Volkan Tuncay MSC, Presenter: Nothing to Disclose

Wisnumurti Kristanto MSC, Abstract Co-Author: Nothing to Disclose

Peter M.a. Van Ooijen PhD, Abstract Co-Author: Nothing to Disclose

Matthys Oudkerk MD, PhD, Abstract Co-Author: Nothing to Disclose

To develop (semi)-automatic computer software tools which can be used in the segmentation of the aortic valve opening area and aortic annulus. This area information can be used for two purposes: The first purpose is to assist the physician for the diagnosis of the cardiac diseases which are caused by dysfunction of the aortic valves. The second purpose is to develop a non-invasive technique to determine the size of the implant. Currently, the size of the aortic annulus is measured by using a physical tool called a “valve sizer” to determine the required size of the implant.  

The software tools were developed with Matlab®. A combination of three different segmentation techniques was used for the semiautomatic segmentation, namely thresholding, region growing and Gradient Vector Flow (GVF) snake algorithm.  The process has two major steps. The first step is producing a binary image from the gray-scale dicom image by thresholding. The second step is using the binary image in the snake algorithm to segment the desired area. After the binary image was produced, the semi-automatic algorithm was started by manual placement of an arbitrary amount of seed points inside the aortic valve opening area to provide an initial contour. Afterwards, the initial contour was allowed to grow to match the aortic valve using a gradient vector flow snake algorithm. The programme was tested on the heavily calcified CT images of 21 patients. The manual segmentations of these 21 images were done by three observers blinded to each other’s results. These results were gold standard for our semi-automatic segmentation algorithm.  

The semi-automatic algorithm succeeded in segmentation of the aortic valve opening area in 17 out of 21 images and the area of the aortic annulus was able to be segmented automatically in all of the 21 images.

The programme can be used on good quality images however the program should be improved in order to be able to segment the aortic valve opening area on the lower quality images.   

This software can be used in the surgical planning to determine the size of the implant that will be used in the implantation.

Tuncay, V, Kristanto, W, Van Ooijen, P, Oudkerk, M, Software Tools to Determine the Aortic Valve Opening Area and the Aortic Annulus.  Radiological Society of North America 2010 Scientific Assembly and Annual Meeting, November 28 - December 3, 2010 ,Chicago IL. http://archive.rsna.org/2010/9006621.html