| Signal distribution |
25 |
Describe the pixel-wise variations of the ROI signal intensities. Particular emphasis was added in the smaller and larger quantiles. |
| Fractals |
25 |
A set of texture features based on signal distribution of the local fractal dimension. |
| Gray Level Co-Occurrence |
24 |
Haralick’s grey-level co-occurrence matrix texture features: energy, entropy, correlation, difference moment, inertia, cluster shade, cluster prominence, and Haralick correlation at 3 distances (1, 2, and 4 pixels). |
| Minkowski functionals |
40 |
Ten values (curvature energy, kurtosis, largest derivate, mean, half of maximum height, skewness, energy of the derivate, order of largest derivate, signal standard deviation, functional standard deviation) of the area, contour length, Euler number, and compactness of the image resulting from the binarization of the ROI at 40 uniformly spaced threshold values. |
| Wavelet decompositions |
400 |
Decomposition of the images using the Daubechies-4 mother wavelet in 2D at 4 multi-resolution levels. The resulting images were subject to the signal distribution and fractal feature extraction process (4x4x25). |
| Local standard deviation |
25 |
Signal distribution of the standard deviation from a 3x3 pixel image. |
