Performance-aware programming for intraoperative intensity-based image registration on graphics processing units.

PURPOSE: Intensity-based image registration has been proven essential in many applications accredited to its unparalleled ability to resolve image misalignments. However, long registration time for image realignment prohibits its use in intra-operative navigation systems. There has been much work on accelerating the registration process by improving the algorithm's robustness, but the innate computation required by the registration algorithm has been unresolved. METHODS: Intensity-based registration methods involve operations with high arithmetic load and memory access demand, which supposes to be reduced by graphics processing units (GPUs). Although GPUs are widespread and affordable, there is a lack of open-source GPU implementations optimized for non-rigid image registration. This paper demonstrates performance-aware programming techniques, which involves systematic exploitation of GPU features, by implementing the diffeomorphic log-demons algorithm. RESULTS: By resolving the pinpointed computation bottlenecks on GPU, our implementation of diffeomorphic log-demons on Nvidia GTX Titan X GPU has achieved ~ 95 times speed-up compared to the CPU and registered a 1.3-M voxel image in 286 ms. Even for large 37-M voxel images, our implementation is able to register in 8.56 s, which attained ~ 258 times speed-up. Our solution involves effective employment of GPU computation units, memory, and data bandwidth to resolve computation bottlenecks. CONCLUSION: The computation bottlenecks in diffeomorphic log-demons are pinpointed, analyzed, and resolved using various GPU performance-aware programming techniques. The proposed fast computation on basic image operations not only enhances the computation of diffeomorphic log-demons, but is also potentially extended to speed up many other intensity-based approaches. Our implementation is open-source on GitHub at https://bit.ly/2PYZxQz .

Interfacing Soft and Hard: A Spring Reinforced Actuator.

Muscular hydrostats have long been a source of inspiration for soft robotic designs. With their inherent compliance, they excel in unpredictable environments and can gently manipulate objects with ease. However, their performance lacks where high force or a fast-dynamic response is needed. In this study, we propose a novel spring reinforced actuator (SRA) that explores the intermediate state between muscular hydrostats and endoskeletal mechanisms. The result is that we dramatically enhance the robot dynamic performance, which is unprecedented in similar kinds of soft robots, while retaining compliant omnidirectional bending. Analytical modeling of the flexible backbone was built and experimentally validated. This is also the first attempt to perform detailed finite element analysis to investigate the strain-stress behavior of the constraining braided bellow tube. The braided interweaving threads are modeled, in which complex thread-to-thread contacts occur. Experimental evaluation of SRAs was performed for actuation force, stiffness, and dynamic response. We showcase the enhanced actuator's performance in several applications such as locomotion and heavy object manipulation.

Electrocoiling-guided printing of multiscale architectures at single-wavelength resolution.

The rope coiling observed in liquid ink with high viscosity has been exploited in additive printing to fabricate architectures with periodically curled structures and tune their mechanical properties. However, the control over the coiling path relying on mechanical motion restricts the spatiotemporal resolution. We develop an electrically assisted high-resolution technique to manipulate coiling paths of viscous ink and structures of the deposited filament. By spatially programming the voltage applied onto the viscous ink, we show that the switching between different filament structures can be accomplished at single wavelength resolution, facilitating the rapid and accurate construction of sophisticated patterns. Furthermore, translational guiding of the electrocoiling enables rapid printing of filaments with complex structures at a line speed of 102 mm s-1. With a simplified trajectory of the printing head, large-area and multiscale patterns can be printed at an unprecedented speed; for instance, centimeter-sized architectures constructed from nanofibers with micron-sized curled structures can be completed in a few minutes. By enabling the printing of complex fiber networks with tunable shape and density, our work provides a route towards custom-design of fiber architectures with unique features such as spatially varying mechanical properties.

Experimental validation of robot-assisted cardiovascular catheterization: model-based versus model-free control.

PURPOSE: In cardiac electrophysiology, a long and flexible catheter is delivered to a cardiac chamber for the treatment of arrhythmias. Although several robot-assisted platforms have been commercialized, the disorientation in tele-operation is still not well solved. We propose a validation platform for robot-assisted cardiac EP catheterization, integrating a customized MR Safe robot, a standard clinically used EP catheter, and a human-robot interface. Both model-based and model-free control methods are implemented in the platform for quantitative evaluation and comparison. METHODS: The model-based and model-free control methods were validated by subject test (ten participants), in which the subjects have to perform a simulated radiofrequency ablation task using both methods. A virtual endoscopic view of the catheter is also provided to enhance hand-to-eye coordination. Assessment indices for targeting accuracy and efficiency were acquired for the evaluation. RESULTS: (1) Accuracy: The average distance measured from catheter tip to the closest lesion target during ablation of model-free method was 19.1% shorter than that of model-based control. (2) Efficiency: The model-free control reduced the total missed targets by 35.8% and the maximum continuously missed targets by 46.2%, both indices corresponded to a low p value ([Formula: see text]). CONCLUSION: The model-free method performed better in terms of both accuracy and efficiency, indicating the model-free control could adapt to soft interaction with environment, as compared with the model-based control that does not consider contacts.

Nonparametric Online Learning Control for Soft Continuum Robot: An Enabling Technique for Effective Endoscopic Navigation.

Bioinspired robotic structures comprising soft actuation units have attracted increasing research interest. Taking advantage of its inherent compliance, soft robots can assure safe interaction with external environments, provided that precise and effective manipulation could be achieved. Endoscopy is a typical application. However, previous model-based control approaches often require simplified geometric assumptions on the soft manipulator, but which could be very inaccurate in the presence of unmodeled external interaction forces. In this study, we propose a generic control framework based on nonparametric and online, as well as local, training to learn the inverse model directly, without prior knowledge of the robot's structural parameters. Detailed experimental evaluation was conducted on a soft robot prototype with control redundancy, performing trajectory tracking in dynamically constrained environments. Advanced element formulation of finite element analysis is employed to initialize the control policy, hence eliminating the need for random exploration in the robot's workspace. The proposed control framework enabled a soft fluid-driven continuum robot to follow a 3D trajectory precisely, even under dynamic external disturbance. Such enhanced control accuracy and adaptability would facilitate effective endoscopic navigation in complex and changing environments.