KOPS - The Institutional Repository of the University of Konstanz

Rekonstruktion der Koronar-Anatomie mittels digitaler Bildverarbeitung von Echokardiogrammen

Rekonstruktion der Koronar-Anatomie mittels digitaler Bildverarbeitung von Echokardiogrammen

Cite This

Files in this item

Checksum: MD5:a499fa0c9b8023d25f85820b61b7c325

GRAICHEN, Uwe, 2005. Rekonstruktion der Koronar-Anatomie mittels digitaler Bildverarbeitung von Echokardiogrammen [Dissertation]. Konstanz: University of Konstanz

@phdthesis{Graichen2005Rekon-6451, title={Rekonstruktion der Koronar-Anatomie mittels digitaler Bildverarbeitung von Echokardiogrammen}, year={2005}, author={Graichen, Uwe}, address={Konstanz}, school={Universität Konstanz} }

application/pdf Rekonstruktion der Koronar-Anatomie mittels digitaler Bildverarbeitung von Echokardiogrammen 2005 One of the main causes of death in the western countries are coronary heart<br />diseases. Presently, the X-ray angiography is the gold standard for the<br />examination of coronary arteries. This invasive, X-ray-based procedure<br />requires the infrastructure of a cardiac catheterization lab. Ultrasonic<br />examinations are getting more and more important as an imaging technique in<br />cardiography. In contrast to the X-ray angiography, patient and physician are<br />not exposed to degrading radiation. Further, ultrasound devices are commonly<br />used already and inexpensive in comparison to other imaging systems. The goal<br />of this project was to develop an image processing system, which can detect<br />and quantify coronary arteries in 3D-ultrasound data and thus can supplement<br />or replace the X-ray angiography examination. The procedure consists of<br />several steps: 1st the interpolation of volumedata without artefacts, 2nd the<br />analysis of the local structure of the datasets, 3rd tracking the run of the<br />vessels and 4th the segmentation of the vessel lumen.<br /><br />3D ultrasound datasets are interpolated out of a series of 2D ultrasound<br />images, which are recorded using a transesophageal transducer. The recording<br />is triggered by the ECG and needs a few minutes. Because of the long recording<br />time, the motion of the heart and of the patient during the recording the<br />image series contain movement artefacts. This artefacts can be reduced by<br />registration adjacent images in the series before interpolation volume data.<br />The registration is done using a rigid, correlation based method. Rotation and<br />scaling parameter are determined from the Fourier-Mellin invariant descriptors<br />of the images. The translation parameters are calculated out of the by<br />rotation and scaling corrected images. Due the special structure of the<br />ultrasound images a windowing function adapted to the ultrasound cone is used.<br />With this modified volume interpolation method movement artefacts can be<br />significant reduced.<br /><br />The analysis of the local structure is used to detect line-like structures in<br />3D ultrasound datasets. The structural analysis is done using the local<br />differential structure of the datasets. Particularly suitable for this purpose<br />are Hessians which are calculated for every voxel of the Gaussian-smoothed 3D<br />dataset. The Eigenvectors of the Hessian establish an orthogonal system. Two<br />Eigenvectors are pointing in the directions with the minimal or maximal<br />absolut value of the second derivation respectively. The third is orthogonal<br />to both. The Eigenvalues connected to the Eigenvectors spezify the value of<br />the second derivation. From the Eigenvalue a similarity measure with Gaussian<br />lines is calculated. By chosing the the standard deviation and by examination<br />the sign of the Eigenvalues it is possible to search selective of a line-like<br />with a specific diameter and progression of contrast.<br /><br />The run of the vessels is described by centerlines. The result of the<br />structural analysis step is 3D dataset with similarity measures with the<br />select line-like structure. The centerlines are calculated in such a way that<br />the run of the centerline is close to voxels with large similarity measures.<br />This done by a modified thinning method. First a threshold is applied. Voxels<br />with a similarity measure smaller than this threshold are set to zero. From<br />the remaining voxels a erasure list is generated. The erasure list contains<br />coordinates of the voxels and is sorted by the similarity measures in<br />ascending order. During the thinning the erasure list is traversed and it is<br />checked to delete the voxel without changing the topology and without removing<br />end points of the skeleton. The resulting resulting skeleton runs close to<br />voxels with large similarity measures.<br /><br />Along the centerlines of the vessels the lumen of the vessels is segmented.<br />For every voxel on the centerline a cross section is calculated. The normal of<br />the cross section is the Eigenvector of the Hessian which is connected to the<br />Eigenvalue with the smallest absolute value. On the cross sections the lumen<br />of the vessels is segmented using a modified Snake-method. The potential<br />function, which describes the energy of the image, is calculated from the<br />magnitude of the gradients and additionally from the similarity measure with<br />line-like structures. The potential forces which affect to the Snake are<br />calculated from this potential function using the gradient-vector-flow<br />method. By this modified Snake method is robust approach for segmenting the<br />lumen of vessels. deu 2011-03-24T16:12:48Z terms-of-use Graichen, Uwe Reconstruction of coronary anatomy by means of digital image processing of echocardiographs Graichen, Uwe 2011-03-24T16:12:48Z

Downloads since Oct 1, 2014 (Information about access statistics)

dis.pdf 361

This item appears in the following Collection(s)

Search KOPS


My Account