Table 16 Performance analysis of statistical significance testing with FCE images.
From: Feature fusion context attention gate UNet for detection of polycystic ovary syndrome
Model | Raw accuracy (%) | FCE accuracy (%) | \(\overline{{{\text{Diff}}}}\) | \(\overline{{{\text{Var}}}}\) | SD | t-value | DOF | p-value | 95% CI (%) |
|---|---|---|---|---|---|---|---|---|---|
DenseNet | 68.27 | 69.43 | 1.16 | 0.16 | 0.40 | 2.90 | 4 | 0.045 | 0.03–2.29 |
VGG | 71.23 | 72.62 | 1.39 | 0.16 | 0.40 | 3.47 | 4 | 0.025 | 0.31–2.47 |
AlexNet | 72.91 | 73.51 | 0.60 | 0.16 | 0.40 | 1.50 | 4 | 0.20 | −0.44–1.64 |
ResNet | 76.25 | 77.44 | 1.19 | 0.16 | 0.40 | 2.98 | 4 | 0.042 | 0.12–2.26 |
U-Net | 78.64 | 79.32 | 0.68 | 0.16 | 0.40 | 1.70 | 4 | 0.16 | −0.30–1.66 |
Attention U-Net | 82.36 | 83.78 | 1.42 | 0.16 | 0.40 | 3.55 | 4 | 0.023 | 0.36–2.48 |
FCAU-Net | 90.51 | 99.89 | 9.38 | 0.25 | 0.50 | 42.00 | 4 | < 0.001 | 8.50–10.26 |
