Panacea or Alchemy? ——Benefits and Risks of Robot Learning in Medical Applications
The Zoom meeting link for remote attendees: Meeting link
October 27, 9:00 to 17:00 Japan Standard Time (JST), UTC +9, hybrid event
Location: Rm2 (Room B-1), Kyoto International Conference Center
For both in-person and remote attendees, registration is required!
*Please make sure you register for our workshop (FOR FREE) via the above link before October 26, no matter you attend in person or remotely.
The organizers will check your name while you are in the Zoom waiting room or before entering room B-1. Without registration, you will not be admitted to the meeting.
AIM AND SCOPE
Medical robots are revolutionizing healthcare. Nowadays, minimally invasive surgeries and interventions assisted by robots are firmly in the mainstream due to less operation time, small trauma, and fast postoperative recovery. Robot-assisted rehabilitation is also widely adopted as it is able to provide repetitive, intensive, and high-quality physics training. However, the current level of autonomy and cognitive ability of medical robots, compared to those in e.g. autonomous driving or mobile robots, are relatively low. A number of challenges/open-ended questions still remain in the medical robotics field due to the complicated clinical environment as well as the strict safety requirements in clinical scenarios, such as:
How to enable robots to learn skills from expert surgeons?
How to manage the interaction between robots and complicated medical environments?
How to promote the human-robots symbiosis in surgery and rehabilitation?
How to better model and control medical robots with complex mechanics (e.g. continuum and soft robots)?
How to promote autonomous planning and decision-making of medical robots?
Robot learning (RL)/Machine Learning (ML) has produced notable achievements in manipulation tasks, and therefore drawn a lot of attention in the robotics and Artificial Intelligence (AI) community. Robot learning intends to enable robots to learn new skills by learning from demonstration or by interacting with environments. There is a large chance to effectively cope with the above challenges by introducing RL to the perception, modeling, control, and navigation of medical robots. We believe “How medical roboticists can better leverage the strength of RL/ML techniques while circumventing their limitations” is a trending topic that could engage a number of interdisciplinary experts and young professionals.
This workshop aims to bring world-class researchers (see speakers list) to present state-of-the-art research achievements and advances that introduce robot learning in medical applications. The organizers encourage participants to demonstrate new challenges and solutions on this topic by submitting Extended Abstracts/Posters. Note that the authors can present their abstracts remotely if accepted.
TOPICS OF INTEREST
This workshop will specifically focus on advancements in machine learning techniques (including deep learning, reinforcement learning) for medical robots’ perception, modeling, control, and navigation around the following key themes:
Machine learning applied to (semi-)autonomous medical robotics
(Deep) Reinforcement learning in medical robotics control
Machine learning in medical robotics modeling
Learning of surgical skills from demonstrations/videos
Learning for compliant behaviors in medical tasks
Medical robot sensing and localization based on machine learning
Machine learning for autonomous (sub-)task learning and execution
Machine-learning-based navigation and motion planning for medical robots
Machine Learning for medical workflow modeling
Machine Learning for evaluation of medical skills
Machine-learning-based visual servoing for medical robotics
Machine-learning-based human-robot interactive control
Machine-learning-based intention recognition in rehabilitation training
ORGANIZERS
Jing Guo
Guangdong University of Technology, China
Di Wu
KU Leuven, Belgium/TU Delft, the Netherlands
Loris Fichera
Worcester Polytechnic Institute, USA
Chao Zeng
Universität Hamburg, Germany
Zhan Li
Swansea University, UK
Farshid Alambeigi
University of Texas at Austin, USA
Important dates
Workshop date: October 27, 2022
Call for sponsorship deadline: July 31, 2022
Call for student assistant organizer deadline: July 31, 2022
Deadline for abstracts/posters submission: September 20, 2022 (Submission link: here)
Notification of acceptance: A notification will be given on a rolling basis within 14 days of submission.
Final version submission: October 10, 2022
IROS Conference: October 23 to October 27, 2022
hybrid participation
This workshop will be held in a hybrid format - allowing for both in-person and remote participation.
The IROS 2022 organizing committee then decided to hold the IROS as a traditional in-person conference. Virtual attendance will also be an option for those who have travel restrictions by their organization or do not feel comfortable participating in large gatherings, with the recognition that the conference cannot be experienced fully while remote.
Registration
No matter you want to attend in-person or remotely, you need to register for the workshop. Note that registration for the workshop is separate from registration for the conference, which means that you need to register for the workshop separately. Please register before August 22 in order to get an early-bird rate.
If you attend in person, you need to pay 62 USD (student member) /81 USD (student non-member).
If you attend online, you need to pay 7 USD (student member)/11 USD (student non-member).
The registration must be done through the IROS official registration portal. (please scroll down to the bottom of the page after you are directed to the IROS official website)
*All registrations must be paid in Japanese yen. The above prices are approximate based on the exchange rate. For more information regarding registration, see here. Registering for a workshop allows you to attend not only our workshop, but also other workshops or tutorials that are held on October 23rd and 27th.
IEEE/RAS TC SUPPORT
This proposed workshop is endorsed by the following IEEE RAS Technical Committees:
STUDENT ORGANIZATION
