Master thesis works
Master Thesis Works
We are looking for students with interest in medical image processing, cardiac physiology, or MR physics. Experience in Matlab programming and relevant courses are valuable merits.
Ideally, we would find projects where the student has a strong interest and that there is a strong need to solve the challenge by the research group, therefore the best is to find mutually interesting projects. Therefore, we typically do not advertise specific projects on this page, but rather find individually adjusted projects. Master thesis works can be performed in conjunction with the spin-off company Medviso AB. For more information about if we have possibility suitable projects and the capacity to supervise students, please contact Einar Heiberg. The following projects are examples of areas where we see possible projects.
Image analytics of large study cohorts
Using machine learning techniques, we want to automate the analysis process so that it will be possible to perform processing of large study cohorts, such as the SCAPIS population consisting of 30 000 subjects.
New 3D printing techniques for clinical utility
We work closely together with neurosurgeons, hand surgeons, and maxillofacial surgeons and develop new applications including moulds for bone cement implants, design of cutting and drilling guides, and patient-specific implants. We have access to a multitude of 3D printers at the department.
Image reconstruction for cardiac fetal images
Acquiring cardiac fetal MR images is very challenging and the image reconstruction is critical for the success. Image reconstruction can be performed using a technique called compressed sensing, or machine learning approaches.
Image processing
There is a multitude of image processing tasks that is needed at the department such as robust image registration to tissue characterization.
Previous Master Thesis projects
Below are some examples of previous master thesis projects that can serve as inspiration.
Jesper Öberg, Segmentation of the right ventricle in cardiac CT images
Andreas Söderlund, Patient-specific blood vessel modeling on data from MR scans of children with congenital heart disease
Anja Lemic, Semi-Automatic Segmentation of Coronary Arteries in CT Images
Gabriella Gleisner, Semi-Automatic Segmentation of the Right Ventricle in MR Images
John Heerfordt Sjökvist, Enhanced Display of Mitoses in Hematoxylin-Eosin Digital Pathology Images
Anton Holm, Deep Learning Algorithms for Cardiac Image Classification and Landmark Detection
Emma Törner, Adaptation of existing cardiovascular simulation model to cardiac pumping physiology
Mattias Nilsson, Myocardial Segmentation in MR images using Convolutional Neural Networks