Table 2 Evaluation results of the external datasets.

From: Automated detection of intracranial aneurysms using skeleton-based 3D patches, semantic segmentation, and auxiliary classification for overcoming data imbalance in brain TOF-MRA

Model

 

Acc

Specificity

PPV

NPV

Sensitivity

DSC

3D U-Net

 

0.772

0.641

0.651

0.754

0.772

0.226 ± 0.301

3D U-Net with auxiliary loss

1:1

0.860

0.703

0.727

0.817

0.834

0.559 ± 0.209

1:2

0.883

0.856

0.857

0.872

0.879

0.682 ± 0.114

1:3

0.874

0.841

0.848

0.851

0.853

0.753 ± 0.106

1:4

0.872

0.822

0.846

0.857

0.862

0.750 ± 0.194

1:5

0.873

0.824

0.834

0.842

0.858

0.748 ± 0.173

nnU-Net

0.791

0.738

0.748

0.759

0.768

0.623 ± 0.198

 
  1. For comparisons between groups, t-tests were performed based on a 1:2 ratio, with p-values less than 0.05 denoted by ** and p-values less than 0.5 denoted by *to indicate statistical significance. Results with p-values less than 0.05 are considered statistically significant. Accuracy (Acc); predictive value (PPV); negative predictive value (NPV); dice similarity coefficients (DSC); three-dimensional (3D).
Back to article page