Table 3. Comparison of detection sensitivity of different lung nodule detection algorithms.
Algorithm | 0.125 | 0.25 | 0.5 | 1 | 2 | 4 | 8 | CPM |
---|---|---|---|---|---|---|---|---|
Xie [2019] (25) | 73.40 | 74.40 | 76.30 | 79.60 | 82.40 | 83.20 | 83.40 | 79.00 |
Zuo [2019] (26) | 67.20 | 69.40 | 71.40 | 73.90 | 76.60 | 78.70 | 82.20 | 74.20 |
You [2019] (27) | 62.40 | 64.30 | 67.70 | 72.80 | 80.00 | 85.80 | 90.00 | 74.70 |
Sun [2021] (28) | 46.7 | 60.2 | 73.0 | 81.2 | 87.7 | 92.0 | 93.1 | 76.20 |
Hong [2021] (29) | 60.05 | 70.18 | 80.01 | 84.39 | 89.99* | 92.16* | 94.42* | 81.60 |
Yang [2021] (30) | 56.40 | 67.40 | 75.70 | 78.20 | 83.50 | 87.60 | 91.20 | 77.20 |
Improved model | 69.83* | 80.00* | 82.36* | 84.42* | 84.79 | 85.23 | 86.16 | 81.83* |
*, the highest sensitivity of the detection at that FP rate. CPM, challenge performance metric; FP, false positive.