Skip to main content

Table 4 Comparison of different versions for causal CNN architecture

From: Interpretable CNN for ischemic stroke subtype classification with active model adaptation

Model

Accuracy

AUC

Recall

Precision

F1-score

Number of parameters

CNN-V1

0.5578

0.6557

0.5578

0.6012

0.4948

11077

CNN-V2

0.5682

0.6505

0.5682

0.5912

0.4683

8997

CNN-V3

0.5652

0.6474

0.5652

0.6081

0.4973

11717

CNN-V4

0.6020

0.6757

0.6020

0.6213

0.5141

12997

  1. The bold values are to highlight our results