Sergiy Pereverzyev Jr.
Medizinische Universität Innsbruck, Austria
Aggregation with Linear Functional Strategy and Mathematics in Neuroradiology
Abstract:
Consider a linear ill-posed inverse problem (IIP), and the estimation of the linear bounded
functionals on its solution. This estimation is known as the linear functional strategy (LFS).
The LFS naturally appears in the aggregation of the regularized solutions of the IIP. In this
talk, we will discuss the use of the so-called quasi-optimality criterion (QOC) for the LFS,
and also we will show that the aggregation equipped with the QOC-based LFS leads to a very
accurate approximation of the solution of the IIP.
Afterwards, we will present several neuroradiological problems that require mathematics for their
solution. These problems are the following. Segmentation of the cervical arteries from the
magnetic resonance angiography images, and subsequent analysis of this segmentation. Prediction
of the neurodevelopmental outcome for the very preterm infants using their magnetic resonance
spectroscopy and diffusion tensor images characteristics. Classification of the Parkinson’s
disease patients using serum neurofilament light chain level and brainstem magnetic resonance
planimetry measurements. Graph-based analysis of the change of the resting state connectivity
networks under the open monitoring meditation.
<< back