of interest were reported as normal, equivocal or suspicious. At MRS imaging, cancer was diagnosed as "possible" if the ratio of choline plus creatine to citrate exceeded 2 SDs above mean normal peripheral zone values or as "definite" if the ratio exceeded 3 SDs. All patients underwent a TRUS 10-core biopsy within 30 days of the imaging study. MR alone showed sensitivity, specificity, positive predictive values, negative predictive values and accuracy for detection of prostate cancer of 85%, 76%, 53%, 91% and 65%, respectively, whereas MRS alone showed 89%, 77%, 78%, 69% and 83%, respectively. These values were 71%, 90%, 89%, 74% and 80% when the prostate was evaluated combining MR and MRS. The addition of MRS to the MR imaging provides a higher specificity in tumour detection and can be recommended as a problem-solving modality for patients with elevated PSA levels and suspicious TRUS before biopsy. Key Words: Prostate, Prostatic cancer, Biopsy, Magnetic Resonance Imaging, Magnetic Resonance Spectroscopic Imaging Prostate carcinoma is an important medical and socio-economical problem due to its increasing incidence among the male population and the steady ageing of the western countries. The American Cancer Society estimated 230,000 new cases of prostate carcinoma and 29,900 cases of related deaths in the United States in 2004 (1). A yearly test of the PSA levels combined with a digital rectal examination (DRE) and a transrectal ultrasonography (TRUS) completed with a biopsy (2) have allowed the detection of cancer at an earlier stage compared to the past. This has enabled the use of treatments such as brachytherapy, intensity-modulated radiotherapy, thermal ablation or cryosurgery that are less aggressive and show less morbidity than radical prostatectomy. It is, therefore, of the outmost importance the precise localization and staging of the neoplasm to evaluate its biological aggressiveness. Magnetic Resonance (MR) imaging allows a better definition of the morphology and of the anatomy of the prostate gland. To date, however, the diagnos- tic accuracy varies widely between authors ranging from 75 to 90% and reaching a 97% (3) in the detection of known lesions, but showing a lower accuracy in case of lesions smaller than 5 mm (4). The morphologic imaging alone, however, has a low specificity (55%) (5). In fact, prostatitis, post-biopsy bleedings or the effects of therapy may show a characteristic low intensity signal in the T2 weighted sequences thus mimicking the neoplastic lesion. The introduction of the 3D Magnetic Resonance Spectroscopic (MRS) imaging has enabled the correlation of both anatomical and morphological data with functional informations about tissue metabolism. Combined MR/MRS imaging studies show more accuracy in the detection of prostate cancer enabling the simultaneous displaying of both extension and biological aggressiveness (6). Purpose of this study was to define the effectiveness of MRS imaging compared to MR imaging in the detection and localization of prostatic neoplastic foci prospectively assessed with postbiopsy histopathological examination. 523 E. Squillaci et al. Materials and Methods Patients. Between April 2004 and January 2005, 65 consecutive patients (mean age 68.5 years, range 55-82) underwent MR and MRS of the prostate. Inclusion criteria were: a) free-PSA level higher than 4.0 ng/ml or; b) suspicious (<15%) free-to-total PSA ratio measured by immunoassay 20 days prior to MR. Patients included in the study had a mean free PSA value of 24.15 ng/ml (range 4.3-44 ng/ml). Patients (with unknown prostate cancer) were referred for MR with evidence of focal or diffuse hypoechoic lesion/s at TRUS performed up to 10 days before MR. Informed consent was obtained from all patients before MR imaging. MR Technique. MR imaging was performed with a 1.5 T scanner (Philips Gyroscan Intera, Best, The Netherlands) with 30 mT/m gradients by using the body coil for radiofrequency (RF) excitation and a surface coil (C1) placed on the pubic region for signal acquisition. To reduce bowel peristalsis and improve imaging quality 1 mg glucagon (Glucagen; Novo Nordisk, Mainz, Germany) or 20 mg butylscopolamine (Buscopan; Boehringer, Ingelheim am Rhein, Germany) were