Student projects

 

MSc projects

Currently the following MSc projects are available:

• Combining relative assessments for melanoma classification (11/15/2018)
• Meta-learning for medical image segmentation (9/3/2018)
• Weakly supervised learning in medical imaging (various projects) (9/3/2018)
• Liver Cancer Recurrence Prediction (6/21/2018)
• Real-time Multimodal Image Registration (6/21/2018)
• Surgical Workflow Analysis (6/21/2018)
• Combining Optics and Acoustics For Real-time Guidance during Cancer Surgery (6/21/2018)

• • Prostate Cancer Cryoablation Tracking (with Philips Research Eindhoven)
  • Image registration of volumetric models (with Philips Research Eindhoven)
  • Pipeline for image based numerical simulations (with Philips Research Eindhoven)
  • A prediction model of the effect of delineation uncertainties on dose in radiotherapy (with Catharina Hospital Eindhoven)
  • Obtaining the actual delivered dose for image-guided online adaptive radiotherapy (with Catharina Hospital Eindhoven)
  • Optimization of breast cancer patient treatment plans in radiotherapy (with Catharina Hospital Eindhoven)
  • Various MSc projects at Maastro Clinic Maastricht: examples 1 and examples 2
  • Various projects at PIE Medical Imaging, Maastricht: Projects PIE Medical
  • Automatic 3D quantitative maps of knee cartilage (with UMC Utrecht)
  • Deep learning for analysis of stroke in infants (with UMC Utrecht)

Here are some examples of past Master Students’ Projects:

• Retinal structure recognition in images reconstructed from A-Scans of an OCT-probe
• X-ray scatter correction using sinusoidal modulation
• Neuromelanin imaging using magnetization transfer MRI at 7T
• Mapping the homunculus in sensorimotor cortex using fMRI
• Study on the excitability of the brain and its relation to functional MRI connectivity in patients with focal epilepsy
• Radiomic analysis of CT imaging features with clinical outcome
• Visualization of variability in radiotherapy dose planning 
• Supervised artery/vein classification in retinal images via local and contextual feature analysis
• Automatic 2D-3D registration of abdominal images during endovascular interventions
• The evaluation of a multi atlas multi organ segmentation for automatic lymph node detection in pelvic MR lymphography
• Software infrastructure for Retina Check : developing multiple software components for the Retina Check screening project 
• Automatic detection of vascular bifurcations and crossovers in retinal images
• Segmentation of the facial nerve in the temporal bone in multi-slice CT images : a semi-automatic approach with a parametric deformable model

BSc projects

This page has examples of past Bachelor End Projects, so you can get an idea of what projects are possible in the Medical Image Analysis group. Each student will have an individual project, but we try to cluster a number of them on a similar topic. In this way students can support each other to some extent. Obviously, you will also have a researcher from the group as supervisor.

Past projects:

• Validation of vessel quantification in fundus images
• Wisdom of the crowds for analysis of medical images
• Automatic assessment of skin capillary density
• Detecting diabetic retinopathy in retinal images using deep neural networks
• Functional MRI to diagnose and treat cognitive problems in epilepsy
• Automatic analysis of brain MRI
• Correction of blocking artifacts in histological images
• Segmentation of acini in breast terminal duct lobular units

Past projects in image acquisition (MRI) at UMCU

• Evaluation of T2 mapping with stimulated echo compensation
• Experimental evaluation of a “sonic needle”
• Reproducibility of cerebral aqueduct flow measurements

 

More details about specific possible projects (including external projects at UMCU) will be available later.  If you think you would like to do your project in our group, you can register your choice with Marielle Bosboom AND by filling in this form before December 5th 2018. We unfortunately cannot guarantee that you will get the project of your choice.

If you have any questions about doing a BEP on Medical Imaging, you can contact Josien Pluim, j.pluim@tue.nl.