How to get the lung-only DRR and bone-only DRR from CT images ?

1. Use rule-based methods (Hounsfield unit) to generate lung and bone masks from CT images.

2. Use these masks to training segmentation model, and follow Step A to Step D.

Step A: Use madel to get lung mask from CT image.

Step B: Clean all information, only save body contour.

Step C: Use lung mask to segmentation lung image from CT image.

Step D: Mix lung image and body image.

3. Repeat Step A to Step D for all CT images.

4. Sum all segmentation images.

5.The GIF show the process of sum images.