The Focused Assessment with Sonography in Trauma (FAST) exam is a critical tool for initial screening and triage of trauma cases. The need to offer this exam despite the lack of access to experienced sonographers in the field motivates this work. While past works have attempted to address automating the FAST exam, the solutions offered do not account for the chaotic nature of field triage with multiple medics working on the patient simultaneously. This paper offers a solution whereby custom low-cost robotic skins can be added to any lightweight robot to allow the robot to operate safely around medics. The skin offers contact detection and localization, along with an added padding for passive safety. A formulation of the contact detection, force sensing, and localization is presented by leveraging solutions of the rigid point-based registration problem, first-order stiffness modeling, and an active compliance control strategy. A high-level active compliance redundancy resolution strategy is presented and verified to allow users to safely move the body of the robot out of the way while the robot maintains a high-quality ultrasound image of a region of interest. These results present a first step toward a mobile deployable ultrasound triage platform for FAST.
This study investigates factors influencing microsuture needle bending in robotic surgery. An experiment was conducted by varying needle shape, curvature, tissue elasticity, suturing velocity, and suturing patterns. Regression analysis and torque measurements revealed that curvature radius and its interaction with needle shape significantly affect deflection. The dominant factors differed across suturing patterns due to variations in tissue contact conditions. These findings support the development of deflection-aware needle designs and control strategies for safe automated suturing.
This work presents a first demonstration of a nonanthropomorphic, underactuated exoskeleton architecture that uses a single linear actuator per leg to transfer loads directly to the ground next to the wearer’s foot. This allows the exoskeleton and user to move somewhat independently, facilitating normal kinematics during walking. We also present the control system that employs inertial measurement units (IMUs) affixed to the exoskeleton and shoes to estimate gait parameters. The performance of the exoskeleton is assessed based on its ability to touch down to the ground and lift off from the ground at the start of the stance and swing phases, respectively, as a participant walks on a treadmill at five speeds (0.67–1.34[Formula: see text]m/s). The exoskeleton exhibited a touchdown delay of 0.093 ± 0.019s and a lift-off delay of 0.007 ± 0.018s, resulting in total support percentages of 88.9 ± 2.9% during the stance phase across all speeds and the best performance at 0.67[Formula: see text]m/s. Other experiments showed that our gait percentage estimation during steady state walking achieved a pooled root-mean-square error (RMSE) of 2.94%, which is less than prior works. This proof-of-concept architecture highlights a promising direction for lightweight, low-complexity exoskeletons for load carriage or bodyweight support.
Endoscopic Sinus Surgery (ESS) suffers from significant targeting inaccuracies due to elastic deformation of the rigid endoscope during vision-based surgical procedures. Current vision-based image-guided surgery systems assume rigid-body behavior, neglecting clinically significant deflections caused when surgeons use the endoscope shaft as a fulcrum against anatomical structures. To address this fundamental limitation, we present a novel real-time deformation compensation system featuring a compact, instrumented sleeve that mounts proximally on the endoscope shaft, external to the surgical field. The sleeve integrates six strain gauges distributed at two axial locations and employs a hybrid compliant mechanism design,combining flexure elements with lever-amplification mechanisms,to precisely capture multi-axis bending. We evaluated four mapping algorithms, including polynomial regression and machine learning approaches, to optimize displacement prediction. Our system reduced displacement error from [Formula: see text] [Formula: see text]mm to [Formula: see text] [Formula: see text]mm RMSE, achieving 85% improvement in target localization accuracy. This work introduces a practical, cost-effective solution that addresses a fundamental source of error in endoscopic navigation without disrupting established surgical workflows.
This study presents a modeling and control framework for a retinal microsurgical robot, the Improved Integrated Robotic Intraocular Snake (I2RIS), which accounts for the coupling between pitch and yaw degrees of freedom. The system is modeled using a multi-input multi-output (MIMO) mass-spring-damping (MSD) formulation, with parameters and states jointly estimated through a dual-extended Kalman filter (dual-EKF) approach. Unlike analytical methods such as Cosserat-based models, the proposed approach is computationally efficient, requires minimal training data, and enables real-time control. The coupled MSD model captures the interdependence between pitch and yaw dynamics using experimentally optimized parameters. An optimal stochastic controller, the Model Predictive Path Integral (MPPI), is then designed and compared with a Linear Quadratic Regulator (LQR) in a trajectory-tracking control problem. Experimental results demonstrate that MPPI achieves superior performance and robustness in controlling the coupled dynamics of the I2RIS robot, offering a promising solution for efficient real-time control of snake-like robotic systems.
