Robust regression and outlier detection
Robust regression and outlier detection
Fast Approximate Energy Minimization via Graph Cuts
IEEE Transactions on Pattern Analysis and Machine Intelligence
MICCAI '01 Proceedings of the 4th International Conference on Medical Image Computing and Computer-Assisted Intervention
Automatic whole heart segmentation in static magnetic resonance image volumes
MICCAI'07 Proceedings of the 10th international conference on Medical image computing and computer-assisted intervention
STACOM'11 Proceedings of the Second international conference on Statistical Atlases and Computational Models of the Heart: imaging and modelling challenges
Hi-index | 0.00 |
Delayed-enhancement magnetic resonance imaging is an effective technique for imaging left atrial (LA) scars both pre- and post- radio-frequency ablation for the treatment of atrial fibrillation. Existing techniques for LA scar segmentation require expert manual interaction making them tedious and prone to high observer variability. In this paper, we propose a novel automatic segmentation algorithm for segmenting LA scar based on a probabilistic tissue intensity model. This is implemented as a Markov random field-based energy formulation and solved using graph-cuts. It was evaluated against an existing semi-automatic approach and expert manual segmentations using 9 patient data sets. Surface representations were used to compare the methods. The segmented LA scar was expressed as a percentage of the total LA surface. Statistical analysis showed that the novel algorithm was not significantly different to the manual method and that it compared more favorably with this than the semi-automatic approach.