Next: List of Tables
Up: UNIVERSITAT DE GIRONA Department
Previous: Contents
Contents
- Different viewpoints of the same breast.
- Shapes and margins of a mass.
- Examples of masses.
- HRIMAC's architecture.
- Scheme of our proposal.
- Breast profile segmentation using the fast breast
segmentation.
- Breast profile segmentation using the contour-based
approach.
- Right and left mammograms of a woman.
- CC and MLO views of the same breast.
- A tumour-like template.
- FROC analysis of the algorithms over the set of
mammograms.
- Grey-level histograms of a breast.
- Mammogram division strategies for feature
extraction.
- Quantitatively reviewed strategies for breast
segmentation.
- Template models classification.
- Four RoIs and their manual segmentations.
- Eigenmasses.
- Constructed templates for a posterior pattern
matching.
- Profile of the constructed templates.
- Potential images of three mammograms.
- Suspicious regions found using the Bayesian pattern
matching approach.
- FROC analysis.
- FROC analysis per lesion size.
- Mean
values for the MIAS database.
- Mean Kappa values for the MIAS database.
- Mean
values for the DDSM database.
- Mean Kappa values for the DDSM database.
- Eigenrois.
- FROC analysis of the proposals.
- FROC analysis comparison.
- FROC analysis comparison per lesion size.
- Influence of the training database in the FROC
analysis.
- FROC analysis comparison using different databases.
- Proposed algorithms FROC comparison.
- Comparison of the proposal with other algorithms.
- Mean
value for the DDSM database.
- Mean Kappa values for the DDSM database.
- Breast tissue information.
- Histogram of a typical mammogram.
- Sequence of the breast profile segmentation.
- Examples of the breast profile segmentation.
- Contour growing scheme.
- Cost functions for robust candidate selection.
- Two examples of the breast-skin line segmentation.
- ROC example.
Arnau Oliver
2008-06-17