Table 2.
The results of ROC analysis of potentially optimal texture parameters in PET and CT images for breast lymphoma versus breast carcinoma.
| Variable | Lymphoma | Carcinoma | AUC | P value | ||
|---|---|---|---|---|---|---|
| Median | Range | Median | Range | |||
| Radiomic parameters on PET | ||||||
| HISTO_entropy_log10 | 1.36 | 0.64–1.76 | 0.97 | 0.32–1.66 | 0.740 | 0.001 |
| SHAPE_volume(# vx) | 183.00 | 22.00–6334.00 | 71.00 | 11.00–4241.00 | 0.725 | 0.002 |
| GLCM_entropy_log10 | 2.04 | 1.00–3.19 | 1.54 | 0.59–3.03 | 0.730 | 0.001 |
| GLRLM_HGRE | 287.95 | 17.14–1919.15 | 75.66 | 9.96–862.28 | 0.752 | <0.001 |
| NGLDM_Contrast | 0.33 | 0.04–2.32 | 0.16 | 0.00–0.59 | 0.712 | 0.003 |
| GLZLM_HGZE | 266.62 | 17.53–1593.64 | 83.77 | 9.67–832.22 | 0.754 | <0.001 |
|
| ||||||
| Radiomic parameters on CT | ||||||
| HISTO_Kurtosis | 5.02 | 1.90–217.17 | 2.84 | 1.49–97.34 | 0.660 | 0.024 |
| SHAPE_volume (mL) | 11.38 | 1.07–403.07 | 4.40 | 0.73–272.41 | 0.728 | 0.001 |
| GLCM_Homogeneity | 0.49 | 0.08–0.60 | 0.44 | 0.33–0.65 | 0.628 | 0.072 |
| GLRLM_RLNU | 977.06 | 135.49–24346.21 | 507.35 | 80.31–19527.32 | 0.685 | 0.009 |
| NGLDM_Busyness | 0.53 | 0.04–6.04 | 0.38 | 0.13–2.77 | 0.602 | 0.152 |
| GLZLM_LZHGE | 14709543.63 | 77721.12–1603305720.99 | 1385779.22 | 44958.07–435410392.46 | 0.681 | 0.011 |
Abbreviations: HISTO, histogram; GLCM, gray-level co-occurrence matrix; GLRLM, gray-level run-length matrix; HGRE, high gray-level run emphasis; GLZLM, gray-level zone-length matrix; HGZE, high gray-level zone emphasis; RLNU, run-length nonuniformity; LZHGE, long-zone high gray-level emphasis.