. 2020 Sep 3;138(10):1070–1077. doi: 10.1001/jamaophthalmol.2020.3269

Table 3. F1 Score of Algorithms for Retinal Diagnosis That Address Bias in Artificial Intelligence.

Algorithm^a	10 Samples	40 Samples	160 Samples	5120 Samples
RES_FT	0.5648	0.4865	0.4925	0.7291
RES_KNN	0.5536	0.5263	0.5863	0.5870
RES_SVM	0.6067	0.5340	0.5812	0.7011
RES_RF	0.5893	0.5962	0.6414	0.6739
DIM	0.6381	0.5846	0.7022^b	0.7360
DIM_KNN	0.5447	0.5674	0.5864	0.6082
DIM_SVM	0.6513^b	0.6086	0.6600	0.7446^b
DIM_RF	0.5660	0.6316^b	0.6498	0.6985

Abbreviations: DIM, Deep InfoMax; KNN, K-nearest neighbors; RES_FT, traditional fine-tuned ResNet algorithm; RF, random forest; SVM, support vector machine.

^{^a}

Algorithms include a traditional fine-tuned ResNet algorithm (denoted as RES_FT), RES_RF and RES_SVM (ResNet encoding fed into an RF or SVM classifier), Augmented Multiscale Deep InfoMax encoding¹¹ yielding local and global features, fed to a classifier, consisting of either ResNet (using only local features, and denoted as DIM), and 3 other classifiers using the global features of DIM and either DIM_SVM, or DIM_RF.

^{^b}

Best algorithm for number of samples used for training.