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

Table 5

Neural Network for breast image classification.

Reference	Descriptor	Image type	Number of images	Key findings
Rajakeerthana et al. [42]	GLCM, GLDM, SRDM, NGLCM, GLRM	Mammogram	322	The classifier achieved Accuracy.
Lessa and Marengoni [43]	Mean, Median, Standard Deviation, Skewness, Kurtosis, Entropy, Range	Thermographic	94	Achieved Sensitivity, Specificity, and Accuracy are 87.00%, 83.00%, and 85.00%, respectively.
Wan et al. [44]	ALBP BBLBP	OCM	46	Achieved Sensitivity and Specificity are 100% and 85.20%. respectively.
				ROC value obtained 0.959.
Chen et al. [40]	19 BI-RADS features have been used	Ultrasound	238	Chi squared method has been utilized for the feature selection.
				Achieved Accuracy, Sensitivity, and Specificity are 96.10%, 96.70%, and 95.70%, respectively.
de Lima et al. [45]	Total 416 features have been used	Mammogram	355	Multiresolution wavelet and Zernike moment have been utilized for the feature extraction.
Abirami et al. [46]	12 statistical measures such as Mean, Median, and Max have been utilized as the features	Mammogram	322	Wavelet transform has been utilized for the feature extraction.
				The achieved Accuracy, Sensitivity, and Specificity are 95.50%, 95.00%, and 96.00%, respectively.
El Atlas et al. [47]	13 morphological features have been utilized	Mammogram	410	Firstly the edge information has been utilized for the mass segmentation and then the morphological features were extracted.
				Achieved best Accuracy is 97.5%.