8/2021 - Development of digital twins for quantitative imaging based on nuclear magnetic resonance

Updated: over 2 years ago
Deadline: 19 Oct 2021

The aim is the development of numerical models and advanced data processing techniques for generating digital twins to support quantitative diagnostics based on nuclear magnetic resonance. The digital twins will be designed by merging experimental data (from inter-laboratory comparisons) with synthetic data. The latter will be produced by means of an MRI simulator, based on the Bloch equations, and specifically developed for creating realistic images, as a ground truth for signal generation and image production process. The simulator, with supervised machine learning techniques, will allow us to develop an extended database, to be used as a validation tool for quantitative analysis, thus improving the reliability of MRI in the quantification of several biomarkers, e.g. iron levels and proton density fat-fraction in tissues. Particular attention will be devoted to the creation of FAIR metadata that can be reprocessed by ad hoc artificial intelligence techniques.



Similar Positions