**4. Discussion**

and evaluating the model using the remaining subjects [46]. The final reported performance was obtained by applying the final model gathered on the training stage and evaluating it in

A total of 1796 mammograms were successfully segmented. The image sets of nine subjects had to be removed from the experiment due to problems with the registration process, six were from MS, two from CS, and one from HS. All the remaining subjects were included in the subsequent stages of the analysis. The 2150 extracted features were filtered by the correla-

**Table 1** shows the features that were selected for each model: the CS versus HS (*n* = 12), and the MS versus HS (*n* = 16). The former achieved an accuracy of 0.825 with an AUC of 0.882 and the latter an accuracy of 0.698 with an AUC of 0.807. **Figure 8** shows the ROC curves for both the models.

**CS versus HS MS versus HS**

Note: Features are grouped by dataset, symmetric features are denoted with:

 , *<sup>σ</sup>*Δavg <sup>=</sup> *<sup>σ</sup><sup>r</sup>* <sup>+</sup> *<sup>σ</sup>* \_\_\_\_\_*<sup>l</sup>* 2

*<sup>l</sup>*|, *H*Δs = |*Hr* − *Hl*|, L Δs = |*Hr* − *Hl*|, *σ*Δ*<sup>s</sup>* = |*Hr* − *Hl*|, *F*Δs = |*Hr* − *Hl*|.

 , *<sup>F</sup>*Δavg <sup>=</sup> *<sup>σ</sup><sup>r</sup>* <sup>+</sup> *<sup>σ</sup>* \_\_\_\_\_*<sup>l</sup>* 2 ,

 , L Δavg <sup>=</sup> *Lr* <sup>+</sup> *<sup>L</sup>* \_\_\_\_\_*<sup>l</sup>* 2

**# View Image Feature View Image Feature** CC *H*<sup>∆</sup> 27 CC *L*<sup>∆</sup> 40 CC *F*<sup>∆</sup> 13 CC *Ir* 29 CC *Ir* 29 CC *Ll* 40 CC *Hl* 29 CC *Fr* 6 CC *Hl* 6 MLO *Hl* 11 MLO *Il* 28 CC *L*∆avg 29 MLO *Hl* 11 CC *I*∆*<sup>s</sup>* 28 MLO *Hl* 21 CC *σ*∆*<sup>s</sup>* 38 CC *I*∆*<sup>s</sup>* 28 CC *F*∆avg 12 CC *σ*∆*<sup>s</sup>* 38 MLO *I*∆*<sup>s</sup>* 40 MLO *L*∆avg 27 MLO *I*∆s 28 CC *H*<sup>∆</sup> 27 MLO *I*∆s 29 MLO *H*∆s 31 MLO *H*∆s 7 MLO *L*∆avg 39 MLO *L*∆avg 27

the blind test set.

118 New Perspectives in Breast Imaging

tion process, removing 826 features.

**3. Results**

 *I* Δavg <sup>=</sup> *<sup>I</sup> <sup>r</sup>* <sup>+</sup> *<sup>I</sup>* \_\_\_\_*<sup>l</sup>* 2

I Δs = |*I <sup>r</sup>* − *I*

 , *<sup>H</sup>*Δavg <sup>=</sup> *Hr* \_\_\_\_\_ <sup>+</sup> *Hl* 2

**Table 1.** Features of the proposed models.

The proposed methodology is fully automated and does not require manual intervention as previous proposals [16, 17]. Although the approach is similar to others [22], we did not attempt to remove the pectoral muscle from the segmentation mask, since the presence of abnormal axillary lymph in this area is an indicator of occult breast carcinoma [47]. However, from the computational point of view, the feature extraction process may be affected if the region processed is not well focused [48].

The proposed registration process achieved a good performance having only 2.0% of the subjects that had to be discarded due to registration issues. This performance is remarkable when considering the amount of deformation undergoing in a mammography procedure. The B-spline deformation is an improvement over rigid or affine co-registration methods [33]. The advantage of the deformable registration has been recognized as a key element in breast analysis and has been successfully used in longitudinal studies [22]. Regarding digital subtraction, the differences in the X-ray projection, and image acquisition and digitizing artifacts may affect the detection of asymmetric patterns. Our results indicate that even in the presence of registration artifacts, the digital subtraction added information that was successfully incorporated during the feature selection process.

The B-Spline transformation algorithm, proposed for the bilateral mammogram registration presented, shows a clear improvement after the registration. Due to lack of temporal mammograms, temporal registration was not tested. Nevertheless, the methodology could be implemented in such task. However, the temporal registration should be re-optimized using a new set of parameters.

The enhanced images and texture maps enriched the feature set providing a four-fold increase in extracting features per patient, which were also incorporated in the final classification models. Regarding symmetry, the strategy of exploring bilateral symmetry has been explored by other researchers where a series of features (signal, texture, breast density, etc.) were computed from each mammogram and the absolute difference between both breasts was obtained to measure breast tissue asymmetry, and used it to predict the likelihood of developing cancer [19]. We extended this idea by registering the left and right images using a deformable transformation, which increased the number of features per patient by 25%.

This study shows that healthy subjects, subjects with calcifications, and subjects with masses can accurately be classified through models generated via mammography registration and a feature selection methodology. The analysis of the feature selection strategy demonstrated that even when using a different approach for the feature selection strategy, the proposed methodology achieved similar results as the previously presented ones. Therefore, we can say that the methodology is robust to the feature selection strategy.

The methodology demonstrated that the image subtraction of registered images generates information that aids in the identification of subjects with lesions, such as malignant masses and calcifications. The methodology also incorporated the use of feature-based asymmetry into the CADx system. The combination performance achieved has the potential to be used to queue cases with a high chance of malignant findings, or may have the practical use of triaging mammograms in developing countries where there is a deficiency of expert readers.
