Fig. 1.
The procedure of acquiring breast thermograms [12].
Fig 2.
Flowchart of the presented approach.
Fig 3.
Fig. 4.
Results of pseudo-coloring algorithms for a normal thermogram of an individual: original gray-level image, and its pseudo-colorized images in the HSI color space under various conditions.
Fig. 5.
Breast thermograms clustering using FCM method.
(a) Pseudo-colorized image, (b) clustered image applying the FCM method, (c) extracted hottest region after removing the axilla and sternal and the final binary image.
Table 1.
The mean std. of chaotic features for all benign and malignant cases.
Table 2.
The mean std. of texture features of benign and malignant cases for all breast thermograms.
Table 3.
The design parameters for feature selection methods.
Table 4.
The final values of the cost function, the selected features number, and the implementation time for the introduced meta-heuristic algorithms.
Fig. 6.
The objective function amounts versus the selected feature number for the NSGA III method.
Table 5.
Further information regarding the designed classifiers.
Table 6.
The achieved outcomes regarding the statistical metrics of the designed classifiers.
Fig. 7.
The bar graph depicting the statistical metrics of the designed classifiers applying 10-fold cross-validation method.
Table 7.
Accuracy related to different feature domains on different classifications.
Fig. 8.
The graph bar of the accuracy comparison for three types of feature sets applying 10-fold cross-validation method.
Table 8.
A comparison between accuracies of the present method and the other published results used the same database (DMR-IR).