Advanced Neuroimaging

– a collaboration involving several departments at the Haukeland University Hospital and three faculties at the University of Bergen

The aim of the MMIV advanced neuroimaging project is to apply quantitative imaging and interactive visualization in studies of the human central nervous system. Deep learning is a core activity in all projects associated with the MMIV advanced neuroimaging project.

Clinical neuroradiological reading explores differential diagnosis, primarily by describing morphological abnormalities, including large vessel mapping. Assessments are generally qualitative, yet each patient is subject to extensive individual follow up with longitudinal information incorporated in all decisions (tissue characterization, treatment planning and response evaluations). Contrary to this, advanced neuroimaging research studies are most commonly cross sectional, evaluating a statistical difference between involved groups or time points. To bridge this gap between clinical imaging practice and neuroimaging research is a core goal of the project which will lead to precision diagnostics.

Our methods aim at:

feature detection (i.e. novelty in data acquisition, reconstruction and visualization)

feature extraction (i.e. novelty in data modelling and quantification)

feature prediction (i.e. linking features across time and scales – combining advanced neuroimaging, patient history and various “omics” data while exploring novel approached in medical visualization).

Core projects in the MMIV advanced neuroimaging project are: