Open Access | Review
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Ushering in the next generation of autonomous surgical robots? current trends and future possibilities with data driven physics simulation and domain randomization
#These authors contributed equally to the manuscript.
*Corresponding author: Charbel Saade, PhD
Mailing address: Faculty of Health Sciences, American University
of Beirut Medical Center, Beirut, Lebanon. P.O.Box: 11-0236
Riad El-Solh, Beirut 1107 2020, Lebanon.
Email: mdct.com.au@gmail.com
Received: 20 May 2020 Accepted: 17 June 2020
DOI: 10.31491/CSRC.2020.06.051
Abstract
As artificial intelligence (AI) plays an ever-increasing role in medicine, various designs that implement
machine learning (ML) are being introduced in an effort to develop surgical robots that perform a variety of
surgical techniques without human interference. However, current attempts in creating autonomous surgical
robots (ASRs) are hindered by the amount of time needed to train a robot on a physical sett, the incredible
amount of physical and/or synthetic (artificial) data needed to be collected and labeled, as well as the
unaccountable and unpredictable characteristics of reality. Progress outside of the medical field is being made
to address the general limitations in autonomous robotics.
Herein, we present a review of the basics of machine learning before going through the current attempts in
creating ASRs and the limitations of current technologies. Finally, we present suggested solutions for these
limitations, mainly data driven physics simulations and domain randomization, in an attempt to create a
virtual training environment as faithful to and as random as the real world that could be transferred to a
physical setting. The solutions suggested here are based on techniques incorporated and strides being made
outside of the medical field that could usher in the next generation of autonomous surgical robotics designs.
Keywords
Autonomous surgical robots; robotic surgery; artificial intelligence; machine learning; physics simulation; domain randomization.
Introduction
Technological advancements in hardware and software
increasingly play an imperative role in the evolution of
contemporary medical/surgical techniques and paradigms. This, in addition to the high liability of medical errors, an increased workload, coupled with a reduced
workforce and an aging population, are incentivizing
experts to acquaint themselves with computerized
assistants or to introduce certain automated surgical
interventions [1–3]. One such technology is artificial intelligence (AI), specifically machine learning (ML). This
article reviews the current role of ML techniques in
surgery with a focus on autonomous robotics surgery
(ARS). Also, we provide a perspective on future possibilities that could help in enhancing the effectiveness of
autonomous surgical robots (ASRs), mainly data driven
physics simulation and domain randomization. The use
of ML in electronic medical record systems, diagnostics
and medical imaging is out of the scope of this review.
Searches were performed on Google scholar, Medline, PubMed, Scopus, Cochrane library and IEEE using various combinations of keywords: Autonomous surgical
robots, surgical and medical robotics, artificial intelligence, machine learning, physics simulation and domain
randomization.
As AI systems are being continuously adopted in medicine, there has been increasing interest in “autonomous” surgical robots that can assist surgeons or even
perform portions of an intervention independent of human guidance or control [3,4]. An autonomous intelligent
robot can be achieved using different variations of AI.
ML is a subset of AI and an increasingly growing field.
It is popular as it permits efficient processing of large
quantities of data for analysis, interpretation and decision making while providing computers with the ability to learn and perform a range of tasks without being
explicitly programmed to do so. Already widely used in
electronic medical record systems, medical imaging and
diagnostics, it is expected that ML will play a pivotal role
in surgical and interventional procedures [1,3,5].
ML agents can acquire surgical skills in a variety of ways,
one of which is, for example, through demonstration by
human experts [3]. Currently, intelligent surgical robots
with varying degrees of autonomy are proving to be
comparable to surgeons at some tasks, such as suturing,
locating wounds and tumor removal. These intelligent
surgical assistants could surpass the current state of the
art commercial surgical robots and promise good results
and a wider access to specialized procedures [1,3,5].
