The lesions in the brain white matter of the brain called white matter hyperintensities may cause significant functional impairment and are commonly associated with neurological or psychiatric diseases. The current analysis of these lesions is manually performed by experts in magnetic resonance imaging, and represents a non-trivial, costly and subjective task. Thus, the proposed approach aims to create a helpful tool in the identification and characterization of injured tissues in the brain white matter, as well as enable the tracking of patients over time through structural metrics extracted from the previously identified lesions. The methodology to be developed will combine and propose techniques in the area of image processing, such as preprocessing, segmentation, representation and description, and in the area of pattern recognition, such as characterization, discrimination between classes, classification and prediction, to solve the present problem.
News published in Agência FAPESP Newsletter about the scholarship: