Unified Tracking and Shape Estimation for Concentric Tube Robots

Abstract

Abstract-Tracking and shape estimation of flexible robots that navigate through the human anatomy are prerequisites to safe intracorporeal control. Despite extensive research in kinematic and dynamic modelling, inaccuracies and shape deformation of the robot due to unknown loads and collisions with the anatomy make shape sensing important for intra-operative navigation. To address this issue, vision-based solutions have been explored. The task of 2D tracking and 3D shape reconstruction of flexible robots as they reach deep-seated anatomical locations is challenging, since the image acquisition techniques usually suffer from low signal-to-noise ratio (SNR) or slow temporal responses. Moreover, tracking and shape estimation are thus far treated independently despite their coupled relationship. This paper aims to address tracking and shape estimation in a unified framework based on Markov Random Fields (MRF). By using concentric tube robots as an example, the proposed algorithm fuses information extracted from standard monoplane X-ray fluoroscopy with the kinematics model to achieve joint 2D tracking and 3D shape estimation in realistic clinical scenarios. Detailed performance analyses of the results demonstrate the accuracy of the method for both tracking and shape reconstruction.

3D Shape Estimation Of Tendon-Driven Catheters Using Ultrasound Imaging

“…The recent advances in 3D shape estimation are based on three categories of alternative emerging techniques [19], [23]: fiber-optic-sensor (FOS) 2 [25], [28], electromagnetic (EM) tracking [29] and intraoperative imaging modality [24], [30]- [31]. In FOS-based shape reconstruction method, multiple Bragg Grating Sensor fibers (typically 3-4) are arranged around the circumference of the catheter at fixed distance about the neutral axis of bending [32] to measure strain, force, torque, displacement, temperature and pressure [20], [33].…”

“…In FOS-based shape reconstruction method, multiple Bragg Grating Sensor fibers (typically 3-4) are arranged around the circumference of the catheter at fixed distance about the neutral axis of bending [32] to measure strain, force, torque, displacement, temperature and pressure [20], [33]. Although FOS/FBG-based shape sensing methods can provide accurate and fast measurements of the shape of tendon-driven catheters, their integration with small and hyper flexible catheters is challenging and adds significant cost [24], [34].…”

“…Other Kalman filter assumptions that may contribute to low accuracy are the inherent constant dynamics and zero mean Gaussian noise model which do not hold in real world processes 24. The in this equation refers to the matrix transpose mathematical operation.…”

Shape Sensing Based on Longitudinal Strain Measurements Considering Elongation, Bending, and Twisting

“…Commonly used measurement systems use imaging technology to reconstruct the shape. However, these require either a direct line of sight to the robot [6] or rely on fluoroscopy [7] which requires elaborate and expensive equipment. Electromagnetic tracking coils integrated into the robot allow measuring the position of distinct points along the robot's centerline [8].…”

Combining Differential Kinematics and Optical Flow for Automatic Labeling of Continuum Robots in Minimally Invasive Surgery

“…As explained before, mechanical models for continuum robots are inaccurate and error-prone. They can nevertheless be combined with image data for enhancing the quality of pose estimation (Tran et al, 2017 ; Vandini et al, 2017 ) or tracking (Pezzementi et al, 2009 ; Reiter et al, 2014 ). Similarly to approaches purely based on images, current methods making use of kinematic data typically build over image segmentation techniques, using either machine learning methods (with hand-labeled images) or manual initialization.…”