Table 1.
Recently published approaches.
| Article | Type | Year | Category | Dataset Size/Availability | Metrics |
|---|---|---|---|---|---|
| Kuo, et al. [3] | S | 2014 | Deformable models | 98/private | DSC |
| Liu, et al. [4] | S | 2010 | Level set-based | 79/private | TP, FP, SI |
| Xian, et al. [5] | F | 2015 | Graph-based | 184/private | TPR, FPR, SI, HD, MD |
| Shao, et al. [6] | F | 2015 | Graph-based | 450/private | TPR, FPR, SI |
| Huang, et al. [7] | S | 2014 | Graph-based | 20/private | ARE, TPVF, FPVF, FNVF |
| Xian, et al. [8] | F | 2014 | Graph-based | 131/private | SI, FPR, AHE |
| Gao, et al. [9] | S | 2012 | Normalized cut | 100/private | TP, FP, SI, HD, MD |
| Hao, et al. [10] | F | 2012 | CRF + DPM | 480/private | JI |
| Moon, et al. [11] | S | 2014 | Fuzzy C-means | 148/private | Sensitivity and FP |
| Shan, et al. [12] | F | 2012 | Neutrosophic L-mean | 122/private | TPR, FPR, FNR, SI, HD, and MD |
| Hao, et al. [13] | F | 2012 | Hierarchical SVM + CRF | 261/private | JI |
| Jiang, et al. [14] | S | 2012 | Adaboost + SVM | 112/private | Mean overlap ratio |
| Shan, et al. [15] | F | 2012 | Feedforward neural network | 60/private | TPR, FPR, FNR, HD, MD |
| Pons, et al. [16] | S | 2014 | SVM + DPM | 163/private | Sensitivity, ROC area |
| Yang, et al. [17] | S | 2012 | Naive Bayes classifier | 33/private | FP |
| Torbati, et al. [18] | S | 2014 | Feedforward Neural network | 30/private | JI |
| Huang, et al. [19] | F | 2020 | Deep CNNs | 325/private + 562/public | TPR, FPR, JI, DSC, AER, AHE, AME |
| Huang, et al. [20] | F | 2018 | Deep CNNs + CRF | 325/private | TPR, FPR, IoU |
| Shareef, et al. [21] | F | 2020 | Deep CNNs | 725/public | TPR, FPR, JI, DSC, AER, AHE, AME |
| Liu, et al. [22] | S | 2012 | Cellular automata | 205/private | TPR, FPR, FNR, SI |
| Gómez, et al. [23] | S | 2010 | Watershed | 50/private | Overlap ratio, NRV and PD |
F: fully automatic, S: semi-automatic, SVM: support vector machine, CRF: conditional random field, DPM: deformable part model, CNNs: convolutional neural networks, TP: true positive, FP: false positive, SI: similarity index, HD: Hausdorff distance, MD: mean distance, DSC: Dice similarity, JI: Jaccard index, ROC: Receiver operating characteristic, ARE: average radial error, TPVF: true positive volume fraction, FPVF: false positive volume fraction, FNVF: false negative volume fraction, NRV: normalized residual value, PD: proportional distance, TPR: true positive ratio, FPR: false positive ratio, FNR: false negative ration, and IoU: Intersection over union.