As promising as this might seem, debilitating limitations
currently hinder substantial progress in medical application of AI generally, and ASRs specifically. Mostly, these
limitations are linked to the current available AI technologies and partially to some unique characteristic of AI
application in medicine. The main limitations include:
The need for high-quality medical/surgical data which
slows the process of developing effective agents while
requiring large scale collaborative efforts- a modeling
challenge that hinders our ability to accurately “model”
a surgical environment that replicates the dynamic and
deforming nature of the living body- and the inability of
intelligent agents in general and surgical robots specifically to adapt to unknown or yet unobserved situations
[3].
Interestingly, new technological advancements in AI
software designs are being currently developed that,
we think, could help us overcome the aforementioned
limitations. Two particular new AI advancements that
could be of use are data driven physics simulation environments and domain randomization. First, we will
go through a general overview on ML parameters implemented in current autonomous surgical robots while
exploring some examples of present automated surgical
robotic technologies, we will then discuss some of the limitations of the current technologies and go through
our proposed solutions to overcome current designs
drawbacks.
Design
A robot is a system that has three main components: a set of sensors that detect the robot’s environment, actuators (or end effectors) that interact with and within the environment and a control architecture that processes sensory data and generates actions [3](Figure 1). ML is mainly involved in the control architecture, enabling the robot to “understand” the sensory input and generate a proper action. In order to do so, the agent must learn to generate a certain action in the context of a set of sensory inputs and desired goals. So, how can a robot learn a surgical skill? First, it could learn from human demonstration by observing experiments conducted by trained experts. The robot could also learn from its own interaction with the environment, by evaluating the appropriateness of its own actions to reach certain target states or goals [3,6]. In order to understand current autonomous robotic surgery technologies and future perspectives, a general knowledge of the main subsets of machine learning is needed. The three main categories of machine learning are supervised learning (SL), unsupervised learning (UL) and reinforcement learning (RL) [7]. Many agents use variations and combinations of these three categories.
In SL, training data are considered “labeled “, ie, the data
consist of a set of known input vectors along with a set of
known matching target vectors. The software creates a
function that links the input object with the corresponding output [8]. For example, consider we have a set of
photos of thousands of lung pathologies (pneumothorax,
pneumonia etcetera). The data is considered labeled if
each photo identifies certain features of the radiograph
such as opacification values (input vectors) and each
type of pathologies (target vector). Eventually, SL seeks
to build a predictor model that predicts target vectors for
new input vectors. Learning consists of finding optimal
parameter values for the predictor model [3].
In UL, the training data is considered unlabeled and consists of a set of input vectors without their corresponding
target vectors. UL aims to discover correlations and structure in the data. Whether using UL or SL, Gaussian Mixture
Model (GMM) and Gaussian Mixture Regression (GMR)
based learning could be added to fine tune the learning
process and get more reliable demonstration [3,9].
Reinforcement learning (RL) is concerned with how intelligent agents ought to take actions in an environment in
order to maximize cumulative reward. The difference between RL and SL is that RL does not need labelled input/
output data pairs, the training data is mostly generated
through direct interaction with the environment, and RL
does not need explicit correction of sub-optimal actions. RL focuses on exploring the environment (states) in a trial
and error approach in order to create and adjust a policy
that permits the agent to perform a certain task (action).
RL’s end goal is to learn a policy that represents a mapping
from states to actions. For example, suppose that we want
to train a robot to tie a knot. The agent will explore the
environment, gradually develop a certain policy (creating
and adjusting a suturing technique) in order to be eventually capable of performing the task of tying the knot. A
value of a state-action pair is created and represents how
good it is for the robot to perform an action in a given
condition or state [3,7].
RL can be accelerated using implicit imitation learning
(IML), a technique that allows the agent to learn a skill
through the observation of an expert mentor [10]. For example, IML can be used to teach an agent a surgical skill
by observing and imitating the performance of an actual
surgeon. The agent observes and analyzes the state transitions of the surgeon’s actions (for example, how the surgeon moves from one position to another, and the composition of every maneuver) and uses the information to
update its own states and actions. In a series of works,
trajectories recorded from human subjects are used to
generate an initial policy (the action generated based on
certain state). Additionally, some algorithms for imitation
learning can learn from several mentors and are being
used to transfer knowledge between agents with different
reward structures [3,10].
