Skip to main content

Advertisement

Table 2 Model performance

From: A whole slide image-based machine learning approach to predict ductal carcinoma in situ (DCIS) recurrence risk

Model and Clinicopathologic Variables 2x2 Performance Metrics
 Training CohortValidation Cohort
VariableRec. Status at 10 YearsMetricsRec. Status at 10 YearsMetrics
Model CensoredRecurredAcc: 0.87PPV: 0.65 CensoredRecurredAcc: 0.85PPV: 0.46
Low Risk1169Sn: 0.71NPV: 0.93Low Risk14413Sn: 0.50NPV: 0.92
High Risk1222Sp: 0.91OR: 23.6High Risk1513Sp: 0.91OR: 9.60
Necrosis CensoredRecurredAcc: 0.41PPV: 0.18 CensoredRecurredAcc: 0.26PPV: 0.13
No4713Sn: 0.58NPV: 0.78No286Sn: 0.77NPV: 0.82
Yes8118Sp: 0.37OR: 0.80Yes13120Sp: 0.18OR: 0.71
Size CensoredRecurredAcc: 0.50PPV: 0.15 CensoredRecurredAcc: 0.49PPV: 0.10
Below6820Sn: 0.33NPV: 0.77Below8318Sn: 0.31NPV: 0.82
Above5810Sp: 0.54OR: 0.59Above768Sp: 0.52OR: 0.49
Age CensoredRecurredAcc: 0.27PPV: 0.18 CensoredRecurredAcc: 0.24PPV: 0.15
Below197Sn: 0.77NPV: 0.73Below212Sn: 0.92NPV: 0.91
Above10924Sp: 0.15OR: 0.60Above13824Sp: 0.13OR: 1.83
Radiotherapy CensoredRecurredAcc: 0.64PPV: 0.19 CensoredRecurredAcc: 0.66PPV: 0.05
No9423Sn: 0.26NPV: 0.80No12124Sn: 0.08NPV: 0.83
Yes348Sp: 0.73OR: 0.96Yes382Sp: 0.76OR: 0.27
Grade CensoredRecurredAcc: 0.41PPV: 0.21*
I/II418Sn: 0.74NPV: 0.84
III8623Sp: 0.32OR: 1.37
  1. The 2 × 2 confusion matrix and performance metrics for the 8-feature model and common clinopathological variables in the training and validation cohorts. For each variable, the positive condition was recurrence within 10 years. A 2 × 2 matrix for grade in the validation cohort was omitted due to all patients belonging to grade III. Margin status was not shown for either cohort due to almost all patients having negative margins. The threshold used for patient age was 50 years, and the threshold for size was 2 cm