Cognitive Radiology for Big Data Projects and In Silico Trials

Radiology departments have large databases of imaging studies acquired over the years, but there is a tremendous opportunity for improvement in the organisation and large-scale exploitation of these images and associated reports, for example for the development of clinical trials. Adelaide, a Eurostars-funded cognitive radiology project developed by Cetir, Medexprim and Vicomtech, aims to provide a solution that improves the search, extraction, anonymisation and labelling capabilities of images and associated clinical data. 
In this framework, Vicomtech brings its expertise in medical image processing by collaborating in the development of deep learning algorithms to classify and detect structures in medical images using AI. Regarding the mining of radiological reports, Vicomtech is also involved in natural language processing and machine learning applied to clinical texts. Image analysis and radiology reports automatically generate relevant labels about image content and pathologies present, which have been introduced in the search tool, facilitating a more efficient and focused extraction and generation of databases of radiological studies. The aDELAIDE solution enables the discovery of new imaging biomarkers and progress towards personalised medicine.