Robotized Dissector Systems

• Robotized dissectors are electromechanical instruments designed for precise, minimally invasive tissue dissection via advanced actuation, sensing, and integrated imaging.
• They utilize diverse architectures—such as concentric push/pull robots, cable-driven manipulators, and continuum systems—to achieve multi-degree-of-freedom dexterity and real-time control.
• Experimental validations demonstrate enhanced surgical accuracy, reduced procedure times, and improved ergonomics, paving the way for safer, more efficient clinical interventions.

A robotized dissector is a class of electromechanical instrument or system designed to enable dexterous, precise, and minimally invasive surgical tissue dissection via robotic actuation, manipulation, and often real-time feedback from advanced sensing or imaging modalities. Robotized dissectors are critical to expanding the anatomical access, safety, and repeatability of complex surgical interventions such as pulmonary thromboendarterectomy, endoscopic submucosal dissection (ESD), transoral laser microsurgery, and automated procedures including boundary-following dissection in cholecystectomy or other organ systems. Architectures range from slender, highly articulated manipulators with multi-function channels and integrated endoscopy to bimanual cable-driven systems, continuum manipulators, and AI-augmented, vision-based automation frameworks (Vrielink et al., 2017, Scott et al., 2024, Oh et al., 2023, Liang et al., 2024, Braglia et al., 10 Jan 2026, Zhu et al., 3 Feb 2026).

1. Core Mechanical and Actuation Architectures

Robotized dissectors span a range of mechanical designs, from rigid modular forceps actuators to slender concentric-tube robots and tendon-driven continuum manipulators. Key examples include:

• Concentric Push/Pull Robot (CPPR)-Based Dissector The CPPR dissector comprises two nested steerable segments, each constructed from laser-patterned, thin-walled medical-grade 316L stainless-steel tubes with tenon–mortise slits. Each segment achieves planar bending by axial push/pull of its inner tube relative to its outer tube, and rotation about the long axis enables 3D articulation. Segment diameters are OD = 3.5 mm (proximal) and OD = 2.9 mm (distal), supporting dual-segment 6-DoF distal dexterity. The hollow center supports endoscopic imaging, irrigation, suction, and tool delivery (Zhu et al., 3 Feb 2026).
• Cable-Driven Parallel Manipulators (CDPM) and Continuum Manipulators ESD CYCLOPS employs a 6-tendon CDPM with a rigid scaffold anchoring each cable peripherally, allowing 5-DoF actuation (x,y,z, pitch, yaw) of each instrument. Bowden cables permit off-site motors. Differential tensioning yields workspace volumes ~12,400 mm³ and force outputs up to 46 N (Vrielink et al., 2017). Low-cost open-source continuum dissectors use a stainless-steel backbone, four tendons for 3-DoF bending, and a modular interface mountable on a UR5e macro-positioner (Scott et al., 2024).
• Robotic Forceps for Microsurgery Systems designed for upper aerodigestive tract microsurgery utilize precision-motorized forceps mounted to a macro manipulator, incorporating a dedicated RCM mechanism for shaft-pivoted actuation. Open/close control is separated and synchronized with a teleoperated haptic interface (Braglia et al., 10 Jan 2026).

2. Kinematic Modeling and Control Paradigms

Robotized dissectors require accurate and efficient forward/inverse kinematic solutions to map actuator inputs (tendon tensions, tube translations/rotations, gripper states) to tip pose/orientation, often under constraints dictated by workspace size, tool shaft dimensions, and remote center-of-motion (RCM) geometries.

• Concentric-Tube Robot Kinematics Let $Q=\{q_p,D_p,\varphi_p,q_d,D_d,\varphi_d\}$ denote translations, pull distances, and base rotation angles of proximal and distal segments. Forward kinematics $P = f(Q)$ and orientation $\Re = g(Q)$ leverage constant-curvature assumptions. Inverse kinematics is formulated as a constrained optimization $$\hat Q = \arg\min_Q\|f(Q)-\tilde P\|_2 + \chi\|g(Q)-\tilde\Re\|_2$$, subject to actuation and nesting constraints (Zhu et al., 3 Feb 2026).
• CDPM and Continuum Manipulator Kinematics For CDPM, the Jacobian maps tendon length rates to tip motion and tendon tension to end-effector forces (Vrielink et al., 2017). Continuum robot pose is parameterized by curvature $\kappa$, bending plane $\phi$, and length $L$, with tip position

$$p(\kappa, \phi) =\frac{1}{\kappa}\left[ (1-\cos(\kappa L))\cos\phi, (1-\cos(\kappa L))\sin\phi, \sin(\kappa L) \right]^T$$

(Scott et al., 2024).

• RCM Enforcement in Teleoperated Forceps RCM constraints are implemented via twist mapping and proportional tip translation correction in the control loop, ensuring the instrument pivots about the desired anatomic entry point (Braglia et al., 10 Jan 2026). Kinematic transformations propagate teleoperation inputs to robot joint velocities, preserving RCM fidelity.

3. Sensing, Perception, and Automation Frameworks

Surgical dissection demands real-time spatial awareness, tissue type discrimination, and, in cutting-edge systems, autonomous or semi-autonomous closed-loop control.

