Advancing Minimally Invasive Precision Surgery in Open Cavities with Robotic Flexible Endoscopy

TL;DR

The ability of this system to address the key limitations of MIS in open spaces is validated in vivo in an ovine model and the platform reconstructs real-time mosaics of the endoscopic scene, providing an extended and continuous visual context.

Abstract

Flexible robots hold great promise for enhancing minimally invasive surgery (MIS) by providing superior dexterity, precise control, and safe tissue interaction. Yet, translating these advantages into endoscopic interventions within open cavities remains challenging. The lack of anatomical constraints and the inherent flexibility of such devices complicate their control, while the limited field of view of endoscopes restricts situational awareness. We present a robotic platform designed to overcome these challenges and demonstrate its potential in fetoscopic laser coagulation, a complex MIS procedure typically performed only by highly experienced surgeons. Our system combines a magnetically actuated flexible endoscope with teleoperated and semi-autonomous navigation capabilities for performing targeted laser ablations. To enhance surgical awareness, the platform reconstructs real-time mosaics of the endoscopic scene, providing an extended and continuous visual context. The ability of this system to address the key limitations of MIS in open spaces is validated in vivo in an ovine model.

Figures (23)

• Figure 1: Overview of our robotic flexible endoscopy platform (A) Relevant endoscopic surgeries in open intraorgan cavities. (B) Platform configuration for the treatment of twin to twin transfusion syndrome: the electromagnetic navigation system is laterally positioned with respect to the patient, and generates a magnetic field to steer the tip of a flexible robotic endoscope. (C) During the exploration phase of the procedure the endoscope can be controlled manually using a joystick (D) while generating image mosaics in real time. (E) It can navigate automatically to targets on the image mosaic or (F) on the endoscopic image for (G) high precision vessel ablations.
• Figure 2: Design and characterization of the robotic endoscope. (A) Robotic endoscope with a flexible tip, which fits through a 10 Fr. trocar and is inserted over an advancer. (B) Flexible distal section consisting of an array of magnets and ball joints. The tip (C), and the advancer (D) containing an active driving wheel, and two passive wheels. (E) Stable positions of the endoscope's tip for a sweep in magnetic field orientations. (F) Best angles $\alpha$ for a conventional rigid endoscope and the robotic endoscope on a uterine model for a lateral insertion point during FLC. The plot on the top left depicts the percentage of the anterior surface area (above the dotted line in the uterus model on the bottom right) that is equal or above the angle thresholds for both devices.
• Figure 3: In vitro evaluation of the image-based control methods Median signal of 5 repetitions for a stepwise input direction change (first row), and for a continuous direction change (second row). (A,B) Integrated velocity of the image center. (C,D) Desired and measured angle over time. (E,F) Endoscope's tip trajectories in 3D space.
• Figure 4: In vivo manual navigation (A-D) Endoscopic images at different points in time during the in vivo manual navigation in an ovine model with the desired and the estimated image center velocity overlaid. (E) Mosaic generated from the endoscopic images during manual navigation.
• Figure 5: In vivo automated navigation (A-C) Endoscopic images of the short range automated navigation to a selected target in the endoscopic image in vivo in an ovine model with (D) the error over time. (E) Endoscopic image mosaic used as user interface for the long range automated navigation in vivo with an illustration of the trajectory to the selected target location. (F) The target is not visible on the initial endoscopic view. (G) The endoscope then moves to the initial estimate based on the magnetic field, where the target is at the border of the image and a waypoint is used for navigating closer to the target until (H) the waypoint is reached within the waypoint threshold and the selected target lies within the target proximity threshold. (I) Short-range automated navigation to reach the selected target.