mediri contributes to MICCAI 2020

 

The first article, titled "Multiple Sclerosis Lesion Segmentation Using Longitudinal Normalization and Convolutional Recurrent Neural Networks", was authored and edited mainly by our Master student Sergio Tascon-Morales. He describes how to improve the automatic segmentation of lesions on brain MRI images by taking into account the correlation to previous examinations of the same patient.

The paper by Alena-Kathrin Schnurr from the Mannheim Institute for Intelligent Systems, in which mediri played a supporting role, also deals with the temporal change of certain brain regions in MS. In "Deep Voxel Guided Morphometry (VGM): Learning regional brain changes in serial MRI", it is shown how AI in the form of neural networks can accelerate lesion segmentation on serial brain MRI images. This technique thus becomes even more interesting for clinical practice.