• Stereo Vision and Image-Assisted AI Automated dissection systems employ stereo endoscopes to provide real-time 3D reconstruction, while AI models (e.g., Mask R-CNN/Detectron2) segment tissues and localize instrument keypoints. Segmentation achieves AP above 94% for instrument tip detection and gallbladder tissue, with lower scores (37.5%) for challenging liver tissue (Oh et al., 2023).
• Boundary Extraction and Visual Servoing 3D tissue boundaries are computed via disparity jumps along segmented masks, generating discrete waypoints for position-based visual servoing (PBVS). Robot tip motion then follows $\Delta p = \alpha\hat e_t$ for discrete advancement along boundaries, with error thresholds for progression (Oh et al., 2023).
• Deformable Physics Models and Jacobian-Based Grasp Optimization The MEDiC framework integrates XPBD tissue simulation with differentiable Jacobian analysis to select assistive positions optimizing wedge-opening and minimizing shear at the dissection site. Average expansion ratios $\rho$ validate significant improvements in tissue exposure, with the APS algorithm achieving $1.74 \pm 0.17$ (100% success) vs. $1.29 \pm 0.13$ without APS (Liang et al., 2024).

4. Performance Metrics and Experimental Validation

Assessment of robotized dissectors spans workspace coverage, dexterity, force output, accuracy, task completion, and ergonomic/usability impacts.

Design	Tip Accuracy (mm RMSE)	Max Tip Force (N)	Workspace/Reach
CPPR Dissector (Zhu et al., 3 Feb 2026)	1.1–1.85	1.7	60 mm length; 3.5 mm Ø
ESD CYCLOPS (Vrielink et al., 2017)	0.217 (ellipse tracing)	up to 46	~12,400 mm³
Continuum Open-Source (Scott et al., 2024)	1.45	1.8	Tip sweep ~50 mm
Teleop. Forceps (Braglia et al., 10 Jan 2026)	—	—	Full TLM workspace

• CPPR-Based Dissector: Demonstrated successful navigation of airways/arteries and tissue stripping with <2 mm error over 60 mm (ex vivo lung, bronchial models), and multi-channel operation (irrigation, suction, tool) (Zhu et al., 3 Feb 2026).
• ESD CYCLOPS: Attained tip forces of up to 46 N, and mean radial errors of 0.217 mm in tracing; validated in chicken and porcine ex vivo ESD (Vrielink et al., 2017).
• Open-Source Continuum: Achieved 1.45 mm RMSE position accuracy on a phantom, 1.8 N max force, but noted limited slicing capability for fibrous tissue (Scott et al., 2024).
• Teleoperation Ergonomics: Robotic manipulation reduced hand tremor by 71% (RMS vibration), and decreased forearm muscle activation by 20–35% compared to freehand (Braglia et al., 10 Jan 2026).

5. Clinical and Procedural Impact

Robotized dissectors are designed to enhance access and safety in demanding anatomical and procedural contexts:

• Pulmonary Thromboendarterectomy (PTE): The CPPR dissector enables entry into third- and fourth-class pulmonary artery branches (down to ~4 mm Ø) not accessible by rigid tools, supports endoscopic vision, and multi-functionality, reducing deep-hypothermic circulatory arrest duration (clot removal per turn reduced from ~20 min to ~5 min) (Zhu et al., 3 Feb 2026).
• Upper Aerodigestive Tract Microsurgery: Introduction of RCM-constrained robotic forceps improves stability, precision, and surgeon ergonomics in transoral laser microsurgery, achieving high usability scores (SUS median 78.5, “Good”) and significantly attenuating involuntary tremor (Braglia et al., 10 Jan 2026).
• Endoscopic Submucosal Dissection (ESD): Bimanual CDPM systems such as CYCLOPS deliver workspace tailored to ~20–30 mm lesions, with validated ex vivo performance and sub-millimeter tracking (Vrielink et al., 2017).
• Automated and AI-Driven Procedures: Frameworks for automated cholecystectomy and deformable-tissue dissection integrate 3D vision, real-time segmentation, and model-based as well as learning-driven control, achieving boundary-following errors ∼0.36 ± 0.34 mm and fully automated sub-tasks (Oh et al., 2023, Liang et al., 2024).

6. Limitations and Future Research Directions

Challenges remain in the fields of soft-tissue interaction, real-time force feedback, modeling accuracy, and integration for real-world clinical deployment.

• Modeling Soft-Tissue Deformation: Current quasi-static simulations (XPBD) and frictional contact models oversimplify viscoelasticity and may lead to unmodeled slippage or force misestimation. Real-to-simulation material property discrepancies can introduce Jacobian errors, limiting closed-loop convergence and planning (Liang et al., 2024).
• Force Sensing and Haptic Feedback: Force feedback, both for safety and enhanced manipulation, is in early stages; many platforms rely on inference via actuator current or require new sensor modalities (e.g., micro-load cell, FBG shape sensing) (Scott et al., 2024, Vrielink et al., 2017).
• Workspace and Channel Limitations: Slender designs must balance sufficient stiffness with extreme slenderness to reach deep anatomical sites without buckling, and maintain multi-channel operability (camera, irrigation, tool) in constrained diameters (Zhu et al., 3 Feb 2026).
• Automation and Multi-Arm Coordination: Incorporation of autonomous grasp/pull strategies, real-time segmentation under deformation, and boundary replanning are key next steps. Extension to fully automated dual-arm operation is under exploration (Liang et al., 2024, Oh et al., 2023).

Planned research directions include model refinement using online residual learning, coordinated multi-arm tensioning for improved tissue control, high-definition endoscopic integration, use of superelastic and sterilizable materials, in vivo clinical validation, and modular open-source expansion enabling broad research access and adaptation (Zhu et al., 3 Feb 2026, Scott et al., 2024, Liang et al., 2024).

7. Open-Source and Accessibility in Robotized Dissector Research

Open-source initiatives significantly lower adoption and R&D barriers. Complete mechanical models, control packages, and firmware are released for certain platforms, supporting adaptation to diverse research or procedural end-uses (Scott et al., 2024). This strategy accelerates dissemination, reproducibility, and cross-institutional benchmarking, addressing the historical challenge of proprietary system inaccessibility. A plausible implication is accelerated translation of robotized dissection capabilities across specialties and broader global research participation.