Research Article
Automated Feature Extraction in Brain Tumor by Magnetic Resonance Imaging Using Gaussian Mixture Models
Table 3
Performance metrics (%) based on the wavelets.
| Classifier | Metrics | T1-WI | T2-WI | FLAIR | db1 | coif1 | db1 | coif1 | db1 | coif1 |
| NB | Accuracy | 67.64 | 79.41 | 82.35 | 85.29 | 70.58 | 64.70 | False alarm | 17.64 | 8.82 | 8.82 | 8.88 | 8.82 | 14.70 | Missed detection | 14.70 | 8.8235 | 8.82 | 8.82 | 20.58 | 20.58 |
| SVM | Accuracy | 50 | 76.47 | 85.29 | 88.23 | 58.82 | 55.88 | False alarm | 20.58 | 14.70 | 2.94 | 0 | 8.82 | 8.82 | Missed detection | 29.41 | 8.82 | 11.76 | 11.76 | 32.35 | 35.29 |
| PNN | Accuracy | 70.58 | 79.41 | 82.35 | 88.23 | 70.588 | 58.82 | False alarm | 20.58 | 20.58 | 11.76 | 5.88 | 20.58 | 29.41 | Missed detection | 8.82 | 0 | 5.88 | 5.88 | 8.82 | 11.76 |
|
|
db1 and coif1 are the first order of Daubechies and Coiflet wavelet, respectively.
|