MICCAI 2020

3 Alaa Bessadok 7 DAILY Mo n d a y minimal clinical data, which in this case is conventional T1-weighted MRI modality. Islem and Alaa say it has the potential to revolutionize the field of medicine by leveraging predictive intelligence tools and lowering the cost of clinical diagnosis by circumventing the need to acquire multiple brain modalities. “The construction of brain graphs is limited by the incompleteness of existing multimodal datasets,” Alaa explains. “The pre-processing pipeline, like cortical parcellation and surface registration, is time consuming for a single MRI scan. These fundamental challenges must be urgently solved so that early diagnosis of a patient with a neurological disorder is possible with high accuracy from limited or poor MRI data while predicting non-existing data observations.” Predictive intelligence is one of the most exciting research fields in medicalimage analysis . Think about it as a system that allows us to predict missing data with minimal medical acquisitions . For example, froma single conventional T1-weighted MRI, could we predict or synthesize functional MRI? Could we predict functional connectivity from morphological or structural connectivity?