Inverse RL (IRL) is a technique for imitation learning that
also consists of an observer agent and a mentor. In IRL
the agent learns the reward function of the environment
from the observer. Then the agent builds the policy that
maximizes the reward function using classical RL’s trial
and error approach [3].
Another layer of ML is deep learning (DL). DL allows for
processing of huge chunks of data to find relationships in
sets that are often impossible to explicitly label, such as the
pixel in a given set of pictures. The basic architecture of DL
consists of what are called neural networks, that are analogous to the neurons and synapses of the human brain,
providing much of the ability to learn. Deep learning can
be used in combination with the previously mentioned
techniques [11].
Other aspects of ML deals with fine tuning the agent to
account for the unpredictability of, or to accurately represent physical systems. These techniques include system
identification, high-quality rendering [12–14], domain adaptation [15] and iterative learning control [16]. Although these
methods have some advantages, they generally require
large amounts of data and are still labile to unexpected
changes and unaccountable environmental elements [17].
It is important to note that ML is generally used with other methods in order to optimize robotic capabilities.
Even though the control architecture (where ML is mostly needed) governs the action performed based on the
current state of the environments, two important components cannot be overlooked; the sensors needed for
environment observation and the effectors to perform a
certain task. Indeed, surgical interventions include interactions with delicate and deformable structures. So, in
order to operate, sensory and motor apparatus are being used to detect, for example, the depth needed for a
given maneuver, tissue consistency and to generate the
required force and direction of a specific maneuver [3,4].
For example, Sozzi et al. used real-time adaptive motion
planner (RAMP) to generate collision-free robot motions
to avoid obstacles within the workspace [18]. For the Smart
Tissue Anastomosis Robot (STAR), in order to achieve
proper suturing, the team used force sensors and infrared
bio-glue to prevent tissue deformation and to guide the
robot for needle insertion [19].
Next, we will be dealing with current examples of autonomous surgical robots that have incorporated one or more
of the methods presented.
What robots do we have so far?
Van den Berg et al. employed imitation learning to develop
an agent that learns tasks from multiple human demonstrations, optimizing speed and smoothness of task execution. The technique was employed on the Berkeley Surgical Robot and used for knot-tying and drawing figures
[16]. Schulman et al. also developed an agent that learns by
human demonstration using a trajectory transfer method.
The agent was able to learn five different types of knots
[20]. Recently, Calinon et al. used inverse RL to transfer
skill from a surgeon teleoperator to a flexible robot. In
this method the agent and the mentor may have different
morphological structures and still handle the case of skill
transfer [21].
Moreover, Mayer et al. apply ML for suturing and knot-tying, using recurrent neural networks (a subset of deep
learning), publishing a series of work [22,23]. Also, they used
imitation learning to create a suturing robot using principles known from fluid dynamics [24].
Weede et al. developed an intelligent autonomous endoscopic guidance system that anticipates the surgeon’s
next action during a procedure and adjusts the position
of the endoscopic camera accordingly. The system uses
information on the movements of the instruments from
previous procedures [25].
Krieger el al. used The STAR robot to suture bowels in pigs.
They used the concept of supervised autonomous suturing, where the surgeon outlines the incision area then the
robot uses sensors and combined 3D imaging to assist in
suturing of intestinal anastomosis. The STAR robot was
able to place evenly spaced and leak-proof sutures in a
trial with live pigs. However, sometimes the surgeons had
to make small adjustments to the thread’s position for accurate suturing [19]. Moreover, Krieger has taught the robot
to remove tumors with infrared markers that were used to
mark cancerous areas, the robot then excises these parts
with, though preliminary, human level accuracy [5].
Mylonas et al. used GMM in designing an algorithm that
learns from human demonstration. They created a basic
autonomous eFast ultrasound scanning by a robotic manipulator [26]. Kassahun et al. used GMM modeled joint
probability densities to make their agent capable of understanding the model of interaction between the aorta
and the catheter in interventional procedures [27].
