Weakly supervised 3D classification of multi-disease chest CT scans using multi-resolution deep segmentation features via dual-stage CNN architecture (DenseVNet, 3D Residual U-Net).
deep-learning convolutional-neural-networks multi-resolution weakly-supervised-learning resnet50 3d-classification tensorflow-estimator ct-imaging segmentation-features densevnet lung-disease feature-aggregation
-
Updated
Oct 7, 2020 - Python