Automatic Detection of Calcified Coronary Plaques in Computed Tomography Data Sets
MICCAI '08 Proceedings of the 11th international conference on Medical Image Computing and Computer-Assisted Intervention - Part I
MICCAI '08 Proceedings of the 11th international conference on Medical Image Computing and Computer-Assisted Intervention - Part I
Fast, quality, segmentation of large volumes – isoperimetric distance trees
ECCV'06 Proceedings of the 9th European conference on Computer Vision - Volume Part III
| Hi-index | 0.00 |
Cardiac calcium scoring is an important step for the diagnosis of coronary heart diseases. Therefore, non-contrast enhanced cardiac computed tomography has been established as the de facto standard method for clinical risk assessment and contrast enhanced computed tomography has proven to be a reliable, non-invasive alternative to traditional coronary angiography. However, calcium scores determined on contrast enhanced data cannot be easily related to the scores determined on non-contrast enhanced data. Hence, an increased number of studies are being performed in order to evaluate the clinical value of calcium scoring on contrast enhanced computed tomography coronary angiography images. While the clinical results with respect to the diagnostic value are promising, the high image contrast variability caused by the contrast agent leads to an increased manual effort in order to accurately segment calcified lesions in the data. Moreover, manual calcium scoring on contrast enhanced computed tomography scans is subject to strong intra- and inter-observer variability.In this paper, we present a novel approach to the fully automatic segmentation and quantification of calcified lesions in coronary computed tomography angiograms. The method includes a robust threshold determination algorithm based on a histogram calculated from an automatically generated vessel tree. Thereby, lesions can be accurately segmented and calcium scores can be determined without user interaction. Validation against manual scores obtained by a radiologist showed a very high correlation, which demonstrates the clinical value of the presented method.