Table 2 Accuracy of ablation study in 10-fold cross-validation experiments
From: A novel fusion algorithm for benign-malignant lung nodule classification on CT images
Model | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | Mean |
|---|---|---|---|---|---|---|---|---|---|---|---|
\(CNN_1\) | 89.71% | 92.71% | 88.54% | 87.36% | 87.50% | 88.24% | 95.83% | 85.12% | 91.37% | 89.67% | 89.61% |
\(CNN_1+R\) | 84.50% | 91.66% | 89.58% | 87.36% | 88.54% | 83.48% | 90.63% | 83.78% | 88.99% | 89.66% | 87.82% |
\(CNN_1+G\) | 92.71% | 92.71% | 93.75% | 89.44% | 92.71% | 91.07% | 95.83% | 93.45% | 94.79% | 89.18% | 92.56% |
\(CNN_1+R+G\) | 92.71% | 91.67% | 93.75% | 89.45% | 92.71% | 91.07% | 95.83% | 93.45% | 94.79% | 89.66% | 92.51% |
\(CNN_2\) | 91.67% | 91.25% | 91.25% | 88.58% | 90.00% | 89.46% | 91.17% | 91.25% | 93.75% | 89.42% | 90.78% |
\(CNN_2+R\) | 88.54% | 92.50% | 91.25% | 88.58% | 91.25% | 92.14% | 95.00% | 91.25% | 91.07% | 90.67% | 91.23% |
\(CNN_2+G\) | 92.71% | 92.50% | 95.00% | 87.42% | 90.00% | 94.64% | 95.00% | 92.32% | 92.50% | 91.63% | 92.37% |
\(CNN_2+R+G\) | 91.67% | 93.75% | 96.25% | 88.67% | 91.25% | 92.32% | 97.50% | 93.75% | 92.50% | 91.63% | 92.93% |
\(CNN_3\) | 91.67% | 86.25% | 91.25% | 88.58% | 90.00% | 89.46% | 91.17% | 91.25% | 93.75% | 89.42% | 90.28% |
\(CNN_3+R\) | 91.67% | 90.00% | 90.00% | 92.33% | 90.00% | 85.89% | 92.50% | 88.57% | 90.00% | 91.92% | 90.29% |
\(CNN_3+G\) | 91.67% | 92.50% | 93.75% | 86.17% | 88.75% | 92.14% | 95.00% | 93.75% | 92.50% | 90.38% | 91.66% |
\(CNN_3+R+G\) | 92.71% | 92.50% | 95.00% | 88.67% | 90.00% | 90.89% | 95.00% | 92.50% | 92.50% | 91.63% | 92.14% |
\(CNN_4\) | 92.86% | 89.29% | 94.05% | 89.29% | 85.90% | 88.33% | 86.25% | 87.50% | 90.83% | 91.25% | 89.56% |
\(CNN_4+R\) | 89.77% | 92.50% | 91.25% | 88.57% | 87.50% | 87.50% | 90.00% | 88.75% | 93.75% | 92.50% | 90.21% |
\(CNN_4+G\) | 93.19% | 93.75% | 95.00% | 87.14% | 88.75% | 93.33% | 96.25% | 93.75% | 93.75% | 91.25% | 92.62% |
\(CNN_4+R+G\) | 93.19% | 93.75% | 95.00% | 87.14% | 90.00% | 94.58% | 97.50% | 93.75% | 92.50% | 92.50% | 92.99% |
\(CNN_5\) | 89.77% | 90.00% | 92.50% | 91.07% | 86.25% | 89.58% | 92.32% | 93.75% | 92.08% | 90.00% | 90.73% |
\(CNN_5+R\) | 89.77% | 90.00% | 95.00% | 88.57% | 88.57% | 91.25% | 93.75% | 95.00% | 91.25% | 90.00% | 91.32% |
\(CNN_5+G\) | 92.05% | 92.50% | 96.25% | 88.57% | 88.75% | 92.08% | 96.25% | 95.00% | 92.50% | 91.25% | 92.52% |
\(CNN_5+R+G\) | 93.19% | 96.25% | 95.00% | 88.39% | 88.75% | 93.75% | 96.25% | 95.00% | 92.50% | 91.25% | 93.03% |
\(CNN_s\) | 93.19% | 92.50% | 90.00% | 87.14% | 88.75% | 92.50% | 93.75% | 93.75% | 90.00% | 95.00% | 91.66% |
\(CNN_s+R\) | 93.19% | 92.5% | 90.00% | 87.14% | 90.00% | 93.75% | 95.00% | 93.75% | 91.25% | 93.75% | 92.03% |
RGD | 93.83% | 95.00% | 95.00% | 88.61% | 90.91% | 93.59% | 96.20% | 94.87% | 92.31% | 92.21% | 93.25% |