simonat2011 / RLCT-v4

source code of Learning to Scan: A Deep Reinforcement Learning Approach for Personalized Scanning in CT Imaging

RLCT-v3

python -u main.py --num-process 8 --save-path '../ctmodel/2pporbfan' --value-clip True --actor-lr 4e-4 --critic-lr 8e-4 --mini-batch 100 --base-noise 0.0 --lr-decay-rate 0.99 --reconstruct-alg SART --tau 0.005 --RL-alg PPO-RB --train-epoch 20 --use-grad-clip True --use-linear-lr-decay True