Another agent used in interventional procedures developed by Fagogenis et al. used what they called “Haptic
vision” to assist in paravalvular leak closure of prosthetic
valves. They designed a robotic catheter that can navigate autonomously (using leak locations localized from
pre-operative imaging) to the aortic valve and deploy an
occluder into the site of a leak. An operator then deploys
the occluder. Haptic vision combines machine learning
with intracardiac endoscopy and image processing algorithms to form a hybrid imaging and touch sensor. Machine learning was primarily used to enable the catheter
to distinguish the blood and tissue from the prosthetic
aortic valve [28].
Last but not least, AI agents are being introduced in Urological procedures. The AquaBeam™ robotic system was
approved as a water ablative therapy for the resection
of the prostate. Although the technology requires human
impute and image planning, the robotic system assists
in resecting, surgeon defined, prostatic tissue using high
velocity saline jet while autonomously adjusting various
flow rates based on the depth, length and width of the
area being resected [29].
Limitations and how to overcome time
All current ML technologies used in ARS, variably share some common drawbacks, mainly, the highly unpredictable nature of the physical world, restraints regarding the training environment, and the amount of data and time needed to train an optimal machine. Training and testing an intelligent agent using available models tends to be time-consuming which generally involves manually collecting and labelling huge amounts of data, for example when using supervised learning. This is problematic when the job requires data that are difficult to obtain in large quantities with necessary variability, labels that are difficult to specify, and/or expert knowledge [30]. The end goal of overcoming these limitations is, broadly, to create a system that can implement the desired actions (suturing for example) with great reliability, flexibility and safety [4]. The question is whether it is possible to create the optimal environment for ML to overcome the physical constraints of the real world (including the data availability issue), and increase the speed of skill acquisition while being adaptable to random changes and the complex nature of the physical reality. We will be focusing on modeling challenges, data limitation and adaptation to physical changes. The solutions suggested here are based on developments outside the medical field, as we suggest the need to incorporate these techniques in future autonomous surgical robotics designs.
Modeling challenges and physical data limitations
One of the major challenges in modeling the surgical environment is the deforming and dynamic nature of the living body due to physiological, pathological and even external phenomena. For that purpose, mechanical, geometric, and physiological behavior of the environment should be considered. The current methods that rely on intraoperative inputs are not optimal as they involve theoretical and technical challenges related to the interpretation of sensory information, such as sensor co-registration, synchronization and information fusion, which are highly fragile as well as the need for annotated real world data [3]. Moreover, applying ML sometimes employs random exploration, which can be hazardous in any real physical training set. ML might also often require thousands or millions of samples, which could take a tremendous amount of time to collect in a real physical word setting, making it impractical for many applications [31]. One way of overcoming this limitation is through learning in simulation.
Data driven physics simulation
Recent results in learning in simulation are promising
for building robots with human-level performance on a
number of complex tasks [32,33]. Ideally, an agent should
learn policies that encode complex behaviors utterly
in simulation and apply those policies successfully via
physical robots.
Moreover, one can speculate that an optimal simulation
should be as faithful as possible to the real world. Here
comes the role of physics simulators. High quality physics simulations are being used in computer graphics to
replicate the physical world, from dynamic fracture animation to fluids and particles simulations [34,35]. Although
they do not provide an interactive environment for real
time simulation, they could be used as a cornerstone
to replicate physical reality. On the other hand, there
are other methods that allow for real time interactions
within the virtual world. For example, Holden et al. developed a data-driven physics simulation method that
supports real time interactions with external objects.
Their method combines ML with subspace simulation
techniques which enables a very efficient physics simulation that supports accurate interactions with external
objects, surpassing existing models [36]. Seunghwan et
al. created a physics-based simulation of a human musculoskeletal model composed of a skeletal system and
300 muscles with a control system, creating a reliable
simulation of anatomical features with robust control of
dynamical systems that generates highly realistic human
movements. Also, their model demonstrates how movements are affected in specific pathological conditions
such as bone deformities and when applying various
prostheses [37]. The same authors also formulated a technique called VIPER that creates realistic muscle models
that simulate controllable muscle movement and even
muscle growth [38]. Moreover, one can use MuJoCo physics
engine which is commonly used to create advanced virtual environments for ML. MuJoCo was used by OpenAI
research to virtually train a robotic hand that can solve
a RubiK’s cube [39]. It is only a matter of time before we
reach ultra-realistic real time physics simulators that
include complex anatomical and physiological elements.
