Automatic analysis of perinatal arterial ischemic stroke with MRI

Image segmentation and artefact reduction techniques for perinatal arterial ischemic stroke








About project

Perinatal arterial ischemic stroke (PAIS) has an incidence of approximately 1:5000 live born infants. Currently, two potential treatments for this type of stroke are being evaluated, these are recombinant human erythropoietin (rhEPO) and mesenchymal stromal cells (MSCs). Evaluating the efficacy of these treatments relies on measurements of the volume of various brain tissues and the ischemic lesion on MRI. Manual segmentation is a time consuming process, which takes approximately a week for each scan. The goal of this project is to first develop an image analysis pipeline that automatically segments these scans, and second, to further improve the quality of the segmentations by developing methods to reduce the impact that image artefacts have on the segmentation quality. This project is done in collaboration with Prof Manon Benders from the Wilhelmina Children’s Hospital of the UMC Utrecht.


Our publications & software