Standardization and data homogenization
Data preparation and processing paradigms: Successful extraction of morphometric biomarkers from imaging data depends critically on processing pipelines that are tuned to specific study constraints. While effective algorithms exist for many of these tasks, the know-how of optimal use and parameterization is a valuable resource. qbig, in collaboration with MIAC, is actively developing systems and development platforms that support building and using tailored image processing tasks in a standardized modular framework. Such modular development and pipelining platforms significantly accelerate the adaptation and development of new pipelines and target metrics.
Phantom development: Phantoms are physical or digitally simulated objects of known geometry and/or composition with a reduced number of experimental confounds with respect to living subjects. They are indispensable tools in any type of MRI validation. Existing physical brain phantoms only very limitedly mimic true tissue properties and anatomy. Indeed, while some phantoms do not match the true relaxation properties of living tissue, others are characterized by substantially simplified contours of the brain or brain regions. For this reason, they rarely find application in the image processing and reconstruction fields. To overcome those limitations, we suggest a modular framework for the construction of 3D-printed brain phantoms, which realistically mimic both anatomy and relaxation properties of human brain regions.