Glaucoma is a multifactorial neurodegenerative illness requiring early diagnosis and strict monitoring of the disease progression. Artificial intelligence algorithms can extract various optic disc features and automatically detect glaucoma from fundus photographs.
Author | Year | N. of Images | Structure | SEN | SPEC | ACC | AUC | ||||
---|---|---|---|---|---|---|---|---|---|---|---|
Kolar et al. [12][16] | 2008 | 30 | |||||||||
OCT Fundus | Thompson et al. [43][ | FD | 93.80% | ||||||||
47 | ] | 2019 | 1. Global BMO-MRW prediction | Nayak et al. [13][17] | 2009 | 61 | Morphological | 100% | 80% | 90% | |
Bock et al. [14][18] | 2010 | 575 | Glaucoma Risk Index | 73% | 85% | 80% | |||||
Acharya et al. [15][19] | 2011 | 60 | SVM | 91% | |||||||
Dua et al. [16][20] | 2012 | 60 | DWT | ||||||||
0.914 | |||||||||||
Author | Year | Outcomes Mesures | Architecture | SEN | SPEC | ACC | AUC | ||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
ResNet34 | |||||||||||||||||||||
Asaoka et al. [65] | 0.945 | ||||||||||||||||||||
[69] | 2016 | Pre-perimetric VFs vs. VFs in healthy eyes | FNN | 0.926 | 2. Yes glaucoma vs. No glaucoma | ||||||||||||||||
Kucur et al. [71][75] | 2018 | Early glaucomatous VF loss vs. no glaucoma | CNN with Voronoi representation | ||||||||||||||||||
Medeiros et al. [49][53] | 2019 | 1. RNFL thickness prediction | ResNet34 | 80% | 83.7% | ||||||||||||||||
Li et al. [8] | 0.944 | ||||||||||||||||||||
[12] | 2018 | Glaucomatous VF loss vs. no glaucoma | VGG I 5 | 93% | 83% | 88% | 0.966 | 2. Glaucoma vs. Suspect/healthy | 93.3% | ||||||||||||
Li et al. [72] | |||||||||||||||||||||
[76] | 2018 | Glaucoma vs. Healthy | VGG | 93% | 3% | 0.966 | |||||||||||||||
Jammal et al. [48][52] | 2020 | RNFL prediction | ResNet34 | ||||||||||||||||||
Berchuck et al. [73][ | 0.801 | ||||||||||||||||||||
77] | 2019 | Rates of VF progression compared to SAP MD; Prediction of future VF compared to point-wise regression predictions | Deep variational autoencoder | Mookiah et al. [17][21] | 2012 | 60 | DWT, HOS | 86.7% | Lee et al. 93.3% | [58]93.3% | [62] | 2021 | RFNL prediction | M2M | |||||||
Wen et al. [74][78] | 2019 | HFA points and Mean Deviation | CascadeNet-5 | Noronha et al. [18][22] | 2014 | Medeiros et al. 272 | [Higher order cumulant features | 100% | 92% | 92.6% | 50][54] | 2021 | Detection of RFNL thinning from fundus photos | CNN | |||||||
Kazemian et al. [70][74] | 2018 | Forecasting visual field progression | Kalman Filtering Forecasting | ||||||||||||||||||
Acharya et al. [19][23] | 2015 | 510 | Gabor transform | 89.7% | 96.2% | 93.1% | |||||||||||||||
OCT 2D | Asaoka et al. [51][55] | 2019 | Early POAG vs. no POAG | Novel CNN | 80% | 83.3% | |||||||||||||||
Garcia et al. [69] | [73 | 0.937 | |||||||||||||||||||
] | 2019 | Forecasting visual field progression | Kalman Filtering Forecasting | Isaac et al. [20][24] | Muhammad et al. [59][63]2015 | 67 | Cropped input image after segmentation | 100% | 90% | 94.1% | |||||||||||
2017 | Early glaucoma vs. health/suspected eyes | CNN + transfer learning | |||||||||||||||||||
DeRoos et al. [68] | 93.1% | 0.97 | |||||||||||||||||||
[72] | 2021 | Forecasting visual field progression | Kalman Filtering Forecasting | Raja et al. [21][25] | 2015 | 158 | Hybrid PSO | 97.5% | Lee et al. [60][98.3% | 98.2% | |||||||||||
64] | 2020 | GON vs. No GON | CNN (NASNet) | 94.7% | 100% | 0.990 | Singh et al. [22][26] | 2016 | |||||||||||||
Devalla et al. [61][65] | 63 | Wavelet feature extraction | 100% | 90.9% | 94.7% | ||||||||||||||||
2018 | Glaucoma vs. normal | Digital stain of RNFL | 92% | 99% | 94% | Acharya et al. [23][27] | 2017 | 702 | kNN (K = 2) Glaucoma Risk index | 96.2% | 93.7% | 95.7% | |||||||||
Wang et al. [52][56] | 2020 | Glaucoma vs. no glaucoma | CNN + transfer learning | 0.979 | Maheshwari et al. [24][28] | 2017 | 488 | Variational mode decomposition | 93.6% | 95.9% | 94.7% | ||||||||||
Thompson et al. [47][51 | |||||||||||||||||||||
