Cooperative vs. Teleoperation Control of the Steady Hand Eye Robot with Adaptive Sclera Force Control: A Comparative Study

Cooperative vs. Teleoperation Control of the Steady Hand Eye Robot with Adaptive Sclera Force Control: A Comparative Study

The paper under review focuses on the implementation and comparative analysis of cooperative and teleoperation control modes in the Steady Hand Eye Robot (SHER) 2.1, particularly targeting its application in retinal surgeries. The primary objective is to control hand tremor and minimize contact forces between the surgical tool and sclera, thereby enhancing the precision and safety of delicate retinal procedures. The implementation of a teleoperation control mode integrated with adaptive sclera force control is a novel contribution of this study, representing a significant step in robotic-assisted ophthalmic surgery.

Overview of SHER 2.1 and Control Modes

The Steady Hand Eye Robot (SHER) is a 5-DoF robotic manipulator designed for precise surgical tasks in ophthalmology. The robot employs two control modes: cooperative and teleoperation. The cooperative mode operates on an admittance control principle, where the robot filters the surgeon's hand tremor and translates their manual input into refined movements. However, this mode fails to maintain safe tool-sclera interaction forces, potentially compromising tissue integrity. The need for precise control of these forces is underscored by the anatomical constraints, such as the retinal vein's diameter (~150 µm) and the RMS value of hand tremor (~182 µm).

Conversely, the teleoperation mode, facilitated by the PHANTOM Omni haptic device, offers advanced features such as hand motion scaling and haptic feedback, which are pivotal for enhancing positioning accuracy and surgeon comfort. The crux of this study lies in integrating an adaptive sclera force control algorithm within the teleoperation mode, thereby allowing SHER to dynamically minimize scleral forces.

Implementation of Adaptive Sclera Force Control

The adaptive force control algorithm utilizes Fiber Bragg Grating (FBG) sensors to monitor and regulate the tool-sclera contact forces. The SHER end-effector's desired motion is determined by the surgeon's inputs through the haptic device, while the adaptive algorithm dynamically adjusts movements to maintain safe interaction forces below the threshold of 120 mN. This technique not only ensures patient safety but also leverages the teleoperation mode's capabilities to offer a more controlled and precise surgical experience.

Experimental Evaluation

The study conducted a series of vessel-following experiments within an eye phantom under a surgical microscope to evaluate the performance of both control modes. Metrics such as mean and maximum sclera forces, handle force/torque, and task completion time were recorded and analyzed.

• Sclera Force: Both adaptive control modes (cooperative and teleoperation) exhibited lower mean and maximum sclera forces compared to their non-adaptive counterparts. Specifically, the mean sclera force for the adaptive cooperative mode was 35.93 ± 8.72 mN, while for the teleoperation mode, it was 41.52 ± 13.40 mN. These metrics confirm the algorithm's efficacy in reducing potentially harmful contact forces.
• Handle Force/Torque: The cooperative control modes recorded higher handle force and torque values due to direct surgeon manipulation. However, in teleoperation modes, these forces were significantly lower, reflecting only the scleral forces.
• Completion Time: The cooperative modes had shorter completion times compared to the teleoperation modes, which is attributed to the direct interaction with the robot. However, the teleoperation mode's potential for motion scaling and repositioning suggests an improved surgical accuracy.

Implications and Future Directions

The findings imply that adaptive sclera force control can significantly enhance the safety and efficacy of robot-assisted retinal surgeries. The study posits that while cooperative control modes offer quicker task completion, the teleoperation modes provide superior control over surgical precision and safety with the added benefits of advanced features like motion scaling and repositioning.

Future research could explore bilateral teleoperation frameworks where both primary and secondary surgical tools employ adaptive force control. Additionally, integrating model predictive control and haptic feedback could further refine force perception and control, providing surgeons with real-time tactile cues about scleral forces. These advancements could potentially set new benchmarks in the domain of robotic-assisted ophthalmic surgery.

In conclusion, this paper contributes substantially to the understanding and implementation of adaptive control mechanisms in teleoperated surgical robots. By effectively minimizing scleral forces while maintaining high precision, SHER 2.1 represents a promising tool in overcoming the limitations posed by human hand tremor in delicate retinal surgeries.