Home | deutsch  | Legals | Data Protection | Sitemap | KIT

Thesis or Internship Development and evaluation of a refraction correction in a SAFTbased image reconstruction for 3D Ultrasound Computer Tomography

Thesis or Internship Development and evaluation of a refraction correction in a SAFTbased image reconstruction for 3D Ultrasound Computer Tomography
job offer: links:

Development and evaluation of a refraction correction in a SAFTbased image reconstruction for 3D Ultrasound Computer Tomography

Foto 1 IPE 04-17
Background At Karlsruhe Institute of Technology a new medical imaging method for early breast cancer diagnosis is currently being developed. The 3D Ultrasound Computer Tomography system allows reconstructing 3D reflectivity images using synthetic aperture focusing technique (SAFT). The resulting volume represents the echotexture of human tissue which is expected to be related to different types of tissue. Though a sound speed and attenuation correction have recently been included in the algorithm to enhance the contrast and resolution of images, refraction is currently neglected.

Aim
The aim of this work would be to integrate a refraction correction based on a Fast Marching Algorithm to determine the fastest travel time according to Fermat’s principle. For fast computation the method needs to be integrated into the existing CUDA implementation of SAFT.

Task description
(topics will be adapted to available time frame)
 • Getting familiar with the currently used SAFT algorithm and extension as well as with the CUDA implementation.
 • Prototype implementation and evaluation of a refraction correction in MATLAB, generation of simulated data to evaluate the effectiveness of the refraction correction.
 • Integration of the refraction correction in the CUDA SAFT implementation
 • Evaluation of the image reconstruction with simulated and real data, performance evaluation

Qualifications
 • Basic knowledge about medical imaging and medical image processing
 • Programming knowledge required, predominantly in MATLAB and C respectively CUDA.

Contact
Torsten Hopp,  torsten hoppNci1∂kit edu