Intraocular microsurgery requires submillimeter precision within an extremely confined and delicate anatomical workspace. Cable-driven continuum robots such as Improved Integrated Robotic Intraocular Snake (I 2 RIS) offer the necessary dexterity but exhibit nonlinear hysteresis, complicating accurate control and localization. To enhance control precision, vision-based localization methods can be incorporated to provide external feedback and to compensate for modeling uncertainties. To support the development and quantitative evaluation of such vision-based approaches, modular, and open-source simulation framework is established, replicating an ophthalmic surgical scene that includes the eyeball model, the I 2 RIS continuum robot, and a calibrated surgical microscope. This environment enables automated and scalable acquisition of data that includes synchronized RGB-D images and corresponding ground-truth 6D pose data under diverse lighting, texture, and background conditions. Using the generated dataset, we conduct representative experiments on geometry-driven 6D pose tracking and appearance-based sim-to-real detection to evaluate its applicability for vision-based localization tasks. In addition, a preliminary domain gap analysis is performed using structure-based image similarity metrics, including Canny edge statistics and structural similarity (SSIM), to quantitatively assess visual alignment between simulated and real microscope images. The resulting dataset serves as a consistent evaluation resource for debugging, training, and assessing localization algorithms in intraocular continuum robotics, supporting reproducible research and sim-to-real generalization studies.
This paper presents a computational model, based on the Finite Element Method (FEM), that simulates the thermal response of laser-irradiated tissue. This model seeks to address a gap in the current ecosystem of surgical robot simulators, which generally lack support for lasers and other energy-based end effectors. In the proposed model, the thermal dynamics of the tissue are calculated as the solution to a heat conduction problem with appropriate boundary conditions. The FEM formulation allows the model to capture complex phenomena, such as convection, which is crucial for creating realistic simulations. The accuracy of the model was verified via benchtop laser-tissue interaction experiments using agar tissue phantoms and ex-vivo chicken muscle. The results revealed an average Root-Mean-Square Error (RMSE) of less than 1.5 ∘ C across most experimental conditions.
In this work, we develop a collaborative path-planning method for robotic soft-tissue surgery. The proposed method aims to fill the gaps between current state-of-the-art teleoperation surgery and ideal fully autonomous surgery, which is the long-term goal of surgical robots. Via the proposed method, a human user issues high-level path planning commands that the surgical robot can autonomously execute. Specifically, a human user selects key points on a target tissue via a haptic device and a multi-camera 3D sensing and overlay system. An autonomous path-planning and filtering method then assists the user in completing the path by providing uniformly distributed waypoints on the surface of the tissue that traverse through the key points and avoid undesired regions. We also develop supplementary force feedback and visual cues to further improve the effectiveness of the proposed method by helping the user detect the 3D surface of the tissue more easily. Our results via a human subject study indicate that compared to the unassisted path generation method, the proposed method can reduce the error in spacing between the waypoints by 67.6%, the probability of making path corrections by 33.3%, the completion time by 40.7%, and the perceived workload by 30.5%.
Recognizing surgical gestures in real time is critical for automated activity recognition, skill assessment, and surgical assistance. The current robotic surgical systems provide us with rich multi-modal data such as video and kinematics. While some recent works in multi-modal neural networks learn the relationships between vision and kinematics data, current approaches treat kinematics information as independent signals, with no underlying relation between tool-tip poses. However, instrument poses are geometrically related, and the underlying geometry can aid neural networks in learning gesture representation. Therefore, we propose integrating motion invariant signals, arc length, dual angle, curvature, and torsion, with vision and kinematics using a relational graph network to capture the underlying relations between different data streams. We show that gesture recognition improves when combining motion invariant signals, achieving 92.9% frame-wise accuracy on Suturing. Our results show that motion invariant signals combined with positional data provide more interpretable representations than conventional position and quaternion signals. These findings highlight the value of geometric-aware modeling of kinematics for surgical gesture recognition and suggest that motion invariant signals can improve generalization across users.
