Involvement of Machine Learning for Breast Cancer Image Classification: A Survey

Table 13

Logic Based.

Reference	Descriptor	Image type	Number of images	Key findings
Angayarkanni and Kamal [102]	GLCM	Mammogram	322	The Achieved Sensitivity and Accuracy are 93.40% and 99.50%, respectively.
Wang et al. [103]	Horizontal Weighted Sum; Vertical Weighted Sum; Diagonal Weighted Sum; Grid Weighted Sum.	Mammogram	322	Surrounding Region Dependence Method (SRDM) utilized for region detection.; Achieved True Positive Rate 90.00% and False Positive Rate 88.80%.
Tambasco Bruno et al. [104]	Curvelet Transform; LBP	Mammogram; Histopathological	—	ANOVA method utilized for feature prioritization.; When they use RF algorithm on Mammogram (DDSM) dataset, obtained Accuracy and ROC are 79.00% and 0.89.
Muramatsu et al. [105]	Radial Local Ternary Pattern (RLTP)	Mammogram	376	Textural features have been extracted from the regions of interest (ROIs) using RLTP.; They claimed that the RLTP feature provides better performance than the rotation invariant patterns.
Dong et al. [106]	NRL margin gradient; Gray-level histogram; Pixel value fluctuation	Mammogram	—	Chain code utilized for extraction of regions of interest (ROIs).; Rough-Set method utilized to enhance the ROIs.; Their achieved ROC value is 0.947 and obtained Matthews Correlation (MCC) is 0.8652.
Piantadosi et al. [107]	Local Binary Pattern-Three Orthogonal Projections (LBP-TOP)	Mammogram	—	Their achieved Accuracy, Sensitivity, and Specificity values are 84.60%, 80.00%, and 90.90%.