Figure 1 | Scientific Reports

Figure 1

From: Deep learning-based histopathological segmentation for whole slide images of colorectal cancer in a compressed domain

Figure 1

Comparison of Dice for each class between the result of standard input in spatial domain (Tile size: 512 by 512, ×20 magnification, standard), the result of small ROI input in spatial domain (Tile size: 256 by 256, ×20 magnification, Small ROI), the result of low resolution input in spatial domain (Tile size: 512 by 512, ×10 magnification, low resolution), the result of weighted average ensemble (WAE) result for wavelet sub-band after grayscale conversion (GRAY-DWT (WAE)), the result of wavelet weighted ensemble (WWE) result for wavelet sub-band after grayscale conversion (GRAY-DWT (WWE)), the result of weighted average ensemble (WAE) result for wavelet sub-band after principle component analysis (PCA) (PCA-DWT (WAE)), and the result of wavelet weighted ensemble (WWE) result for wavelet sub-band after principal component analysis (PCA) (PCA-DWT (WWE)).

Back to article page