This new medical imaging method, based on standard MRI sequences, can obtain nosological maps of the lesion. These are colour-coded to differentiate regions and tissue sub-types of the lesion according to the patient’s biological processes, diagnosis and/or prognosis.

Applications

• Diagnosis, therapeutic planning, treatment monitoring and clinical prognosis of patients affected by a Glioblastoma-type tumour based on the analysis and generation of advanced imaging biomarkers.
• Information and Decision Support Systems (DSS) for medical professionals.

Technology Overview

The invention consists of a method that obtains a single map from standard magnetic resonance sequences where the different types of tissues and sub-regions of the lesion are represented in a differentiated and recognisable way by the medical expert. The result is called a nosological image and is obtained automatically without professionals needing intervention.

The system comprises an imaging module and an image processing module, which manages the multi-parametric image (MPI) stack and creates the nosological map of the lesion.

The method comprises three stages: first, pre-processing is performed for MPI image stack enhancement where noise filtering, inhomogeneity correction, extra-meningeal tissue removal, image registration and super-resolution operations are performed. The second phase is image feature extraction, which feeds the unsupervised classification and segmentation module. Finally, based on the classification results, the generation of the nosological map and the automatic identification of biological profiles is carried out using probabilistic maps of registered tissues corrected with patient-specific information and statistical techniques for comparison and characterisation of the identified tissue distributions.