Virtual simulations have served as important tools for designing and testing robotic frameworks in various domains. Increasingly, they have been used to train and experiment in surgical robotics applications. Having seen the success of an open simulation environment for robot-assisted surgical suturing in the Asynchronous Multi-Body Framework (AMBF), we sought to reproduce the surgical suturing environment from AMBF in NVIDIA Isaac Sim, a simulator with impressive visual fidelity and GPU-accelerated computation capabilities. We present a virtual scene for robotic suturing comprising two robotic manipulators, an endoscope, two surgical phantoms and a needle attached to a thread. The thread is modeled using three different techniques. This work also provides interactive manipulation of the virtual scene via haptic input devices using the Robot Operating System interface. Although Isaac Sim possesses impressive and realistic visuals, AMBF outperforms in speed by up to 15 frames per second, depending on the task. AMBF also responded faster than Isaac Sim by approximately 0.1 s when using a haptic input device. In this context, our objective is to provide a comprehensive overview of the capabilities and limitations of NVIDIA Isaac Sim in the surgical robotics challenge relative to AMBF.
Accurate surgical activity recognition from surgical video can support intraoperative decision-making, post-operative analysis, and surgical education. However, complex and variable workflows across different procedures challenge the generalization of current models. Consistently learning temporal dependencies at the frame-level, activity-level, and case-level across different procedure types remains an open problem. Here, we apply the existing Frame-Action Cross Attention for Temporal modeling (FACT) architecture to the surgical domain, since it was designed for this purpose. We evaluate its performance across three datasets: Cholec80 (cholecystectomy, 7-phase), AutoLaparo (hysterectomy, 7-phase), and MultiBypass140 (gastric bypass, 12-phase and 46-step). To distinguish between the relative importance of spatial and temporal representation learning, we investigate both, domain-general and surgically fine-tuned image feature extractors with FACT which jointly reasons over frame-level and action-level dependencies across the entire case via bidirectional cross-attention. To better understand workflow variability in our datasets, we model the variance in surgical workflows across these datasets using metrics. We compare FACT, which processes an entire case as a single sample, against multiple other architectures, including traditional windowed methods to evaluate how different approaches perform across these datasets. Across Cholec80, AutoLaparo, and MultiBypass140, FACT delivers competitive performance under matched protocols including 94.1% accuracy on Cholec80 phases, 77.3% on MultiBypass140 steps, and 89.5% accuracy on MultiBypass140 phases. Importantly, FACT performs comparably when using either domain-general or surgically-finetuned spatial representations, in some cases, even being relatively robust to older generations of domain-general image feature extractors, like RotNet. We emphasize method clarity, cross-dataset consistency, and workflow-variability analysis as our key contributions. While our results suggest that FACT is a flexible and robust temporal architecture for surgical activity modeling, challenges remain in generalizing to finegrained or highly variable workflows. We also include a discussion of how the performance of different architectures may be affected by the measured variance in surgical workflows and propose future work that may help close this gap.
Bleeding from noncompressible, penetrating, deep wounds is a significant healthcare challenge. Failure to control bleeding in these wounds is primarily due to inability to tamponade bleeding by applying even pressure to inaccessible bleeding blood vessels. We have designed a device intended for rapid treatment of complex non-compressible wounds: the Rapid Everting Tamponade (RET) is designed to apply internal pressure to affected vessels through the everting growth of a flexible, inflated tube so that the tube quickly grows into any wound shape and applies even pressure to the wound cavity. We investigate the optimal sizing of an inelastic tube for a simulated knife wound as a design trade-off between open-air burst pressure and minimum required eversion pressure and calculate a safety factor and nominal eversion pressure for each tube size. The candidate tubes were everted into a rigid phantom of the simulated knife wound at the nominal eversion pressure and their ability to stop pulsatile blood flow near healthy human diastolic pressures was measured with a new, low-cost, load cell-based flow measurement system. Rectangular tubes with 40 and 50 mm diameters, with areas 3 and 5 times that of the wound area, respectively, were found to have a sufficient safety factor above 1.75 and resulted in minimal blood loss. In a silicone knife wound phantom, bleeding was stopped by 30, 40, and 50 mm diameter tubes in all trials. Not only has eversion been shown to be a rapid and effective means of applying internal pressure to non-compressible wounds, this work also characterizes the increased minimum eversion pressure when an everting tube size exceeds the size of a lumen.
