TY - GEN
T1 - Collaborative Multi-agent Reinforcement Learning for Landmark Localization Using Continuous Action Space
AU - Kasseroller, Klemens
AU - Thaler, Franz
AU - Payer, Christian
AU - Štern, Darko
N1 - Publisher Copyright:
© 2021, Springer Nature Switzerland AG.
PY - 2021
Y1 - 2021
N2 - We propose a reinforcement learning (RL) based approach for anatomical landmark localization in medical images, where the agent can move in arbitrary directions with a variable step size. Using a continuous action space reduces the average number of steps required to locate a landmark by more than 30 times compared to localization using discrete actions. Our approach outperforms a state-of-the-art RL method based on a discrete action space and is inline with state-of-the-art supervised regression based methods. Furthermore, we extend our approach to a multi-agent setting, where we allow collaboration between agents to enable learning of the landmarks’ spatial configuration. The results of the multi-agent RL based approach show that the position of occluded landmarks can be successfully estimated based on the relative position predicted for the visible landmarks.
AB - We propose a reinforcement learning (RL) based approach for anatomical landmark localization in medical images, where the agent can move in arbitrary directions with a variable step size. Using a continuous action space reduces the average number of steps required to locate a landmark by more than 30 times compared to localization using discrete actions. Our approach outperforms a state-of-the-art RL method based on a discrete action space and is inline with state-of-the-art supervised regression based methods. Furthermore, we extend our approach to a multi-agent setting, where we allow collaboration between agents to enable learning of the landmarks’ spatial configuration. The results of the multi-agent RL based approach show that the position of occluded landmarks can be successfully estimated based on the relative position predicted for the visible landmarks.
KW - Collaborative multi-agent
KW - Continuous action space
KW - Landmark localization
KW - Reinforcement learning
UR - http://www.scopus.com/inward/record.url?scp=85111455725&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-78191-0_59
DO - 10.1007/978-3-030-78191-0_59
M3 - Conference paper
AN - SCOPUS:85111455725
SN - 9783030781903
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 767
EP - 778
BT - Information Processing in Medical Imaging - 27th International Conference, IPMI 2021, Proceedings
A2 - Feragen, Aasa
A2 - Sommer, Stefan
A2 - Schnabel, Julia
A2 - Nielsen, Mads
PB - Springer Science and Business Media Deutschland GmbH
T2 - 27th International Conference on Information Processing in Medical Imaging, IPMI 2021
Y2 - 28 June 2021 through 30 June 2021
ER -