A Computer Aided Diagnosis System To Detect Prostate Cancer With Endorectal Multiparametric MRI
Alessandro Savino, Valentina Giannini, A. Vignati, Simone Mazzetti, Alfredo Benso, Christian Bracco, Stefano Di Carlo, Gianfranco Michele Maria Politano, Michele Stasi, F. Russo, Daniele Regge · PORTO Publications Open Repository TOrino (Politecnico di Torino) · 2013
The tool is designed to assist Prostate Cancer diagnosis, from the patient data management to the final medical report, helping to limit the subjectivity in the MR data analysis, and reducing the requested reporting time. It provides a voxel-wise malignancy probability map, obtained by feeding into a SVM classifier a set of quantitative and semi-quantitative features, extracted from morphological (T2-w) and functional (DCE and DW) MR sequences [1]. It allows the simultaneous visualization of all MR sequences, improving the comparison of images. It supports the user with a set of image interacting tools. It supplies quantitative or semi-quantitative measurements on selected areas. It allows the reporting of annotations related to one of the 16 prostatic regions of interest