This study develops a process for evaluating the impact of hardware design parameters on the performance of a stylet embedded with a multicore fiber (MCF) for shape sensing, to be used to guide the insertion of an interstitial brachytherapy needle. The MCF consists of seven cores (one central and six outer), with each core inscribed with fourteen fiber Bragg gratings (FBGs), called active areas (AAs). Hardware performance was evaluated using two datasets from distinct constant-curvature jigs. First, the influence of the number and spacing of AAs along the fiber on reconstruction accuracy was evaluated, which identified the AA configuration that yielded the lowest reconstruction errors. Channel configurations of seven-core and four-core fibers were analyzed similarly. Finally, recognizing that AA and channel performance are not entirely independent, a joint analysis was conducted to determine the globally optimal configuration of the stylet. Reconstruction with the optimized stylet achieved a reduction in tip error of 34% relative to the full-sensor configuration, a difference that is statistically significant ( α = 0.05 , p = 0.00005 ) for the designated calibration and validation dataset. This work provides practical guidance for MCF selection and establishes a new framework for post-fabrication optimization of multicore fibers.
High-dose-rate (HDR) brachytherapy is an effective treatment for prostate and gynecologic cancers, yet current rigid catheter systems constrain trajectories to straight paths, limiting conformity to patient-specific anatomy and compromising dose distribution. To address this, we developed and preliminarily evaluated a semi-automated robotic platform for curvilinear catheter implantation. The system integrates (i) a handheld tendon-driven active needle manipulator with joystick-based control, (ii) an innovative template for initial angulation, and (iii) a novel gynecologic applicator incorporating concentric tube robotics for steerable tandem delivery. Patient-specific curvilinear trajectories were preplanned from anonymized MRI/CT datasets, and experimental evaluations were conducted in air and tissue-mimicking phantoms. The handheld manipulator achieved tip displacement errors below 2.89 mm in air and 2.22 mm in phantom, while the combined device-template system reproduced prostate curvilinear trajectories with root mean squared deviations (RMSDs) under 3.67 mm. Integration of the concentric tube robot with the gynecologic applicator achieved uterine trajectories with RMSDs below 1.17 mm. These results demonstrate, for the first time, the feasibility of semi-automated active needle manipulation and concentric tube-assisted applicator design for anatomically conformal HDR brachytherapy. This approach has the potential to improve catheter placement accuracy, enable patient-specific dose escalation, and reduce toxicity to organs at risk in both prostate and cervical cancer treatment.
Focal prostate treatment (aka “male lumpectomy”) has the potential to reduce invasiveness for prostate cancer patients. However, clinical deployment has been impeded by the difficulty of performing surgery through the urethra, which we hypothesize relates to both instrument dexterity and visualization limitations. To evaluate this hypothesis, in this paper we propose a system consisting of an endoscopic robot to enhance dexterity and an image-guidance display updated periodically during surgery based on MRI images. To evaluate the system, in this paper we compare four conditions: unaided manual resection using an endoscope, robot-aided surgery without image guidance, image guidance without the robot, and both robot and image guidance together. We find that while the robot and image guidance improve performance individually, the combination of the two provides the greatest improvement.
Soft tissue simulation can play an essential role in the automation of robotic surgery by providing contextual information during surgery and generating datasets for training. Any time tissue deformations are simulated, computational speed, accuracy, and stability are key concerns. State-of-the-art tissue simulation resolves inertial dynamics solutions using position-based computational methods. However, existing methods fail to efficiently resolve steady-state solutions at surgical size scales because of transient inertial dynamics and the small time step required for stability at such size scales. We propose a position-based tissue simulation framework which is based on large-deformation Neo–Hookean elasticity and enables fast resolution to steady-state for efficient simulation. Our method replaces the inertial terms in the model with a virtual viscous damping term. This enables realistic tissue motion while eliminating the transient vibrations that require more computation. It also enables smooth and stable dynamic transitions between disparate static states. Using our method, we develop an interactive simulator capable of stable, real-time tissue manipulation with a deformable concentric tube robot (CTR) model. Stable collision and simulator realism are achieved through the inclusion of local iterations of collision areas and a novel hydrostatic strain energy formulation.