administered before imaging. Morphologic study was obtained with high-spatial-resolution sagittal T2-weighted turbo spin-echo (TSE) sequences to assess coil position, then coronal and axial T2-weighted TSE images were obtained from the apex to the seminal vesicles with the following parameters: repetition time/echo time msec (TR/TE) 4.000/130 msec, 2.5 section thickness, no intersection gap, 6 signal acquired (NSA), 19-cm field of view (FOV), 240 and 256 (acquisition and reconstruction) matrix, scan time 8:47 min. T2weighted with fat suppression (SPIR) sequences were also obtained (4.750/90 msec TR/TE). The morphologic study was completed with 1H-MRS imaging and dynamic contrast-enhanced MR (DCE-MR) imaging. 3D MR spectroscopic imaging volume was selected on the high-spatial-resolution axial T2-weighted images. The sequence applied was the point resolved spectroscopic (PRESS) with the following parameters: TR 1.600 msec, TE 136 msec (echo delay optimized) for quantitative citrate and choline multivoxel detection, spectral bandwidth 1.000 Hz, 512 points, FOV 14-cm, n. image 4, scan resolution 16, voxel spatial resolution 0.24 cm3, scan time 17:25 min. Band-selective inversion with gradient dephasing 524 (BASING) sequence was used for the optimal water and lipid suppression. DCE-MR imaging was performed after the spectroscopic acquisition in all patients with fast field-echo (FFE) T1-weighted sequence and 20 mL /sec singledose bolus injection of gadopentate dimeglumine (GdDTPA, Magnevist®; Schering, Berlin, Germany) with the following parameters: TR 20 msec, TE 5 msec, flip angle 20°, slice thickness 3 mm, FOV 18-cm, scan number 10, scan time 2:57 min. These sequences were used to show the contrast media uptake for the evaluation of quick contrast-enhancement of the peripheral zone, synchronous with the central gland suggesting neoangiogenesis of the lesion. MR imaging and MRS imaging data analysis. Images were examined by an expert radiologist (E.S.) aware of the TRUS findings. Only morphologic images included in the MRS imaging were evaluated to allow a direct comparison between MR and MRS imaging. Peripheral zone limits were assessed on the axial T2-weighted where the prostatic capsule and pseudo-capsule characteristically show a low-intensity signal band (7). The presence of a peripheral zone neoplasia was evaluated basing on the following criteria: 1) the absence of low-intensity signal areas was considered a normal pattern; 2) the presence of a non-homogeneous low-intensity signal area with irregular edges or more triangular streaky lesions was considered equivocal; 3) the presence of a homogeneous lowintensity signal smudgy-appearing area with rounded sides and blurred outlines was considered highly suspicious for neoplasia. Single voxels of the spectroscopic selected volume sequence were two-dimensionally plotted on a grid superimposed on the related T2-weighted axial images. The spectra related to the prostatic metabolites were automatically generated with the manual pointing of ROIs (Region of interests) for each voxel represented on the grid. Voxels with abnormal contrast-enhancement in the dynamic sequences were further analyzed (Fig.1). The integral values of the areas under the peak of the prostatic metabolites citrate, choline and creatine were calculated using a dedicated software for the spectroscopic analysis, after frequency and baseline correction. To detect the presence of prostatic cancer the ratio of the integral values of choline plus creatine to citrate (Cho+Cr/Cit) was calculated for each voxel. "Possible" presence of carcinoma was defined when this ratio was greater than 2 standard deviations (SDs) MR Spectroscopy of Prostate Cancer Fig. 1 - Dynamic contrast-enhanced axial T1-weighted FFE images: acquisitions before (A) and 20 seconds after contrast media injection (B) show an area of early wash-in Gd-DTPA contrast enhancement in the middle-left peripheral zone (arrow). In the acquisition after 45 seconds (C) the high intensity signal increases (arrow) compared with the remaining peripheral zone. After 120 seconds there is a synchronized Gd-DTPA wash-out of the peripheral zone lesion and of the adenomyoma (D). compared with normal prostatic tissue (>0.75); whereas a ratio greater than 3 SDs compared with normal tissue (>0.86) was defined as "definite" cancer. Voxels with ratio values lower than 0.75 were defined as normal tissue (8). Suspicious or equivocal areas at the morphological imaging and voxels with (Cho+Cr/Cit) ratio >3 SDs and between 2 and 3 SDs were spatially assigned to different regions of the prostate gland: apical, intermediate, lateral and basal region for each side. This kind of subdivision was used to correlate MR and MRS imaging results with the histopathologic findings (Fig.2). Prostatic Biopsy. Each patient underwent systematic US-guided prostatic biopsy (trans-perineal) performed with an ATL (Advanced Technology Laboratories) HDI 5000 SonoCT Diagnostic Ultrasound Machine with convex 7.5 MHz sector probe within 30 days from the spectroscopic session. In all patients, a 10-core biopsy was performed according to the octant subdivision of the prostate 525 E. Squillaci et al. cious, >3 SDs and >2 - <3 SDs) and negative (normal and < 2 SDs). DCE-MR imaging data were not used for the statistical analysis. Results Fig. 2 - Coronal view scheme of the subdivision of the prostate gland. This scheme was used to correlate the MR and MRS imaging findings with the biopsy results. gland (1 sample from the apex, the intermediate zone and the basal region, 2 samples from the lateral zone, for each side). At least 2 samples were obtained from the suspicious zone or from the area showing pathologic characteristics at MR and MRS. This additional sampling was used to compensate for the inaccuracy in the correlation between the sextant biopsy technique and the morphologic and "voxelby-voxel" spectroscopic imaging. All samplings were performed using a BioPince® Full Core Biopsy Gun with an 18 gauge needle (MD Tech., Florida, USA). Specimens were classified referring to the biopsy sites and then an individual histological analysis for each sample site was performed. Each biopsic specimen was analyzed by a single pathologist specialized in prostatic pathology. Statistical Analysis. All data were analyzed using the Statistical Package for Social Sciences, (SPSS® V.12, Chicago, Illinois). On the basis of the histopathological results, MR imaging, MRS imaging and the combination of MR and MRS imaging (MR/MRS) findings were encoded in descriptive statistics including sensitivity, specificity, accuracy and positive (PPV) and negative (NPV) predictive value. The Pearson χ2 test was used for the evaluation of the statistical significance (p < 0.05) of the single techniques and their combination, compared with the histopathologic finding. For the statistical analysis the imaging results, classified as normal, equivocal or suspicious, were divided in two groups: positive (equivocal and suspi526 In 43 out of the 65 patients (66%) a focal or diffuse low-intensity signal area was detected, mono- or bilaterally in the axial T2-weighted images, whereas in the remaining 22 patients (34%) no pathologic distortion of the intensity signal was detected. Particularly, the morphologic findings of the MR imaging were correlated to the spatial localization of the prostate gland, with 8 suspicious areas in the apex (5 left side, 3 right side), 13 in the intermediate zone (4 left side, 9 right side), 15 in the lateral zone (8 left side, 7 right side) and 7 in the basal region (3 left side, 4 right side). Of the 43 patients with suspicious neoplastic lesions, 28 (65% true-positive TP) had histopathologic confirmation of adenocarcinoma, whereas 15 (35% false-positive FP) had diagnosis of benign lesions (9 prostatitis and 6 nodular benign hyperplasia). Of the 22 patients without suspicious images at MR imaging, 17 (77% true-negative TN) had histopathologic confirmation of the absence of malignant lesions (10 cases of chronic prostatitis, 7 cases of nodular benign hyperplasia), whereas the remaining 5 patients (23% false-negative FN) had diagnosis of adenocarcinoma. On the basis of these results, MR imaging has shown a sensitivity equal to 85%, a specificity of 53%, PPV 65%, NPV 77% and an accuracy of 69% with a p value 0.013 (Table II). MRS imaging has detected pathologic spectra (Cho+Cr/Cit ratio > 3 SDs or >2 - <3 SDs) in 28 (43%) patients, whereas in 37 (57%) the spectral analysis has produced no-pathologic ratios (Cho+Cr/Cit < 2 SDs). Suspicious areas were also related to the spatial octant partitioning of the prostate gland therefore allowing the detection of 3 areas with pathologic ratios in the prostatic apex, 12 in the intermediate zone (7 left side, 5 right side), 11 in the lateral zone (5 left side, 6 right side), 2 in the basal region. Of the 28 patients with Cho+Cr/Cit pathologic ratio, 25 (89% TP) confirmed the presence of adenocarcinoma (Fig.3), whereas 3 (11% FP) were diagnosed as benign lesions (2 cases of chronic prostatitis, 1 case of nodular benign hyperplasia). In the 37 patients with Cho+Cr/Cit < 2SDs, 29 (78% TN) generated histopathologic findings of MR Spectroscopy of Prostate Cancer benign lesion (18 cases of prostatitis and 11 cases of nodular benign hyperplasia) (Fig.4) while in the other 8 patients (22% FN) a diagnosis of adenocarcinoma (Table I). MRS imaging demonstrated a 76% sensitivity, 91% specificity, 89% PPV, 78% NPV and an 83% accuracy with a p value of 0.000 (Table II). MR imaging combined with MRS imaging results showed a sensitivity equal to 71%, a specificity of 90%, PPV 89%, NPV 74% and an accuracy of 80% with a p value of 0.000 (Table II). No significant correlations were found between the MR and the MRS results and the bioptic sampling findings, except in case of MRS imaging results and histopathologic findings for the intermediate and lateral zones (p = 0.003 and p = 0.002, respectively). Discussion Nowadays, prostatic cancer represents the most common cause of cancer-related death in the male population (1). Despite the high mortality rate, many cases of prostate cancer are subclinical and accidentally detected at autopsy (9). Fig. 3 - Sagittal (A) and axial (B, C) T2-weighted images of the prostate showing a peripheral middle-right area of low intensity signal. The spectrum obtained from this area shows elevated levels of choline and reduced levels of citrate, a pattern consistent with definite cancer (D). Histopathologic analysis confirmed the neoplastic nature of the lesion. 527 E. Squillaci et al. Fig. 4 - Axial T2-weighted images of the prostate showing an area of low intensity signal in the left peripheral zone (A, B, C). The MRS spectrum demonstrated choline and citrate levels indicative of normal tissue (D). Post-biopsy histopathologic analysis showed the absence of neoplastic cells in this area, confirming what MRS had correctly demonstrated whereas MR imaging findings could have been misleading. The available non-invasive diagnostic techniques have large limitations in discriminating between latent and aggressive or progressive neoplastic lesions (10,11). Olson et al. demonstrated that TRUS depicts up to 30% of palpable DRE lesions (12). In addition, these authors showed that this technique has a high level of FP because only 20% of hypoechoic lesions are malignant. US-guided biopsy has also a high percentage of sampling errors due to the small part of the prostatic gland effectively analyzed. Thus, so many patients with abnormal PSA levels (>4ng/ml) that undergo systematic sextant biopsies 528 have negative results and, therefore, need to be subjected to repeated biopsy (13,14). The development of new therapies for prostatic cancer, less invasive than radical prostatectomy, requires detailed localization and the assessment of the extent of the disease to be able to treat the patients with a targeted therapy. This kind of approach increases the therapy effectiveness and reduces the related morbidity. The detailed knowledge of the lesion location significantly helps patients with persistent elevated PSA levels that undergo repeated biopsies. MR has demonstrated a good sensitivity (78%) MR Spectroscopy of Prostate Cancer Table I - Correlation of MR imaging, MRS and biopsy findings Biopsy findings Negative Positive Total MR MRS Negative Positive Negative Positive 17 5 22 15 28 43 29 8 37 3 25 28 and a low specificity (55%) in the detection of neoplasm site/neoplastic foci due to the high percentage of FP (post-biopsy bleedings, inflammation, fibrosis etc.) (15). The combination of metabolic data using the spectroscopic imaging add information to the morphologic imaging in the