Improving QMRI By Realizing trustworthy integration of AI in Neuro-imaging

MRI is a key methodology in modern neuroimaging, but conventional MRI relies on visual interpretation of intensity differences in the images, which is heavily dependent on scanner settings.

Quantitative MRI (qMRI) is an attractive alternative MRI method that allows quantitative measurement of physical tissue parameters, enabling objective comparison between patients and across time.

Moreover, qMRI facilitates early detection of pathological changes in the brain resulting from neurological disorders such as multiple sclerosis.

Unfortunately, and despite the demonstrated potential in research settings, the implementation of qMRI in routine clinical practice remains limited due to long scan and post-processing times.

While recent developments in artificial intelligence have the potential to accelerate and improve medical imaging pipelines, reduced transparency about the underlying processes, the lack of training data sets and limited information about the accuracy of the results has limited its use for clinical qMRI applications so far.In IQ-BRAIN, we propose a unique research and training programme that tackles this urgent need for improved and accelerated qMRI methodology for neuroimaging applications.

By integrating both physics-based models and trustworthy artificial intelligence methods along the qMRI pipeline, our innovative approach combines the best of both worlds.

IQ-BRAIN will prepare the next generation of qMRI specialists trained in the different aspects of the qMRI-neuroimaging pipeline, that can bridge the gap between qMRI method development and clinical need.

Through a training programme of network-wide events, international secondments, and strong interaction between partners from academia, industry and hospitals, IQ-BRAIN offers early-stage researchers a rich combination of knowledge, expertise and essential transferable skills that prepares them for a thriving career as R&D professionals in the qMRI field.