Research Article
Automated Segmentation and Object Classification of CT Images: Application to In Vivo Molecular Imaging of Avian Embryos
Table 1
Classification accuracy of the SVM after 10-fold cross-validation on the training data from Group 1 eggs.
| Classes | True positive rate | False positive rate | Precision | Recall | F-measure | ROC-area |
| Humerus | 0.941 | 0.003 | 0.842 | 0.941 | 0.889 | 0.999 | Radius | 0.929 | 0.001 | 0.929 | 0.929 | 0.929 | 0.996 | Ulna | 0.918 | 0.001 | 0.938 | 0.918 | 0.928 | 0.979 | Carpometacarpus | 0.865 | 0.002 | 0.900 | 0.865 | 0.882 | 0.988 | Femur | 1.000 | 0.001 | 0.962 | 1.000 | 0.981 | 1.000 | Tibiotarsus | 0.925 | 0.000 | 0.980 | 0.925 | 0.951 | 0.999 | Tarsometatarsus | 0.941 | 0.001 | 0.960 | 0.941 | 0.950 | 0.991 | Not of interest (NOI) | 0.994 | 0.043 | 0.994 | 0.994 | 0.994 | 0.996 | Weighted avg. | 0.986 | 0.038 | 0.987 | 0.986 | 0.986 | 0.996 |
|
|