Lapgym: an open source framework for reinforcement learning in robot-assisted laparoscopic surgery
Recent advances in reinforcement learning (RL) have increased the promise of introduc ing cognitive assistance and automation to robot-assisted laparoscopic surgery (RALS). However, progress in algorithms and methods depends on the availability of standardized learning environments that represent sk...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article (Journal) |
| Language: | English |
| Published: |
12/23
|
| In: |
Journal of machine learning research
Year: 2023, Volume: 24, Pages: 1-43 |
| ISSN: | 1533-7928 |
| Online Access: | Verlag, kostenfrei, Volltext: https://dl.acm.org/doi/pdf/10.5555/3648699.3649067 Verlag, kostenfrei, Volltext: https://jmlr.org/papers/v24/23-0207.html 
|
| Author Notes: | Paul Maria Scheikl, Balázs Gyenes, Rayan Younis, Christoph Haas, Gerhard Neumann, Martin Wagner, Franziska Mathis-Ullrich |
| Summary: | Recent advances in reinforcement learning (RL) have increased the promise of introduc ing cognitive assistance and automation to robot-assisted laparoscopic surgery (RALS). However, progress in algorithms and methods depends on the availability of standardized learning environments that represent skills relevant to RALS. We present LapGym, a frame work for building RL environments for RALS that models the challenges posed by surgical tasks, and sofa env, a diverse suite of 12 environments. Motivated by surgical training, these environments are organized into 4 tracks: Spatial Reasoning, Deformable Object Manipulation & Grasping, Dissection, and Thread Manipulation. Each environment is highly parametrizable for increasing difficulty, resulting in a high performance ceiling for new algorithms. We use Proximal Policy Optimization (PPO) to establish a baseline for model-free RL algorithms, investigating the effect of several environment parameters on task difficulty. Finally, we show that many environments and parameter configurations reflect well-known, open problems in RL research, allowing researchers to continue explor ing these fundamental problems in a surgical context. We aim to provide a challenging, standard environment suite for further development of RL for RALS, ultimately helping to realize the full potential of cognitive surgical robotics. LapGym is publicly accessible through GitHub (https://github.com/ScheiklP/lap_gym |
|---|---|
| Item Description: | Online veröffentlicht: 12/23 Gesehen am 25.07.2024 |
| Physical Description: | Online Resource |
| ISSN: | 1533-7928 |