Data-Driven Shape Sensing in Continuum Manipulators via Sliding Resistive Flex Sensors (2311.18154v1)

Abstract: We introduce a novel shape-sensing method using Resistive Flex Sensors (RFS) embedded in cable-driven Continuum Dexterous Manipulators (CDMs). The RFS is predominantly sensitive to deformation rather than direct forces, making it a distinctive tool for shape sensing. The RFS unit we designed is a considerably less expensive and robust alternative, offering comparable accuracy and real-time performance to existing shape sensing methods used for the CDMs proposed for minimally-invasive surgery. Our design allows the RFS to move along and inside the CDM conforming to its curvature, offering the ability to capture resistance metrics from various bending positions without the need for elaborate sensor setups. The RFS unit is calibrated using an overhead camera and a ResNet machine learning framework. Experiments using a 3D printed prototype of the CDM achieved an average shape estimation error of 0.968 mm with a standard error of 0.275 mm. The response time of the model was approximately 1.16 ms, making real-time shape sensing feasible. While this preliminary study successfully showed the feasibility of our approach for C-shape CDM deformations with non-constant curvatures, we are currently extending the results to show the feasibility for adapting to more complex CDM configurations such as S-shape created in obstructed environments or in presence of the external forces.

• The paper presents a novel method for precise shape sensing in continuum manipulators using sliding resistive flex sensors, achieving an average error of 0.968 mm.
• The authors integrated the sensors into a 3D-printed CDM and calibrated the data with a ResNet framework, resulting in an R² value of 99.8% for accurate shape reconstruction.
• These results offer a cost-effective, robust alternative for minimally-invasive robotic surgery, potentially reducing operational costs and improving clinical feasibility.

Data-Driven Shape Sensing in Continuum Manipulators via Sliding Resistive Flex Sensors

The paper "Data-Driven Shape Sensing in Continuum Manipulators via Sliding Resistive Flex Sensors" by Zhang et al. introduces a novel method for shape sensing in cable-driven Continuum Dexterous Manipulators (CDMs) utilized in minimally-invasive surgery. The method employs resistive flex sensors (RFS) that slide inside the CDM, capturing resistance metrics from various positions to enable accurate shape estimation. Despite their lower cost and robustness, these sensors offer accuracy near that of more expensive alternatives.

Methodology

The authors developed a mechatronic system integrating RFS units into a three-times scaled 3D-printed CDM prototype. The CDM, characterized by its degrees of freedom in one plane and designed for stability in minimally-invasive procedures, includes channels accommodating the RFS units that record deformation data. This data is further refined through an overhead camera system and a machine learning-based calibration framework (ResNet). The design features ensure the real-time positional tracking of each CDM joint through extensive data collection and processing, resulting in a comprehensive dataset used for training and evaluation.

Experimental Setup

Experiments were meticulously structured, involving a series of trials controlling the CDM’s deflection across various configurations. Using a joystick to manipulate linear step motors, the CDM’s bending and positional metrics were recorded by the embedded RFS and optical rotary encoders. Calibration and validation utilized the intrinsic and extrinsic parameters of an Intel RealSense D435 camera, yielding an accurate 2D reconstruction of the CDM shape.

Results and Discussion

Data indicated an average shape estimation error of 0.968 mm with a standard error of 0.275 mm, and a response time of 1.16 ms. These metrics highlight the feasibility of the proposed RFS-based shape sensing method. Notably, the authors employed a ResNet architecture for sensor calibration, achieving robust model performance, as evidenced by an $R^2$ value of 99.8%, denoting high predictive accuracy. The CDM shape reconstruction closely aligned with ground-truth data, confirming the method’s precision and reliability.

While the methodology demonstrated potential, some limitations persisted. The technique's reliance on scanning discrete joint positions led to certain inaccuracies, which future automation in sensor scanning could mitigate. Additionally, adapting the ResNet model to accommodate dynamic input could enhance real-time shape adjustment capabilities. Further research into the miniaturization of the RFS unit could extend application feasibility to smaller continuum manipulators, crucial for more delicate minimally-invasive surgeries.

Implications and Future Work

The practical implications of this work are substantial. By providing an accurate, cost-effective alternative to existing shape sensing technologies, like Fiber Bragg Grating (FBG) sensors, for CDMs, this method significantly reduces operational costs. This cost-effectiveness supports the transition of CDMs to broader clinical applications, where their use has been limited by expense and sensor sensitivity to environmental changes. Notably, RFS units are not affected by temperature variations, enhancing reliability during procedures involving hard tissue manipulation.

Theoretically, the integration of data-driven techniques with traditional sensing mechanisms paves the way for innovations in robotic surgery. Future work could see the adaptation of the RFS units for more complex CDM configurations, continuous scanning capability, and improvements in sensor calibration models to accommodate dynamic, real-time feedback. The pursuit of these avenues holds promise for increasingly sophisticated and precise applications in minimally-invasive surgical robotics.

In conclusion, the paper successfully establishes an innovative approach to shape sensing in CDMs using sliding resistive flex sensors, demonstrating significant potential for enhancing the precision and reliability of minimally-invasive surgical procedures. This work represents a meaningful advancement in the field of robotic surgery, promising both practical and theoretical contributions to the broader domain of continuum robotics.