] | 2020 | POAG vs. no glaucoma | ResNet34 | 95% | 81% | 0.96 | Raghavendra et al. [25][29] | 2017 | 1000 | RT, MCT, GIST | 97.80% | 95.8% | 97% | ||||||||
Pre-perimetric vs. no glaucoma | 95% | 70% | 0.92 | Ting et al. [3][7] | 2017 | 494,661 | VGGNet | 96.4% | 87.2% | ||||||||||||
Glaucoma with any VF loss (perimetric) vs. no glaucoma | 0.942 | ||||||||||||||||||||
95% | 80% | Kausu et al. [26][30] | 2018 | 86 | Wavelet feature extraction, Morphological | 98% | 97.1% | 97.7% | |||||||||||||
0.97 | |||||||||||||||||||||
Mild VF loss vs. no glaucoma | 95% | 85% | 0.92 | Koh et al. [27][31] | 2018 | ||||||||||||||||
2220 | Pyramid histogram of visual words and Fisher vector | 96.73% | 96.9% | 96.7% | |||||||||||||||||
Moderate VF loss vs. no glaucoma | 95% | 93% | 0.99 | Soltani et al. [28][32] | 2018 | 104 | Randomized Hough transform | 97.8% | 94.8% | 96.1% | |||||||||||
Li et al. [8][12] | 2018 | 48,116 | Inception-v3 | 95.6% | 92% | 92% | 0.986 | ||||||||||||||
Fu et al. [29][33] | 2018 | 8109 | Disc-aware ensemble network (DENet) | 85% | 84% | 84% | 0.918 | ||||||||||||||
Raghavendra et al. [25][29] | 2018 | 1426 | Eighteen-layer CNN | 98% | 98.30% | 98% | |||||||||||||||
Christopher et al. [30][34] | 2018 | 14,822 | VGG6, Inception-v3, ResNet50 | 84–92% | 83–93% | 0.91–0.97 | |||||||||||||||
Chai et al. [31][35] | 2018 | 2000 | MB-NN | 92.33% | 90.9% | 91.5% | |||||||||||||||
Severe VF loss vs. no glaucoma | 95% | 98% | 0.99 | ||||||||||||||||||
Mariottoni et al. [62][66] | 2020 | Global RNFL thickness value | ResNet34 | ||||||||||||||||||
OCT 3D | Ran et al. [54][58] | 2019 | Yes GON vs. No GON | CNN (NASNet) | 89% | 96% | 91% | 0.969 | |||||||||||||
78–90% | 86% | 86% | 0.893 | ||||||||||||||||||
Maetschke et al. [53][57] | 2019 | POAG vs. no POAG | Feature-agnostic CNN | 0.94 | |||||||||||||||||
0.92 | Ahn et al. [32][36] | 2018 | 1542 | Inception-v3 Custom 3-layer CNN |
84.5% 87.9% |
||||||||||||||||
Russakoff et al. [55][59] | 0.93 | 2020 | Referable glaucoma vs. non-referable glaucoma | gNet3D-CNN | 0.94 | ||||||||||||||||
0.88 | Shibata et al. [33][37] | 2018 | 3132 | ResNet-18 | |||||||||||||||||
AS-OCT | Fu et al. [56] | [60] | 2019 | Open angle vs. Angle closure | 0.965 | ||||||||||||||||
VGG-16 + transfer learning | 90% | 92% | 0.96 | Mohamed et al. [34][38] | 2019 | 166 | Simple Linear Iterative Clustering (SLIC) | 97.6% | 92.3% | 98.6% | |||||||||||
Fu et al. [63][67] | 2019 | Open angle vs. Angle closure | CNN | 0.9619 | Bajwa et al. [35][39] | 2019 | 780 | R-CNN | 71.2% | 0.874 | |||||||||||
Xu et al. [57][61] | 2019 | 1. Open angle vs. angle closure | CNN (ResNet18) + transfer learning | 0.928 | Liu et al. [36][40] | 2019 | 241,032 | ResNet (local validation) | 96.2% | 97.7% | 0.996 | ||||||||||
2. Yes/PACD vs. no PACD | Al-Aswad et al. [37][41] | 2019 | 110 | ResNet-50 | 83.7% | 88.2% | 0.926 | ||||||||||||||
0.964 | |||||||||||||||||||||
Hao et al. [64][68] | 2019 | Open angle vs. Narrowed Angle vs. Angle closure | MSRCNN | Asaoka et al. [38][42] | 2019 | 3132 | ResNet-34 | 0.965 | |||||||||||||
ResNet-34 without augmentation | 0.905 | ||||||||||||||||||||
VGGI I | 0.955 | ||||||||||||||||||||
VGGI 6 | 0.964 | ||||||||||||||||||||
Inception-v3 | 0.957 | ||||||||||||||||||||
Kim et al. [39][43] | 2019 | 1903 | Inception-V4 | 92% | 98% | 93% | 0.99 | ||||||||||||||
Orlando et al. [40][44] | 2019 | 1200 | Refuge Data Set | 85% | 97.6% | 0.982 | |||||||||||||||
Phene et al. [41][45] | 2019 | 86,618 | Inception-v3 | 80% | 90.2% | 0.945 | |||||||||||||||
Rogers et al. [42][46] | 2019 | 94 | ResNet-50 | 80.9% | 86.2% | 83.7% | 0.871 | ||||||||||||||
Thompson et al. [43][47] | 2019 | 9282 | ResNet-34 | 0.945 | |||||||||||||||||
Hemelings et al. [11][15] | 2020 | 8433 | ResNet-50 | 99% | 93% | 0.996 | |||||||||||||||
Zhao et al. [44][48] | 2020 | 421 | MFPPNet | 0.90 | |||||||||||||||||
Li et al. [45][49] | 2020 | 26,585 | ResNet101 | 96% | 93% | 94.1% | 0.992 |
Author | Year | Outcome Measures | Arch | SEN | SPEC | ACC | AUC |
---|