So how do these approaches attempt to replicate reality? The aforementioned models and other approaches
sometimes make the simulator to closely match physical
reality by performing a variety of techniques including
system identification, high-quality rendering [12–14], domain adaptation [15] and iterative learning control [16]. Although they are the best methods to account for known
physical entities and generally do not directly rely on
physical data, the problem in these techniques is that
they are still suboptimal in accounting for the randomness of the real world, requires large sets of synthetic
data impute (both data demanding and time consuming)
and many times still requires additional training on real-world data [17].
So, we have techniques to simulate the real world that
can be trained fast and does not require real physical
data input, now the question is how can we ensure that
our model can deal with uncounted randomness, can be
trained fast, and does not require large synthetic data
input?
Adaptation to unknown situations and overcoming synthetic data limitation
Any system with decision-making power in the operating room should guarantee the safety of the patient
while being able to cope with unpredictable events and
the uncertainty of the living body. A critical challenge is
to develop intelligent agents that are able to adapt the
learned skills to unexpected and novel situations [3]. For
solving the modelling problem, we suggested the use of
virtual physics simulation. Unfortunately, incongruities
between reality and simulators make transferring skills from simulation problematic. For example, system identification, a process used for adjusting the parameters
of the simulation to match the characteristics of the
real world and the behavior of the physical system (i.e
robot), is error-prone and time-consuming. Even with
other techniques such as high-quality rendering [12–14],
domain adaptation [15] and iterative learning control [16],
the real world has physical effects that are hard to model
and are not captured by current real-time physics simulators like gear backlash, nonrigidity, fluid dynamics
and wear-and-tear. Furthermore, simulators are often
unable to produce the noise and richness embedded in
the real physical world. These differences, known as the
reality gap, are considered bottleneck to the usage of
simulated data on real physical robots [17,39].
Bridging the ‘reality gap’ that separates experiments
on hardware from simulated robotics might accelerate
autonomous robotic developments through improved
synthetic data availability. This brings us to domain
randomization (DR), a new method for training agents
on simulated environments that transfer to the real environment by randomizing rendering in the simulator.
In other words, the parameters of the simulation are
randomized in a way that with enough variability in the
simulator, the real world appears to the agent as just
another variation. So, the underlying hypothesis is this: if
the variability in simulation is significant enough, agents
trained in simulation will generalize to the real world.
It is important to note that researchers can also try to
use DR in combination with other techniques that optimize the physics simulation, which might improve the
results [17,39].
Domain randomization
In DR, the parameters of the simulator—like lighting,
pose, object textures, and other physical aspects—are
randomized to oblige the agent to learn the essential
features of the object and task of interest. DR requires
us to specify what aspects we want to randomize, and
specify the variable testing states. Although in its early
form, the importance of DR is that it allows for the possibility to produce an agent with strong performance
using low-fidelity synthetic data. This introduces the
possibility of using inexpensive artificial (synthetic)
data for training agents while avoiding the need to collect and label incredible amounts of real-world data or to
generate highly realistic artificial worlds [30]. One of the
earliest works on domain randomization was presented
by Tobin el al. The team used DR in the setting of RL and
managed to train a robot virtually to localize presented
objects. They were able to train an accurate real-world
object detector that is resistant to partial occlusions and
other distractors using synthetic data from a simulator
with non-realistic random textures. The detectors were also used to perform grasping in a messed up real environment [17]. Also, OpenAI used RL with DR to enable a
robot to learn dexterous in-hand manipulations [40].
