Keywords:
Cancer, Staging, Computer Applications-General, Computer Applications-Detection, diagnosis, MR-Spectroscopy, MR-Diffusion/Perfusion, MR, Pelvis, Oncology, Computer applications
Authors:
J. Woetzel1, M. Barann1, A. Genz1, J. Plath1, O. Konrad1, U. Siems1, B. Merkel1, M. Röthke2; 1Bremen/DE, 2Heidelberg/DE
DOI:
10.1594/ecr2013/C-1391
Methods and Materials
Our novel approach for workflow-optimized classification of prostate lesions is following the ideas of the scoring system described by the authors of the well-known ESUR prostate MR guidelines 2012 [1,2].
Technical foundation of our implementation method is the Rondo software platform [3].
A collection of framework functions for data access,
pre-processing,
and visualization is offered,
which includes spatially aligned viewers of T2w,
diffusion,
perfusion (both ADC and DCE),
and spectroscopic MRI data by either overlaid views or side-by-side automatic diagnostic hanging layouts.
Flexible extension points based on MeVisLab interfaces and Python scripting are programmatically used to integrate a complementary,
dedicated PI-RADS reporting solution into thorough clincial workflows.
A standardized 27 region localization scheme and a five level modality scoring system is used to describe lesions.
Semi-automatic tools were optimized for volumetric measurements and furthermore an optional detailed analysis of suspicious findings.
The implementation is based on a client/server protocol and allows for remote usage of the software.