Figure 1 | Scientific Reports

Figure 1

From: Classifying central serous chorioretinopathy subtypes with a deep neural network using optical coherence tomography images: a cross-sectional study

Figure 1

Performance comparison of models on spectral domain-optical coherence tomography images. Test accuracies of each fold during five-fold cross-validation. Our proposed model achieved the highest accuracy of 73.5% when using the second fold as the test set. An illustration of the confusion matrix of classification results in five-fold cross-validation. While the x-axis denotes the predicted class, the y-axis denotes the ground truth. Therefore, diagonal values from the top left to bottom right denote the correct predictions made by our proposed model. Our proposed model displays robust performance on every fold.

Back to article page