Diagnosis of Diabetic Retinopathy through Retinal Fundus Images and 3D Convolutional Neural Networks with Limited Number of Samples
Table 7
Class-wise GM values for the multiclass classification tasks without augmentation, with random weak Gaussian blurred augmentation, with random shift augmentation, and with combined augmentations.
Method
GM (mild)
GM (moderate)
GM (no)
GM (proliferate)
GM (severe)
Without augmentation
0.5297
0.4162
0.5131
0.7558
0.5056
With random weak Gaussian blurred augmentation
0.4972
0.4523
0.4557
0.6307
0.4815
With random shifted augmentation
0.4635
0.3925
0.4513
0.6817
0.4889
With combined augmentations
0.5001
0.4639
0.4754
0.6891
0.476
Test set validated on the model trained without augmentation
