Prediction of Chemotherapy Response of Osteosarcoma Using Baseline 18F-FDG Textural Features Machine Learning Approaches with PCA
Table 2
Receiver operating characteristic curve analysis and univariate logistic regression analysis for evaluation of response to chemotherapy.
Variable
AUC
Sen (%)
Spe (%)
-value
SUVmax
Baseline
0.553
51.61
67.86
0.488
After NAC
0.839
61.29
92.86
<0.001
% change
0.863
93.55
71.43
<0.001
TLG
Baseline
0.538
45.16
82.14
0.626
After NAC
0.816
77.42
71.43
<0.001
% change
0.838
80.65
82.14
<0.001
MTV
Baseline
0.536
45.16
78.57
0.645
After NAC
0.764
96.77
46.43
<0.001
% change
0.838
80.65
82.14
<0.001
1st entropy
Baseline
0.538
22.58
92.86
0.616
After NAC
0.767
70.97
82.14
<0.001
% change
0.713
70.97
71.43
0.0018
GLCM entropy
Baseline
0.543
41.94
75
0.575
After NAC
0.775
70.97
82.14
<0.001
% change
0.71
67.74
75
0.0022
AUC, area under the curve; Sen, sensitivity; Spe, specificity; SUV, standardized uptake value; NAC, neoadjuvant chemotherapy; TLG, total lesion glycolysis; MTV, metabolic tumor volume.