This paper presents an automated, contactless posterior eye examination system that can perform nonmydriatic retinal examinations at safe distances from unconstrained individuals. The system operates in a dimly lit room, tracking the patient’s face and anatomy using infrared cameras, collecting screening-quality visible light images using flash photography when the patient’s pupil is aligned. The sensor system is mounted on a robot arm, which tracks the center of the patient’s head motion, locks onto the pupil, and captures images of the retina. The behavior control system completes an examination of both eyes and ensures that images are captured with minimal artifact. Feasibility studies on a phantom and users from the research team indicate that the system completes screening-quality bilateral retinal imaging in under 2[Formula: see text]min.
Cervical cancer accounts for a significant portion of the global cancer burden among women. Interstitial brachytherapy (ISBT) is a standard procedure for treating cervical cancer; it involves placing a radioactive source through a straight hollow needle within or in close proximity to the tumor and surrounding tissue. However, the use of straight needles limits surgical planning to a linear needle path. We present the OncoReach stylet, a handheld, tendon-driven steerable stylet designed for compatibility with standard ISBT 15- and 13-gauge needles. Building upon our prior work, we evaluated design parameters like needle gauge, spherical joint count and spherical joint placement, including an asymmetric disk design to identify a configuration that maximizes bending compliance while retaining axial stiffness. Free space experiments quantified tip deflection across configurations, and a two-tube Cosserat rod model accurately predicted the centerline shape of the needle for most trials. The best performing configuration was integrated into a reusable handheld prototype that enables manual actuation. A patient-derived, multi-composite phantom model of the uterus and pelvis was developed to conduct a pilot study of the OncoReach steerable stylet with one expert user. Results showed the ability to steer from less-invasive, medial entry points to reach the lateral-most targets, underscoring the significance of steerable stylets.
Magnetic resonance imaging (MRI)-guided transperineal targeted prostate biopsy has potential to provide precise cancer diagnosis. However, it is challenged by targeting errors and needle deflection. The main issue is that the traditional needle guides have limited degrees of freedom (DoF) and therefore restrict insertion path options. Thus, the targeting error can be larger than 5 mm. This study evaluates the performance of a custom-built 4-DoF needle guiding device designed to enable alternative insertion paths. The 4-DoF Smart Template was designed to improve image quality and stability, including redesigned fiducials and titanium positioning frames. Four MRI-guided biopsy cases were conducted. Targets were identified on T2-weighted images, and insertion paths were planned and adjusted to avoid critical structures. The targeting error was measured as the in-plane distance between the target and the actual needle position. The average targeting error of the best insertion attempt was 3.7mm±2.6mm (mean ± standard deviation). When discarding the case where the device had to be removed due to patient anatomy, the average targeting error was reduced to 2.7mm±1.1mm. No adverse events were reported. The new 4-DoF Smart Template device has demonstrated improved accuracy as compared to the traditional needle guide device, successfully reaching the target location with just one insertion in four out of six targets.
This work presents the design, development, and evaluation of a continuum robot for neurovascular surgery (CRONOS), addressing the lack of commercially available robots specifically designed for neurointervention. As the field evolves, assessing the feasibility of robot-assisted procedures, particularly for remote stroke treatment, is crucial. Designed for research purposes, CRONOS aims to explore the feasibility of such procedures. It features a trajectory planner, an open-source 3D-printed design, and independent control of three devices, providing four degrees of freedom. The system enables millimeter control and remote operation via a client–server network. To assess functionality, we simulated three stroke cases using a virtual system, categorizing data by operation mode: manual, robotic training, and robotic experiments. The robot was locally controlled via CAN bus during training and remotely operated in the experimental phase. Key performance metrics, including procedure success rate, device translation, and fluoroscopy time (FT), were compared to manual operations. The robotic system significantly reduced device translation, with robotic experiments showing a 52.2% decrease compared to manual mode. FT showed variability across modes, though manual procedures had the lowest values, with only a 9.7% difference from robotic experiments. These findings suggest that robot-assisted neurovascular interventions can enhance procedural control while maintaining comparable FT outcomes.
