RSNA 2010 

Abstract Archives of the RSNA, 2010


SSA10-06

Multiparametric 3 T MR Imaging of Prostate Cancer: Histopathologic Correlation Using Customized MRI-based Specimen Molds

Scientific Formal (Paper) Presentations

Presented on November 28, 2010
Presented as part of SSA10: ISP: Genitourinary (Prostate Imaging)

Participants

Baris I. Turkbey MD, Presenter: Nothing to Disclose
Haresh Mani, Abstract Co-Author: Nothing to Disclose
Vijay Shah PhD, MS, Abstract Co-Author: Nothing to Disclose
Marcelino Bernardo, Abstract Co-Author: Nothing to Disclose
Tom Pohida, Abstract Co-Author: Nothing to Disclose
Yuxi Pang PhD, Abstract Co-Author: Employee, Koninklijke Philips Electronics NV
Dagane Daar ARRT, Abstract Co-Author: Nothing to Disclose
Yolanda Mckinney, Abstract Co-Author: Nothing to Disclose
Maria Merino MD, Abstract Co-Author: Nothing to Disclose
Peter Pinto, Abstract Co-Author: Nothing to Disclose
Peter L. Choyke MD, Abstract Co-Author: Research agreement, Koninklijke Philips Electronics NV Research agreement, General Electric Company Research agreement, Siemens AG

PURPOSE

The correlation between endorectal coil MRI and pathology has been limited by freehand cutting of the specimen by pathologist. Herein, we describe a customized specimen mold that is based on data extracted from the MRI and correlate multi-parametric MRI with the resulting registered slices.

METHOD AND MATERIALS

This prospective study includes 22 patients (mean age 59.3 years ranging between 52-74 years) with a mean PSA level of 5.76ng/mL (range, 0.49–16ng/mL). MRI of the prostate was performed on a 3T scanner using 16-channel cardiac and endorectal coils. MRI protocol included triplane T2W TSE MRI, diffusion weighted  MRI (DW MRI), MR spectroscopy (MRS), dynamic contrast enhanced MRI (DCE MRI). Following imaging, patients had robotic prostatectomy and prostatectomy specimens were placed in customized 3D molds based on extracted features from their MRI and printed on a 3D printer. MRI images were slice to slice correlated with annotated whole mount specimens obtained from the customized mold.

RESULTS

Sensitivity, specificity, and ROC area values in the peripheral zone (PZ) for T2W MRI, DW MRI, MRS, DCE MRI were 0.66, 0.6, 0.26, 0.41 and 0.93, 0.93, 0.99, 0.99 and 0.79, 0.76, 0.62, 0.7, respectively. Sensitivity, specificity and ROC area values in the central gland (CG) for T2W MRI, DW MRI, MRS, DCE MRI were 0.41, 0.56, 0.23, 0.47 and 1.0, 0.97, 0.99, 0.99 and 0.7, 0.765, 0.61, 0.73, respectively. Overall, sensitivity, specificity and ROC area values for multi-parametric MRI in PZ and CG were 0.76, 0.89, 0.82 and 0.56, 0.97, 0.76, respectively. Sensitivity, specificity and ROC area values were found to be higher in larger lesions (>5mm in greatest diameter).

CONCLUSION

Multi-parametric MRI of prostate at 3T enables accurate tumor detection in most cases. Among MR sequences, DW MRI and DCE MRI are the two most helpful for tumor detection in the central gland. This patient-specific mold technique improves co-registration and therefore correlation of MRI with prostatectomy specimens.

CLINICAL RELEVANCE/APPLICATION

 Multi-parametric MRI of prostate at 3T enables accurate tumor detection. Patient-specific mold technique improves co-registration and therefore correlation of MRI with prostatectomy specimens.

Cite This Abstract

Turkbey, B, Mani, H, Shah, V, Bernardo, M, Pohida, T, Pang, Y, Daar, D, Mckinney, Y, Merino, M, Pinto, P, Choyke, P, Multiparametric 3 T MR Imaging of Prostate Cancer: Histopathologic Correlation Using Customized MRI-based Specimen Molds.  Radiological Society of North America 2010 Scientific Assembly and Annual Meeting, November 28 - December 3, 2010 ,Chicago IL. http://archive.rsna.org/2010/9010419.html