Multi-atlas Propagation Whole Heart Segmentation from MRI and CTA Using a Local Normalised Correlation Coefficient Criterion
Accurate segmentation of the whole heart from 3D image sequences is an important step in the developement of clinical applications. As manual delineation is a tedious task that is prone to errors and dependant on the expertise of the observer, fully automated segmentation methods are highly desirable. In this work, we present a fully automated method for the segmentation of the whole heart and the great vessels from 3D images. The method is based on a muti-atlas propagation segmentation scheme, that has been proven to be succesful in brain segmentation. Based on a cross correlation metric, our method selects the best atlases for propagation allowing the refinement of the segmentation at each iteration of the propagation. We show that our method allows segmentation from multiple image modalities by validating it on computed tomography angiography (CTA) and magnetic resonance images (MRI). Our results are comparable to state-of-the-art methods on CTA and MRI with average Dice scores of 90.9% and 89.0% for the whole heart when evaluated on a 23 and 8 cases, respectively.
Unable to display preview. Download preview PDF.
- 5.Rikxoort, E.M., Išgum, I., Arzhaeva, Y., Staring, M., Klein, S., Viergever, M., Pluim, J.P.W., van Ginneken, B.: Adaptive local multi-atlas segmentation: Application to the heart and the caudate nucleus. Medical Image Analysis (2010)Google Scholar
- 7.Kirişli, H.A., Schaap, M., Klein, S., Papadopoulou, S.L., Bonardi, M., Chen, C.H., Weustink, A.C., Mollet, N.R.A., Vonken, E.P.A., van der Geest, R.J., van Walsum, T., Niessen, W.J.: Evaluation of a multi-atlas based method for segmentation of cardiac CTA data: a large-scale, multicenter, and multivendor study. Medical Physics 37(12), 6279–6292 (2010)CrossRefGoogle Scholar
- 12.Cardoso, M.J., Modat, M., Keihaninejad, S., Cash, D., Ourselin, S.: Multi-STEPS: multi-label similarity and truth estimation for propagated segmentations. In: IEEE Workshop on Mathematical Methods in Biomedical Image Analysis (MMBIA), pp. 153–158 (2012)Google Scholar
- 14.Leung, K.K., Barnes, J., Ridgway, G.R., Bartlett, J.W., Clarkson, M.J., Macdonald, K., Schu, N., Fox, N.C., Ourselin, S.: ADNI: Automated cross-sectional and longitudinal hippocampal volume measurement in mild cognitive impairment and Alzheimer’s disease. NeuroImage 51(4), 1345–1359 (2010)CrossRefGoogle Scholar