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

Table 7

Neural Network for breast image classification.

Reference	Descriptor	Image type	Number of images	Key findings
Chen et al. [61]	Variance Contrast of Wavelet Coefficient	Ultrasound	242	The achieved ROC curve 0.9396 0.0183
	Autocorrelation of Wavelet Coefficient
Silva et al. [62]	22 different morphological features such as convexity and lobulation have been utilized	Ultrasound	—	The best obtained Accuracy and ROC curve are 96.98% and 0.98, respectively
Saritas [63]	Age of patient, mass shape, mass border, Mass density, BIRADS	Mammogram	—	Disease prediction rate is 90.5%
				Neural Network utilized 5 neurons in input layers and one hidden layer.
López-Meléndez et al. [64]	Area, perimeter, etc. have been utilized	Mammogram	322	The achieved Sensitivity and Specificity are 96.29% and 99.00%, respectively.