Lung volume estimation based on chest X-ray images

OData support
Dr. Horváth Gábor
Department of Measurement and Information Systems

The thesis is divided to four parts; each of these sections builds upon the results of the previous ones. The first part describes the problem, and sets the goals for the thesis. The second part describes a CT reconstruction algorithm (SART), with it's background theory and implementation. The role of reconstruction algorithms is to merge ordinary X-ray chest radiographs into a pseudo-3D voxel array. The implementation uses classic GPGPU approach with the Microsoft DirectX 11 API to achieve acceptable performance (for clinical use).

The third part describes a segmentation method, which will be capable of segmenting the lungs in the voxel array. It will use a Level Set Method based approach. One advantage of the Level Set Method approach, that it could be generalized to 3D volumes with the same algorithmic complexity as the 2D (slice-by-slice) approach.

The fourth part will describe the algorithm which is capable of measuring the lung volume based on the segmentation of the previous algorithm.


Please sign in to download the files of this thesis.