PhD project by Maria Ines Meyer

Name: Maria Ines Meyer
Project Title: Multi-scanner/multi-center evaluation of the progression of multiple sclerosis
Group: Medical Image Computing
Supervisor: Koen Van Leemput
Co-Supervisor: Diana M. Sima

Project Description:
Multiple Sclerosis (MS) is a chronic disease where the immune system attacks the central nervous system. Over time, this condition can lead to severe disability, but a correct course of treatment can slow down the progression of the disease. To optimize treatment, patients with MS undergo regular follow-up, including the acquisition of brain images through Magnetic Resonance Imaging (MRI). These patients exhibit an increased rate of brain atrophy (brain shrinkage) when compared to healthy subjects, which has been linked to impairment. However, due to the high variability in the image characteristics caused by scanner and center factors (e.g., manufacturer, magnetic field), this type of biomarker cannot be reliably estimated over periods of less than five years. Nevertheless, patients are often scanned in different machines over the course of their disease, making a quantitative follow-up of these patients very difficult and unreliable.

The main goal of the project was to investigate new approaches to overcome quantitative differences in MRI extracted measurements arising from multi-center and multi-scanner settings. We explored the relation between image-specific properties and brain volume measurements and used the findings to develop different types of approaches based on machine learning and image processing. We consider that methods which promote the generalization ability of the volume estimation models are the most promising for future research.

Contact

Koen Van Leemput
Groupleader, Professor
DTU Health Tech
+45 45 25 52 95