The Colibri project: A multicenter shared database of magnetic resonance images about rare neurological diseases

Rare diseases occur infrequently or rarely in a population and among them, nearly 75% affect children. To date, their low prevalence and heterogeneity resulted in a lack of scientific knowledge and prevented an access to a quick and correct diagnosis. The sharing of multiple sources of information from different centers and actors may have a key role in this field. In order to address this issue, a network of Italian centers of excellence in pediatric neuroradiology launched the Colibri project with the aim to collect, share and process data about rare disease patients. The goal is to constitute a large dataset including Magnetic Resonance images and other clinical data of pediatric subjects in order to provide a service for the research and the consultation. The dataset will be made available through an architecture whose components are compliant with the DICOM Standard and will encompass a security mechanism to ensure subject privacy.

I. Introduction

In modem health care, the integration of multi-center and multi modal medical image data into comprehensive archives is an important issue for indexing an ever increasing amount of information and represents the foundation for fostering medical collaboration. Systems for collecting and integrating multi-source and multivariate data are already known in the literature such as the Cardiac Atlas Project system [1]. In the field of neuroimaging, examples also exist witnessing the importance of data sharing, such as the INDI Project collecting resting state functional Magnetic Resonance (MR) images [2], the ADNI Project addressing the Alzheimer disease [3], or the Function BIRN investigating the causes of schizophrenia [4]. All those systems envision a network of clinical sites that share images and clinical data fed by multiple repositories and datasets. They leverage systems already available, such as the LONI Image Data Archive (IDA) [5] or the eXtensible Neuroimaging Archive Toolkit (XNAT) [6] which facilitate the data storage, management and sharing. In this perspective the work illustrated in the paper represents the first Italian infrastructure for collecting, sharing and processing MR images and associated clinical data. The aim is to create an archive with information about pediatric subjects including both healthy ones and others affected by rare diseases.