Page 40 - Read Online

Page 124 Ding et al. Art Int Surg 2024;4:109-38 https://dx.doi.org/10.20517/ais.2024.16

Simulators serve as the essential infrastructure for creating, maintaining, and visualizing a DT of the
physical world, crucially facilitating the collection and transmission of data back to the real world [268,269] . The
Asynchronous Multi-Body Framework (AMBF) has demonstrated success in this regard through its
applications in the surgical domain [270,271] . Building on this foundation, the fully immersive virtual reality for
skull-base surgery (FIVRS) infrastructure, developed using AMBF, has been applied to DT frameworks and
applications, demonstrating significant advancements in the field [7,272-275] . Twin-S, a DT framework designed
for skull base surgery, leverages high-precision optical tracking and real-time simulation to model, track,
and update the virtual counterparts of physical entities - such as the surgical drill, surgical phantom, tool-to-
tissue interaction, and surgical camera - with high accuracy . Additionally, this framework can be
[7]
[256]
integrated with vision-based tracking algorithms , offering a potential alternative to optical trackers, thus
enhancing its versatility and application scope. Contributing further to the domain, a collaborative robot
framework has been developed to improve situational awareness in skull base surgery. This framework
[274]
introduces haptic assistive modes that utilize virtual fixtures based on generated signed distance fields (SDF)
of critical anatomies from preoperative CT scans, thereby providing real-time haptic feedback. The effective
communication between the real environment and the simulator is facilitated by adopting the Collaborative
Robotics Toolkit (CRTK) convention, which promotes modularity and seamless integration with other
[276]
robotic systems. Additionally, an open-source toolset that integrates a physics-based constraint formulation
framework, AMBF [270,271] , with a state-of-the-art imaging platform application, 3D Slicer , has been
[277]
[275]
developed . This innovative toolset enables the creation of highly customizable interactive digital twins,
incorporating the processing and visualization of medical imaging, robot kinematics, and scene dynamics
for real-time robot control.

In addition to AMBF-empowered DT models, other DT models have also been proposed and explored for
various surgical procedures. In liver surgery, a novel integration of thermal ablation with holographic
augmented reality (AR) and DTs offers dynamic, real-time 3D navigation and motion prediction to
improve accuracy and real-time performance . Similarly, in the realm of cardiovascular interventions, the
[278]
development of patient-specific artery models for coronary stenting simulations employs digital twins to
personalize treatments. This approach uses finite element models derived from 3D reconstructions to
validate in silico stenting procedures against actual clinical outcomes, underlining the move toward
[279]
personalized care . Orthopedic surgery benefits from applying DTs in evaluating the biomechanical
effectiveness of stabilization methods for tibial plateau fractures generated from postoperative 3D X-ray
[280]
images, aiding in optimizing surgical strategies and postoperative management . The utilization of DT,
AI, and machine learning to identify personalized motion axes for ankle surgery also marks a significant
advancement, promising improvements in total ankle arthroplasty by ensuring the precise alignment of
implants according to the specific anatomy of each patient . Moreover, introducing a method to
[281]
synchronize real and virtual manipulations in orthopedic surgeries through a dynamic DT enables surgeons
to monitor and adjust the patient’s joint in real time with visual guidance. This technique not only ensures
accurate alignments and adjustments during procedures but also significantly improves joint surgery
outcomes.

The potential of DTs extends further when considering their role in enhancing higher-level downstream
applications, ranging from surgical phase recognition and gesture classification to intraoperative guidance
systems. Leveraging geometric understanding, DTs can interpret the broader context and flow of surgical
operations, thereby increasing the precision and safety of interventions. Surgical phase recognition, for
instance, utilizes insights from both direct video sources and interventional X-ray sequences to accurately
identify the stages of a surgical procedure [11-18,282] . This facilitates a more structured and informed approach
to surgery, enhancing the decision-making process and the efficacy of robotic assistants [14,19-21] .