detection, staging, size evaluation and assessment of the degree of aggressiveness of the neoplasia (8,16,17). However, most experiences are the result of retrospective studies where diagnostic accuracy of the MR and MRS imaging was assessed on the basis of the histopathologic data after radical prostatectomy. The purpose of this study was to evaluate the MR and MRS imaging capability in the detection and detailed localization of prostate cancer. Every single technique result was prospectively compared with bioptic findings. Surface coil was used to avoid the scarse tolerability and the typical artifacts of the endorectal coil (18). MR imaging has demonstrated an 85% sensibility and a 53% specificity in detecting prostate cancer in accordance with previous studies (8). MRS imaging has shown high specificity in depicting prostate cancer compared to the MR (91% against 53%). MR and MRS imaging results suggesting the presence of prostate cancer, resulted in high probability to be neoplastic lesions (PPV 88%); otherwise, a negative result excludes the presence of prostate cancer with lower probability (77% NPV) due to the number of FN. These may be ascribed to the typical feature for the 1.5 T magnetic fields high nominal value (0.24cm3) of the voxel, with overlapping of the pathologic and notpathologic data in the same sampled voxel. Furthermore, the limited aggressiveness of some carcinomas may explain the lack of detection of the neoplastic lesions at the MRS imaging. This has been already described by some authors who have demonstrated that small tumors with low grade differentiation Table II - Sensitivity, specificity, PPV, NPV, accuracy of MR, MRS and combined MR/MRS imaging Sensitivity Specificity PPV NPV Accuracy MR (%) MRS (%) MR/MRS (%) 85 53 65 77 69 76 91 89 78 83 71 90 80 89 74 (Gleason 4 and 5) may be missed because of the slight modifications of the citrate and choline levels (19). Initial results with the 3T MRS imaging in prostate cancer patients offer higher signal-noise ratio (SNR) (20,21) and more accurate spectral resolution with increased accuracy in the spatial detection of the metabolic data (22,23). Inadequate correlation between the MR/MRS imaging and the histopathologic findings as to the spatial localization of the lesions impairs the ability of our study to demonstrate if tumor localization affects the two imaging techniques to detect it (24). That is directly related to the fact that the US-guided biopsies do not assure an exact correspondence between bioptic sites and the suspicious areas at the MR and MRS imaging. Only radical prostatectomy results can afford to definitely answer this question. In conclusion, this initial study, performed on a small population, demonstrates that MR/MRS imaging might be indicated in patients with elevated PSA levels and TRUS findings suspicious for neoplasia, who are candidate to bioptic sampling. In view of the excellent feasibility of these two techniques performed with a surface coil, this imaging session provides new important information to those patients with random bioptic findings negative for pathologies and persisting elevated PSA levels that are candidate to a second bioptic sampling. The detection of suspicious areas at the MR and MRS imaging offers the potentiality to perform targeted MR-guided biopsies (25,26) without the need to compare MR imaging and TRUS findings (24). References 1. American Cancer Society, Inc. Estimated new cancer cases and deaths by gender: US 2004. Cancer Facts and Figures 2004:4. 529 E. Squillaci et al. 2. Waterbor J.W., Bueschen A.J.: Prostate Cancer screening (United States). Cancer Causes Control 6:267-274, 1995. 3. Ikonen S., Karkkainen P., Kivisaari L., et al.: Magnetic resonance imaging of clinically localized prostatic cancer. J. Urol. 159:915-919, 1998. 4. Ellis J.H., Tempany C., Sarin M.S., Gatsonis C., Rifkin M.D., McNeil B.J.: MR imaging and sonography of early prostatic cancer: pathologic and imaging features that influence identification and diagnosis. AJR. 162:865-872, 1994. 5. Hricak H., White S., Vigneron D., et al.: Carcinoma of the prostate gland: MR imaging with pelvic phased-array coils versus integrated endorectal-pelvic phased-array coils. Radiology 193:703-709, 1994. 