From: Mammographic texture and risk of breast cancer by tumor type and estrogen receptor status
Analysis groups | Texture features | Texture feature name | References |
---|---|---|---|
Gray-level histogram | Standard deviation | STD | [7, 22, 24–26] |
Skewness | Skewness | ||
Kurtosis | Kurtosis | ||
Balance | Balance | ||
Gray-level co-occurrence matrix (GLCM) | GLCM Energy | Energy | [24, 25, 27, 29] |
GLCM Entropy | Entropy | ||
GLCM Dissimilarity | Dissimilarity | ||
GLCM Contrast | Contrast | ||
GLCM Homogeneity | Homogeneity | ||
GLCM Correlation | Correlation | ||
GLCM Mean | GLCM Mean | ||
GLCM Variance | GLCM Variance | ||
Neighborhood gray-tone difference matrix (NGTDM) | NGTDM Coarseness | NGTDM Coarseness | [24, 28, 29] |
NGTDM Contrast | NGTDM Contrast | ||
NGTDM Complexity | Complexity | ||
NGTDM Strength | Strength | ||
NGTDM Busyness | Busyness | ||
Edge frequency analysis | Mean gradient | Mean_Gradient | [29] |
Fourier transform (FT) analysis, power spectrum | RMS (root mean square) | FT_RMS | [29] |
FMP (first moment of power spectrum) | FT_FMP | ||
SMP (second moment of power spectrum) | FT_SMP | ||
FD (fractal dimension) from power spectrum exponent | FT_FD | ||
Fractal analysis | Intercept of the plot of the standard deviation of the high frequency image as a function of the size the kernel | CD_Yint | [29–31] |
Continuous dimension (CD), slope and intercept | CD_Slope | ||
HZ_PROJ | HZ_PROJ | ||
FD of the standard deviation | FD_Sigma | ||
FD of image using thresholds from 5%-85% | FD_TH_5: FD_TH_85 | ||
FD of the surface of the breast considering the gray value representing the height | FD_CALDWELL | ||
FD, Minkowski method | FD_Minkowski |