The Similarity Cloud Model: A novel and efficient hippocampus segmentation technique
Fredy Edgar Carranza Athó, Agma J. M. Traina, Caetano Traina, Paula Rejane Beserra Diniz · 2011
This work presents a new segmentation model called Similarity Cloud Model (SCM) based on hippocampus feature extraction. The segmentation process is divided in two main operations: localization by similarity and cloud adjustment. The first process uses the cloud to localize the most probable position of the hippocampus in a target volume. Segmentation is completed by a reformulation of the cloud to correct the final labeling, based on a new computation of arc-weights. This method has been tested in an entire dataset of 235 MRI combining healthy and epileptic patients. Results indicate superior quality segmentation in comparison with similar graph and bayesian-based models.