Automatic domain randomization (ADR), is a variation
of DR that randomizes the parameters of the simulator
automatically, without the need to specify what elements
of the simulation we want to change. ADR automatically
generates a distribution over randomized environments
of ever-increasing difficulty, thus creating millions for
scenarios for the learning algorithm. The latest implementation of ADR was conducted by the OpenAI team,
who trained a robotic arm to solve a rubik’s cube. The
robot was able to solve the cube in a real world setting,
even with intentional disturbance of the environment
by the researchers [39].
Compared to iterative learning control and domain adaptation which are important tools for addressing the
reality gap, DR does not require additional training on
real-world data. Although DR requires no additional real
world training, it can also be combined easily with most
parallel techniques, and we should consider using it in
combination with realistic physics simulations and possibly even, when possible, other training methods such
as imitation learning [17]. DR (or ADR), through randomization, reduces the need of synthetic data while creating
an agent robust to changes in the real world. It is important to note that DR is still a new technique and will
require further optimization to improve its usefulness
and applicability.
In brief, what we are suggesting in this review is to try
to combine optimal realistic physical simulation techniques with DR (or ADR) and other parallel techniques
in a sense that we can create an environment as faithful
to the real world, and as random as the real world in
order to produce optimal training environment for surgical agents thus creating the most reliable autonomous
surgical robots.
Conclusion
Autonomous robots will be needed to address the decreased work force, the increased demand for surgery and the high risk of medical errors. Current technologies used in medical robotics require huge amounts of data input, are difficult to train, and prone to minor changes in the environment. Using physics simulation techniques combined with DR might be what is needed to overcome these limitations in order to create the advances desirable in autonomous robotic surgery. Our approach focuses on virtual training with domain randomization. Of course, to apply the learned skills in the real world certain specific hardware will be needed, however these technologies are currently available in the form of sensors and other hardware already in use in many current medical and non-medical models. Here we are adding this layer of training that if linked properly to a physical hardware, the aforementioned limitations should be eventually overcome. In essence, more trial and less error.
Declarations
Conflicts of interest
Te authors declare that they have no conflict of interest.
Authors’ contributions
Dr. Youssef Ghosn, Dr. Mohammed Hussein Kamareddine, Mr. Geroge Salloum, Dr. Elie Najem, Mr. Ayman Ghosn and Dr. Charbel Saade, designed and conceptualized the study and drafed the manuscript for intellectual content. All authors approved the fnal version and agreed to be accountable for all aspects of the work.
References
1. Aruni, G., Amit, G., & Dasgupta, P. (2018). New surgical
robots on the horizon and the potential role of artificial
intelligence. Investigative and clinical urology, 59(4),
221-222.
2. Anderson, J. G., & Abrahamson, K. (2017). Your Health
Care May Kill You: Medical Errors. In ITCH (pp. 13-17).
3. Kassahun, Y., Yu, B., Tibebu, A. T., Stoyanov, D., Giannarou, S.,
Metzen, J. H., & Vander Poorten, E. (2016). Surgical robotics
beyond enhanced dexterity instrumentation: a survey of
machine learning techniques and their role in intelligent
and autonomous surgical actions. International journal of
computer assisted radiology and surgery, 11(4), 553-568.
4. Taylor, R. H., Kazanzides, P., Fischer, G. S., & Simaan, N.
(2020). Medical robotics and computer-integrated
interventional medicine. In Biomedical Information
Technology (pp. 617-672). Academic Press.
5. Svoboda, E. (2019). Your robot surgeon will see you
now. Nature, 573, S110-S111.
6. Taylor, R. H., Kazanzides, P., Fischer, G. S., & Simaan, N.
(2020). Medical robotics and computer-integrated
interventional medicine. In Biomedical Information
Technology (pp. 617-672). Academic Press.
7. Kaelbling, L. P., Littman, M. L., & Moore, A. W. (1996).
Reinforcement learning: A survey. Journal of artificial
intelligence research, 4, 237-285.
8. Russel, S., & Norvig, P. (2013). Artificial intelligence: a
modern approach. Pearson Education Limited.
9. Hinton, G. E., Sejnowski, T. J., & Poggio, T. A. (Eds.).
(1999). Unsupervised learning: foundations of neural
computation. MIT press.
