Table 1 Different preprocessing steps and their respective AUC score.
From: Automatic detection of pathological myopia using machine learning
Preprocessing step | AUC score |
|---|---|
Grayscale image, resize, shuffle, normalization | 0.9593 |
Grayscale, resize, shuffle, normalization, batch normalization | 0.9745 |
Red channel image, resize, shuffle, normalization | 0.9829 |
Red channel image, resize, shuffle, normalization, batch normalization | 0.9845 |
Grayscale image, histogram equalization, resize, shuffle, normalization, batch normalization | 0.9723 |
Grayscale image, adaptive histogram equalization (CLAHE), resize, shuffle, normalization, batch normalization | 0.9761 |
Red channel image, adaptive histogram equalization (CLAHE), resize, shuffle, normalization, batch normalization | 0.9707 |
