Background

Osteoarthritis (OA) of the knee is a degenerative disease resulting in reduced knee function and pain. It can affect all structures of the knee including the knee cap (patella) and its cartilage. To understand the disease state through symptoms like bone deformations and cartilage defects precise automatic segmentations of larger cohorts are needed.


Thesis Objective

The objective of this thesis is to implement a full segmentation pipeline for patellar bone and cartilage in 3D MRI data based on Convolutional Neural Networks (CNNs) and Statistical Shape Models (SSMs). Similar techniques are already in use in our workgroup for segmentation of different knee structures and thus can serve as suitable starting point.

A detailed evaluation of segmentation accuracy for the pipeline has to be carried out.

In addition, the segmentation results can be used to perform correlation analysis between bone geometry/cartilage volume and OA severity as rated by physicians.


Pre-requisites

1. Experience in Python (and C++) programming
                - Experience with CNNs and SSMs would be a huge bonus

2. Knowledge of TCL and/or Bash
                - Experience in 'gluing' different program steps together would be helpful

We Offer

Insight into medical image data processing in combination with deep learning techniques in a stimulating research environment.

We offer co-supervision for the thesis and close collaboration within the research group over the entire duration of the project.

Automated Limb Alignment Determination using Convolutional Neural Networks
MA 01/18
Classifying Lesions of Knee Menisci using nD Views
MA 03/17