Kavya Raju

A total hip replacement is one of the most common orthopaedic procedures in which the hip joint is replaced with an endoprosthesis. The correct positioning and orientation of the implant components are of major importance for a successful patient outcome after surgery. Hence, a new automated method(Mask RCNN) for measuring these parameters on anterior-posterior x-ray images is proposed and validated in this master thesis.

1. Literature review of state-of-the-art approaches 2. Generation of 3D point cloud data 3. Labeling and preparation of data 4. Generation of data from simulation (V-rep) 5. Implementation of deep learning / computer vision algorithm for semantic segmentation, object detection, pose estimation 6. Evaluation of experiments on real and simulated data

A malfunction of the mitral valve is the cause of many heart disorders and valvular diseases. Automatic segmentation of the mitral valve using CNN's has a huge impact on surgical outcomes. This project uses the U-Net architecture to identify important landmarks and segment the annulus and leaflets of the mitral valve using 4D ultrasound images.

1. Processing and Segmentation of Aneursym Datasets. 2. Hands on experience with MevisLab

Role as Developer 1. Worked as a Sitecore developer. 2. Coordinated with the team by designing web pages and web services. 3. On hand experience with Sitecore 4. Training was focused on SQL, C and C#

Grade : 2.0

Grade: 1.3