Harald F. Busse PhD, Presenter: Nothing to Disclose

Josephin Otto, Abstract Co-Author: Nothing to Disclose

Gregor Thörmer, Abstract Co-Author: Nothing to Disclose

Nikita Garnov, Abstract Co-Author: Nothing to Disclose

Thomas Kurt Kahn MD, Abstract Co-Author: Nothing to Disclose

Michael Moche MD, Abstract Co-Author: Nothing to Disclose

Multiparametric (MP) MRI examinations of the prostate may easily generate over 500 individual MR images and, when added, a similar number of spatially resolved proton MR spectra. A dedicated software environment can be very helpful to assist the radiologist with analysis, review and reporting of these data to potentially improve detection, localization and local staging of prostate cancer.

The presented software allows for an advanced analysis and visualization of multiparametric MRI data of the prostate and is considered a valuable add-on tool for cancer diagnostics.

While this investigational software may not provide the range of features of commercial solutions, it is considered to allow for a more intuitive, complementary access to the MP information compared with routine tools on a standard MRI workstation. The software runs on any standard PC under a freely available runtime environment (IDL VM). One key option, the review of apparent findings, is the interactive display of the underlying raw data (contrast-time curves, diffusion images, MR spectra) as the mouse is moved over the corresponding maps. A unique option is the overlay of error-weighted CSI data from a widely used MRS analysis software.

A custom-made software tool has been developed under IDL (ITT VIS, Boulder, CO). Functionality and operation of the software are evaluated for different patients with biopsy-proven prostate cancer who underwent endorectal MRI examination at 3 T prior to radical prostatectomy. MP prostate images from T2-weighted (T2w), diffusion-weighted (DWI) and dynamic contrast-enhanced (DCE) scans are automatically searched for, read and sorted. Semi-quantitative perfusion parameters are computed from logistic and linear fits to the DCE time curves and overlaid as pseudo color maps on T2w and ADC maps at user-defined levels of transparency. Metabolite information from a commercial software (LCModel) may also be overlaid. Color maps of the (choline+creatine)/citrate ratio are only displayed for those voxels where the estimated standard error of the metabolite peak does not exceed a user-defined threshold (e.g. 20%). Three colors are assumed to indicate the degree of malignancy.

Busse, H, Otto, J, Thörmer, G, Garnov, N, Kahn, T, Moche, M, Custom-made Software Tool for the Analysis and Advanced Visualization of Multiparametric MRI Data of the Prostate.  Radiological Society of North America 2012 Scientific Assembly and Annual Meeting, November 25 - November 30, 2012 ,Chicago IL. http://archive.rsna.org/2012/12028575.html