|
Decision tree model improves accuracy of the zAvatar-test. a Decision tree model takes into account tumor stage, zAvatar-metastatic potential and zAvatar-apoptosis fold change (“Apoptosis_FC”). b The “Patient Response” (Progression/No-Progression) was the dependent variable, while the “Stage Tumor”, “Apoptosis_FC” and presence of “Metastasis” were the independent variables. Parent node = 5 cases; child node = 5 cases; overall percentage of correlation = 91%. c New confusion matrix according to the decision tree displays the number of patients with actual and predicted responses in zAvatars. Stage II/III sensitive (S) patients refers to patients’ whose zAvatars have no metastasis or patients’ whose zAvatars present metastasis but the induction of apoptosis is >1.47. Stage IV sensitive (S) patients refers to patients’ whose zAvatars fold induction of apoptosis is >1.185. Conversely, stage II/III zAvatars with presence of metastasis and fold induction of apoptosis ≤1.47, along stage IV zAvatars whose fold induction of apoptosis ≤1.185 are classified as resistant (R). d New values for sensitivity, specificity, PPV and NPV for the zAvatar-test taking into account the tree decision model. Source data are provided as a Source Data file.
|
