Table 3 Average performance of the Little W-Net model for each of the datasets in Table 2 over 5 training runs, with a confidence interval containing the true mean with probability \(p = 95\%\), under a normality assumption of the performances.
From: State-of-the-art retinal vessel segmentation with minimalistic models
DRIVE | CHASE-DB | HRF | ||||||
|---|---|---|---|---|---|---|---|---|
AUC | DICE | MCC | AUC | DICE | MCC | AUC | DICE | MCC |
\(98.10\pm 0.04\) | \(82.79\pm 0.11\) | \(80.24\pm 0.12\) | \(98.47\pm 0.02\) | \(81.69\pm 0.22\) | \(79.74\pm 0.23\) | \(98.25\pm 0.05\) | \(81.03\pm 0.08\) | \(79.09\pm 0.09\) |
