Development and Validation of a Magnetic Resonance Imaging-Based Machine Learning Model for TMJ Pathologies

Table 3

Radiomic features selected for quantifying the heterogeneity differences.

Radiomic group	Associated filter	No. of features ()	Radiomic features
First-order statistics	None	18	Energy, total energy, entropy, minimum, 10 percentile, 90 percentile, maximum, mean, median, interquartile range, range, mean absolute deviation, robust mean absolute deviation, root mean square, standard deviation, skewness, kurtosis, variance
Shape	None	8	Volume, surface area, surface volume ratio, spherical disproportion, maximum 3D diameter, maximum 2D diameter column, maximum 2D diameter row, elongation
Texture features	GLCM	15	Autocorrelation, average intensity, cluster prominence, cluster shade, cluster tendency, contrast, difference average, difference entropy, difference variance, dissimilarity, entropy, sum average, sum entropy, sum variance, sum squares
Texture features	GLSZM	8	Large-area emphasis, gray-level nonuniformity, size zone nonuniformity, gray-level variance, zone entropy, high-gray-level zone emphasis, small-area high-gray-level emphasis, large-area high-gray-level emphasis
Texture features	GLRLM	7	Gray-level nonuniformity, run length nonuniformity, gray-level variance, run entropy, high-gray-level run emphasis, short-run high-gray-level emphasis, long-run high-level emphasis

Label: GLCM = gray-level cooccurrence matrix; GLSZM = gray-level size zone matrix; GLRLM = gray-level run length matrix.