Table 2 AI-based structure-function in fundus-controlled perimetry (FCP).
From: AI-based structure-function correlation in age-related macular degeneration
Author/Ref. | Title | Disease | Technique | Prediction | Outcome measure | Outcome |
|---|---|---|---|---|---|---|
Kihara et al. [44] | Estimating retinal sensitivity using optical coherence tomography with deep-learning algorithms in macular telangiectasia type 2 | Macular telangiectasia type 2 | 1. Million-variable deep-learning model (convolutional neural network) 2. Letnet modell 3. Linear regression | • Fundus-controlled perimetry (FCP) (a) Mesopic | • Mean absolute error (MAE) | 1. 3.36 dB 2. 3.66 dB 3. 4.51 dB |
von der Emde and Pfau et al. [25] | Artificial intelligence for morphology-based function prediction in neovascular age-related macular degeneration | Neovascular age-related macular degeneration | • Random forest (LOO-CV) • Two scenarios 1. Without FCP 2. With FCP data | • Fundus-controlled perimetry (FCP): (a) Mesopic (b) DA Cyan (c) DA Red | • Mean absolute error (MAE) • Root mean squared error (RMSE) | • Scenario 1 • 3.94 dB (a) 4.89 dB (b) 4.05 dB (c) • Scenario 2 • 2.8 dB (a) 3.7 dB (b) 2.85 dB (c) |
Pfau et al. [24] | Determinants of cone- and rod function in geographic atrophy: AI-based structure-function correlation | geographic atrophy secondary to age-related macular degeneration | • Random forest (LOO-CV) • Two scenarios 1. Without FCP 2. With FCP data | • Fundus-controlled perimetry (FCP): (a) mesopic (b) DA Cyan (c) DA Red | • Mean absolute error (MAE) • Root mean squared error (RMSE) | • Scenario 1 • 4.64 dB (a) 4.89 dB (b) 4.4 dB (c) • Scenario 2 • 2.89 dB (a) 2.86 dB (b) 2.77 dB (c) |
Sumaroka et al. [46] | Treatment potential for macular cone vision in leber congenital amaurosis due to CEP290 or NPHP5 Mutations: predictions from artificial intelligence | Retinitis pigmentosa (RP; training) Congenital Amaurosis (LCA; prediction) | • Random forest (LOO-CV) • Two models 1. Thickness and eccentricity 2. Reflectivity | • Dark-adapted static perimetry (a) DA cyan (b) DA red | • 95th Percentile limits of agreement (LOA) | RP 1. 9.6 dB (a) 8.8 dB (b) 2. 11.9 dB (a) 10.8 dB (b) LCA 1. 4.6–17.6 dB |
Sumaroka et al. [45] | Foveal therapy in blue cone monochromacy: predictions of visual potential from artificial intelligence | Blue Cone Monochromacy | • Random forest (a) Layer thickness (b) reflectivity | • Fundus-controlled Perimetry (a) Mesopic | • Root mean squared error (RMSE) | (a) 2.91 dB (b) 2.69 dB |
Heß et al. [47] | Mesopic and scotopic light sensitivity and its microstructural correlates in pseudoxanthoma elasticum | Pseudoxanthoma Elasticum | • Random forest (LOO-CV) | • Fundus-controlled perimetry: (a) Mesopic (b) DA Cyan (c) DA Red | • Mean absolute error (MAE) | (a) 4.91 dB (b) 5.44 dB (c) 4.99 dB |
