Skip to content

Latest commit

 

History

History
20 lines (10 loc) · 921 Bytes

README.md

File metadata and controls

20 lines (10 loc) · 921 Bytes

Optimising radiation therapy treatment schedules with deep reinforcement learning

  • The model folder contains the code of the Python implementation of the simulation.

  • The model_cpp folder contains the code of the C++ implementation of the simulation.

  • The nnets folder contains the neural networks trained with different algorithms and reward functions as explained in the manuscript.

  • The training_logs folder contains the training log files of the four agents described in the manuscript inside zip archives.

  • The eval folder contains the performance evaluations of the different agents.

  • The tmp folder contains the images created during the evaluation of the agents.

  • The misc folder contains files that couldn't be classified in the preceding folders.

  • main.py is used to train an agent.

  • use_network.py evaluates a neural network using the performance indicators described in the manuscript.