Skip to main content

Table 1 Detailed performance of univariate analysis for imaging feature with each modality and fusion method

From: PET/MR fusion texture analysis for the clinical outcome prediction in soft-tissue sarcoma

Class Feature number AUC value
No fusion-based features
 T1-weighted MR images 52 0.7196 ± 0.0340
 T2-weighted MR images 71 0.6985 ± 0.0228
 PET images 79 0.7254 ± 0.0366
Image-level fusion based features
 T1/PET Image Fusion (0.1) 85 0.7459 ± 0.0414
 T1/PET Image Fusion (0.2) 87 0.7462 ± 0.0504
 T1/PET Image Fusion (0.3) 78 0.7533 ± 0.0463
 T1/PET Image Fusion (0.4) 83 0.7523 ± 0.0420
 T1/PET Image Fusion (0.5) 77 0.7567 ± 0.0386
 T1/PET Image Fusion (0.6) 76 0.7619 ± 0.0448
 T1/PET Image Fusion (0.7) 75 0.7631 ± 0.0432
 T1/PET Image Fusion (0.8) 70 0.7407 ± 0.0338
 T1/PET Image Fusion (0.9) 73 0.7306 ± 0.0365
 T2/PET Image Fusion (0.1) 90 0.7411 ± 0.0442
 T2/PET Image Fusion (0.2) 83 0.7519 ± 0.0438
 T2/PET Image Fusion (0.3) 83 0.7503 ± 0.0420
 T2/PET Image Fusion (0.4) 75 0.7392 ± 0.0328
 T2/PET Image Fusion (0.5) 70 0.7215 ± 0.0283
 T2/PET Image Fusion (0.6) 81 0.7084 ± 0.0244
 T2/PET Image Fusion (0.7) 72 0.7041 ± 0.0226
 T2/PET Image Fusion (0.8) 66 0.6957 ± 0.0195
 T2/PET Image Fusion (0.9) 69 0.6885 ± 0.0167
Matrix-level fusion based features
 T1/PET Matrix Fusion 90 0.7216 ± 0.0355
 T2/PET Matrix Fusion 95 0.7441 ± 0.0464
Feature-level fusion based features
 T1/PET Feature Concatenation 131 0.7231 ± 0.0359
 T2/PET Feature Concatenation 150 0.7126 ± 0.0335
 T1/PET Feature Average 85 0.7249 ± 0.0318
 T2/PET Feature Average 95 0.7366 ± 0.0416
  1. The number in the parentheses indicated the MR weight