Table 5 Comparation on the performances of model 1, model 2 and model 3 in identifying LVI in the training and validation cohorts.

From: Predictive value of MRI-based deep learning model for lymphovascular invasion status in node-negative invasive breast cancer

Models

AUC

95% CI

Sensitivity

Specificity

Accuracy

P value

Lower

Upper

Model 1

< 0.001*

 Training set

0.910

0.871

0.948

0.891

0.702

0.792

 

 Validation set

0.896

0.806

0.964

0.806

0.814

0.810

 

Model 2

< 0.001*

 Training set

0.845

0.793

0.897

0.778

0.767

0.772

 

 Validation set

0.720

0.611

0.830

0.762

0.667

0.713

 

Model 3

 Training set

0.927

0.893

0.955

0.991

0.380

0.671

 

 Validation set

0.835

0.719

0.929

0.935

0.333

0.655

 
  1. AUC area under the ROC curves, CI confidence interval.
  2. *DeLong test for comparing with model 3 in each validation set.