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

Table 14

SVM for breast image classification (Page-1).

Reference	Descriptor	Image type	Number of images	Key findings
Malik et al. [108]	Speed of sound; Attenuation image vector; Reflection image vector	QTUS	—	Glands, fat, skin, and connective tissue have been classified.; Both linear and nonlinear SVM classifier have been utilized.; Their experiment obtained 85.20% Accuracy.
Chang et al. [109]	Textural features such as; (i) Autocorrelation Coefficient; (ii) Autocovariance Coefficient	Ultrasound	250	Benign and malignant images have been classified.; Accuracy, Sensitivity, Specificity, positive predictive values, and negative predictive value are 85.60%, 95.45%, 77.86%, 77.21%, and 95.61%, respectively.
Akbay et al. [110]	52 features have been extracted	Mammogram	—	Microcalcification (MC) Classification Accuracy 94.00%
Levman et al. [111]	Relative Signal Intensities; Derivative of Signal Intensities; Relative Signal Intensities and their derivatives in one vector; (i) Maximum of signal intensity enhancement; (ii) time of maximum enhancement; (iii) time of maximum washout	MRI	76	Benign and malignant lesions are investigated.; Linear kernel, a polynomial kernel, and a radial basis function kernel utilized along with the SVM method for the breast image classification.
de Oliveira Martins et al. [112]	Ripley’s function	Mammogram	390	Benign and malignant image classification.; The achieved Accuracy, Sensitivity, and Specificity are 94.94%, 92.86%, and 93.33%, respectively.