Here we need to update the data processing pipeline, by modifying dataset.py(SliceDataset). Instead of slicing the Ct scans into 2D images, process them as 3D volumes.
Furthermore, we need to change all 2D layers to 3D counterparts in the model architecture(Unet.py). Moreover we need to adapt the loss functions and metrics to match the 3D data
Here we need to update the data processing pipeline, by modifying dataset.py(SliceDataset). Instead of slicing the Ct scans into 2D images, process them as 3D volumes. Furthermore, we need to change all 2D layers to 3D counterparts in the model architecture(Unet.py). Moreover we need to adapt the loss functions and metrics to match the 3D data