10. Price, B., & Boutilier, C. (2003, August). A Bayesian
approach to imitation in reinforcement learning.
In IJCAI (pp. 712-720).
11. Guo, Y., Liu, Y., Oerlemans, A., Lao, S., Wu, S., & Lew, M.
S. (2016). Deep learning for visual understanding: A
review. Neurocomputing, 187, 27-48.
12. Planche, B., Wu, Z., Ma, K., Sun, S., Kluckner, S., Lehmann,
O., ... & Ernst, J. (2017, October). Depthsynth: Real-time
realistic synthetic data generation from cad models for
2.5 d recognition. In 2017 International Conference on 3D
Vision (3DV) (pp. 1-10). IEEE.
13. James, S., & Johns, E. (2016). 3d simulation for robot
arm control with deep q-learning. arXiv preprint
arXiv:1609.03759.
14. Richter, S. R., Vineet, V., Roth, S., & Koltun, V. (2016,
October). Playing for data: Ground truth from computer
games. In European conference on computer vision (pp.
102-118). Springer, Cham.
15. Cutler, M., & How, J. P. (2015, May). Efficient reinforcement
learning for robots using informative simulated priors.
In 2015 IEEE International Conference on Robotics and
Automation (ICRA) (pp. 2605-2612). IEEE.
16. Van Den Berg, J., Miller, S., Duckworth, D., Hu, H., Wan, A., Fu,
X. Y., ... & Abbeel, P. (2010, May). Superhuman performance
of surgical tasks by robots using iterative learning from
human-guided demonstrations. In2010 IEEE International
Conference on Robotics and Automation (pp. 2074-2081).
IEEE.
17. Tobin, J., Fong, R., Ray, A., Schneider, J., Zaremba, W., &
Abbeel, P. (2017, September). Domain randomization for
transferring deep neural networks from simulation to the
real world. In 2017 IEEE/RSJ international conference on
intelligent robots and systems (IROS) (pp. 23-30). IEEE.
18. Sozzi, A., Bonfè, M., Farsoni, S., De Rossi, G., & Muradore,
R. (2019). Dynamic Motion Planning for Autonomous
Assistive Surgical Robots. Electronics, 8(9), 957.
19. Leonard, S., Wu, K. L., Kim, Y., Krieger, A., & Kim, P. C.
(2014). Smart tissue anastomosis robot (STAR): A visionguided robotics system for laparoscopic suturing. IEEE
Transactions on Biomedical Engineering, 61(4), 1305-
1317.
20. Schulman, J., Ho, J., Lee, C., & Abbeel, P. (2016). Learning
from demonstrations through the use of non-rigid
registration. In Robotics Research (pp. 339-354). Springer,
Cham.
21. Calinon, S., Bruno, D., Malekzadeh, M. S., Nanayakkara,
T., & Caldwell, D. G. (2014). Human–robot skills transfer
interfaces for a flexible surgical robot. Computer methods
and programs in biomedicine, 116(2), 81-96.
22. Mayer, H., Gomez, F., Wierstra, D., Nagy, I., Knoll, A.,
& Schmidhuber, J. (2008). A system for robotic heart
surgery that learns to tie knots using recurrent neural
networks. Advanced Robotics, 22(13-14), 1521-1537.
23. Mayer, H., Nagy, I., Burschka, D., Knoll, A., Braun, E. U.,
Lange, R., & Bauernschmitt, R. (2008, March). Automation
of manual tasks for minimally invasive surgery. In Fourth
International Conference on Autonomic and Autonomous
Systems (ICAS'08) (pp. 260-265). IEEE.
24. Mayer, H., Nagy, I., Knoll, A., Braun, E. U., Lange, R., &
Bauernschmitt, R. (2007, April). Adaptive control for
human-robot skilltransfer: Trajectory planning based on
fluid dynamics. In Proceedings 2007 IEEE International
Conference on Robotics and Automation (pp. 1800-1807).
IEEE.
25. Weede, O., Mönnich, H., Müller, B., & Wörn, H. (2011, May).
An intelligent and autonomous endoscopic guidance
system for minimally invasive surgery. In 2011 IEEE
International Conference on Robotics and Automation(pp.
