Table 2 Datasets used for testing of the PRIMAGE platform

Imaging data

Imaging data represents the highest challenge in terms of storage and processing. In PRIMAGE data repositories, for each patient, imaging data is linkable to their available pseudonymised biological, pathological, and genetics. The use of common metadata frameworks and image analysis techniques for automated data annotation for each image is proposed to generate common repositories

Genetics and other molecular data

This project uses existing knowledge on biological biomarkers (currently on clinical use or at advanced clinical validation stage). This type of data is used in combination with imaging and clinical data, facilitating multidisciplinary big data analytics.

Clinical data

Use of natural language processing tools for automated extraction of relevant pathological data, including data on patient response to specific treatment will be extracted from the electronic health record. Data will then be structured, curated, and stored.