54 posters,  4 topics,  329 authors,  96 institutions

ePostersLive® by SciGen® Technologies S.A. All rights reserved.

PP10
DCE-MRI dispersion imaging in prostate cancer

Primary tabs

Rate

No votes yet

Statistics

888 reads

Introduction

Multiparametric magnetic resonance imaging (mpMRI) is a promising option for prostate cancer (PCa) localization. Nevertheless, large multicenter studies are still lacking and recent studies show room for improvement [1]; as a result, mpMRI has not yet proven sufficient reliability to guide focal therapy and replace systematic biopsies. Additional parameters might improve mpMRI performance. In particular, due to the extravascular leakage of gadolinium contrast agents, current mpMRI methods investigate functional rather than structural features of the microvasculature, overlooking changes in the microvascular architecture. Here we propose and evaluate a method to characterize the microvascular structure by dynamic contrast-enhanced (DCE) magnetic resonance dispersion imaging (MRDI), providing a new parameter that is sensitive to those changes in the microvascular architecture associated with cancer angiogenesis.

Methods

DCE-MRI acquisition

•Magnetom Avanto (Siemens), 1.5 T, transrectal coil
•IV injection of Gadobutral (Gadovist, Bayer) 0.1 mmol/kg
•2D multislice spoiled GRE with phase oversampling
•TR/TE/FA =  50 ms / 3.9 ms / 70o
•Voxel size = 1.67 x 1.67 x 4 mm3
•Time resolution = 3.1 s/volume (7 slices)
 

DCE-MRI dispersion modeling

The proposed approach combines the extravasation model by Tofts et al. [2] with the description of the intravascular transport of the contrast agent as a convective-dispersion process [3]. Fitting measured CTCs by the obtained model leads to the generation of a parametric map related to the microvascular architecture (dispersion parameter, k) [4].

Validation

A preliminary validation was performed with 90 MRI slices recorded in 15 patients referred for radical prostatectomy due to proven Pca (AMC, University of Amsterdam, Amsterdam, the Netherlands). The histological results were compared with the classification (voxel level) by the obtained parametric maps.

Conclusions

The proposed MRDI method provided accurate localization of PCa by the intravascular dispersion parameter k, outperforming the standard leakage parameter Ktrans. MRDI may therefore contribute to the performance of mpMRI by integrating the currently lacking information on the microvascular architecture. Moreover, the extension of the method to other forms of cancer where angiogenesis is involved can also be envisaged.

 

References

[1] Isebaert et al, JMRI, 2013. 37(6):1392-1401.

[2] Tofts et al, JMRI, 1999. 10(3):223-232.

[3] Kuenen et al, IEEE TMI, 2011. 30(8):1493-1502.

[4] Mischi et al, Invest Radiol, 2014. 49(8): 561-569

Enter Poster ID (e.gGoNextPreviousCurrent