5762-5768). IEEE.
26. Mylonas, G. P., Giataganas, P., Chaudery, M., Vitiello, V.,
Darzi, A., & Yang, G. Z. (2013, November). Autonomous
eFAST ultrasound scanning by a robotic manipulator
using learning from demonstrations. In 2013 IEEE/
RSJ International Conference on Intelligent Robots and
Systems (pp. 3251-3256). IEEE.
27. Kassahun, Y., Yu, B., & Vander Poorten, E. (2013). Learning
catheter-aorta interaction model using joint probability
densities. InProceedings of the 3rd joint workshop on new
technologies for computer/robot assisted surgery (pp.
158-160).
28. Fagogenis, G., Mencattelli, M., Machaidze, Z., Rosa, B.,
Price, K., Wu, F., ... & Dupont, P. E. (2019). Autonomous
robotic intracardiac catheter navigation using haptic
vision. Science robotics, 4(29), eaaw1977.
29. Navaratnam, A., Abdul-Muhsin, H., & Humphreys,
M. (2018). Updates in urologic robot assisted
surgery. F1000Research, 7.
30. Tremblay, J., Prakash, A., Acuna, D., Brophy, M., Jampani, V.,
Anil, C., ... & Birchfield, S. (2018). Training deep networks
with synthetic data: Bridging the reality gap by domain
randomization. In Proceedings of the IEEE Conference on
Computer Vision and Pattern Recognition Workshops(pp.
969-977).
31. Mnih, V., Kavukcuoglu, K., Silver, D., Rusu, A. A.,
Veness, J., Bellemare, M. G., ... & Petersen, S. (2015).
Human-level control through deep reinforcement
learning. Nature, 518(7540), 529-533.
32. Levine, S., Finn, C., Darrell, T., & Abbeel, P. (2016). End-toend training of deep visuomotor policies. The Journal of
Machine Learning Research, 17(1), 1334-1373.
33. Schulman, J., Levine, S., Abbeel, P., Jordan, M., & Moritz,
P. (2015, June). Trust region policy optimization.
In International conference on machine learning (pp.
1889-1897).
34. Wolper, J., Fang, Y., Li, M., Lu, J., Gao, M., & Jiang, C. (2019).
CD-MPM: continuum damage material point methods
for dynamic fracture animation. ACM Transactions on
Graphics (TOG), 38(4), 1-15.
35. Gao, M., Pradhana, A., Han, X., Guo, Q., Kot, G., Sifakis, E.,
& Jiang, C. (2018). Animating fluid sediment mixture
in particle-laden flows. ACM Transactions on Graphics
(TOG), 37(4), 1-11.
36. Holden, D., Duong, B. C., Datta, S., & Nowrouzezahrai, D.
(2019, July). Subspace neural physics: fast data-driven
interactive simulation. In Proceedings of the 18th annual
ACM SIGGRAPH/Eurographics Symposium on Computer
Animation (pp. 1-12).
37. Lee, S., Park, M., Lee, K., & Lee, J. (2019). Scalable muscleactuated human simulation and control.ACM Transactions
on Graphics (TOG), 38(4), 1-13.
38. Angles, B., Rebain, D., Macklin, M., Wyvill, B., Barthe,
L., Lewis, J., ... & Tagliasacchi, A. (2019). Viper: Volume
invariant position-based elastic rods. Proceedings
of the ACM on Computer Graphics and Interactive
Techniques, 2(2), 1-26.
39. Li, T., Xi, W., Fang, M., Xu, J., & Meng, M. Q. H. (2019). Learning
to Solve a Rubik's Cube with a Dexterous Hand. arXiv
preprint arXiv:1907.11388.
40. Andrychowicz, O. M., Baker, B., Chociej, M., Jozefowicz, R.,
McGrew, B., Pachocki, J., ... & Schneider, J. (2020). Learning
dexterous in-hand manipulation. The International
Journal of Robotics Research, 39(1), 3-20.