6. Kurhanewicz J., Swanson M.G., Nelson S.J., Vigneron D.B.: Combined magnetic resonance imaging and spectroscopic imaging approach to molecular imaging of prostate cancer. JMRI 16:451-463, 2002. 7. Hricak H., Dooms G.C., McNeal J.E. et al.: MR imaging of prostate gland: normal anatomy. AJR. Am. J. Roentgenol. 148:51-55, 1987. 8. Scheidler J., Hricak H., Vigneron D.B., et al.: Prostate cancer: localization with threedimensional proton MR spectroscopic imaging-clinicopathologic study. Radiology 213:473480, 1999. 9. Stamey T.A., Mcneal J.E.: Adeno Carcinoma of the Prostate. In: Walsh PC, Retik AB, Stamey TA, Vaughan ED Jr (eds.). Campbell's Urology. Philadelphia, WB Saunders, 1992, 1159-1221. 10. Mettlin C.J., Black B., Lee F., Littrup P.J., DuPont A., Babaian R. et al.: Workgroup #2: screening and detection-reference range/clinical issues of PSA. Cancer 71:2679-2680, 1993. 11. Smith J.A., Scardino P.T., Resnick M.I., Hernandez A.D., Rose S.C., Egger M.J.: Transrectal ultrasound versus digital rectal examination for the staging of carcinoma of the prostate: results of a prospective multi-institutional trial. J. Urol. 157:902-906, 1997. 12. Olsson C.: Prostatic cancer. Kidney Int. 43:955-965, 1993. 13. Keetch D.W., Catalona W.J., Smith D.S.: Serial prostatic biopsies in men with persistently elevated serum prostate specific antigen values. J. Urol. 151:1571-1574, 1994. 14. Ellis W.J., Brawer M.K.: Repeat prostate needle biopsy: who needs it? J. Urol. 153:1496-1498, 1995. 15. White S., Hricak H., Forstner R. et al.: Prostate cancer: Effect of postbiopsy hemorrhage on interpretation of MR images. Radiology 195:385-390, 1995. 16. Yu K.K., Scheidler J., Hricak H., et al.: Prostate cancer: prediction of extracapsular extension with endorectal MR imaging and three-dimensional proton MR spectroscopic imaging. Radiology 213:481-488, 1999. 17. Coakley F.V., Kurhanewicz J., Lu Y., et al.: Prostate cancer tumor volume: measurement with endorectal MR and MR 530 spectroscopic imaging. Radiology 223:91-97, 2002. 18. Husband J.E., Padhani A.R., Macvicar A.D., Revell P.: Magnetic Resonance Imaging of Prostate Cancer: Comparison of Image Quality Using Endorectal and Pelvic Phased-Array Coils. Clinical Radiology 53:673-681, 1998. 19. Vigneron D.B., Males R., Noworolski S., et al.: 3D MRSI of prostate cancer: correlation with histologic grade (abstr). In: Proceedings of the Sixth Meeting of the International Society for Magnetic resonance in Medecine. Berkeley, California: International Society for Magnetic Resonance in Medicine: 487, 1998. 20. Sosna J., Pedrosa I., Dewolf W.C., Mahallati H., Lenkinski R.E., Rofsky N.M.: MR Imaging of the Prostate at 3 Tesla: Comparison of an External Phased-array Coil to Imaging with an Endorectal Coil at 1.5 Tesla. Acad. Radiol. 11:857862, 2004. 21. Bloch B.N., Rofsky N.M., Baroni R.H., Marquis R.P., Pedrosa I., Lenkinski R.E.: 3 Tesla Magnetic Resonance Imaging of the Prostate with Combined Pelvic Phased-array and Endorectal coils: Initial Experience. Acad. Radiol. 11:863867, 2004. 22. Futterer J.J., Scheenen T.W., Huisman H.J., et al.: Initial Experience of 3 Tesla Endorectal Coil Magnetic Resonance Imaging and 1H-Spectroscopic Imaging of the Prostate. Invest. Radiol. 39(11):671-680, 2004. 23. Kim H-W., Buckley D.L., Peterson D.L., et al.: In Vivo Prostate Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy at 3 Tesla Using a Transceive Pelvic Phased Array Coil. Invest. Radiol. 38:443-451, 2003. 24. Beyersdoff D., Taupitz M., Winkelmann B., et al.: Patients with a History of Elevated Prostate-Specific Antigen Levels and Negative Transrectal US-guided Quadrant or Sextant Biopsy Results: Value of MR Imaging. Radiology 224:701706, 2002. 25. D'Amico A.V., Tempany C.M., Cormack R., et al.: Transperineal magnetic resonance image guided Prostate biopsy. J. Urol. 164:385-387, 2000. 26. Beyersdoff D., Bernd Hamm A.W., Lenk S., Loening S.A., Taupltz M.: MR Imaging-guided Prostate Biopsy with a Closed MR Unit at 1.5T: Initial Results. Radiology 234:576581, 2005. Received: June 16, 2005 Prof. Ettore Squillaci University of Rome "Tor Vergata" P.T.V. Viale Oxford, 81 00133 Rome, Italy Tel.: +39 06-20902401; Fax: +39 06-20902404 E-mail: [email protected]
© Copyright 2024