Cancer, with high morbidity and high mortality, is one of the major burdens threatening human health globally. Intervention procedures via percutaneous puncture have been widely used by physicians due to its minimally invasive surgical approach. However, traditional manual puncture intervention depends on personal experience and faces challenges in terms of precisely puncture, learning-curve, safety and efficacy. The development of puncture interventional surgery robotic (PISR) systems could alleviate the aforementioned problems to a certain extent. This paper attempts to review the current status and prospective of PISR systems for thoracic and abdominal application. In this review, the key technologies related to the robotics, including spatial registration, positioning navigation, puncture guidance feedback, respiratory motion compensation, and motion control, are discussed in detail.

Cancers are a major contributor to disease burden worldwide, and projections forecast that global cancer burden will continue growing in the near decades.1 Routine screening and early diagnosis as well as early treatment have become an effective way to prevent cancer, greatly reducing the death risk of patients with malignant tumors.2 Currently, puncture intervention surgery plays an important role in tumor treatment and the main operating treatments include puncture ablation, particle implantation, and biopsy sampling.3–5 In the traditional puncture interventional surgery, clinicians obtain the location of patient's lesion according to the preoperative medical images and determine the entry point of the needle and the patient's bed position. During the operation, the clinicians need to perform several CT scans to confirm the insertion depth and the current insertion needle position.6 Then, the needle is adjusted by comparing the target position and the current needle tip's location, which is repeated until the target is reached.

In recent decades, puncture interventional surgical robot (PISR) systems, primarily performing puncture procedures, have shown great potential in precise positioning, visualization imaging, and automatic control.7 In general, PISR systems consist of planning information generating module, navigation control module, sensing module, execution module, and central scheduling module. The operation of PISR systems generally includes three stages: modeling, planning, and execution. In the modeling stage, CT, MRI, and other medical image information of the lesion is collected by medical imaging equipment. Then, the 3D reconstruction is performed and presented to clinicians for identifying the lesion targets more accurately. In the planning stage, the spatial systems of robot, image, and patient are coordinated and registered before the path is planned. In the execution stage, the manipulator moves to the recommended entry point according to the planned path. The whole process is shown in Figs. 1(a)–1(c). However, due to the lack of real-time positioning of the needle tip, the interventional clinicians need at least twice CT scans to adjust the position of needle to reach the target location and the insertion process of needle relies on the clinicians' experience. Based on the application scenario, PISR systems can be categorized into orthopedic robotic surgery systems and solid tumor robotic surgery systems. This means that the patient's respiratory motion and natural movements intraoperatively hugely affect the positioning accuracy and puncture effect,8 especially for solid tumor robotic surgery. Additionally, the advancement of needle in flexible tissues, such as liver and stomach, is subjected to uneven friction force and cutting force, causing positioning accuracy errors.9 These issues hinder the further development of PISR and should be tackled with intelligent and reliable techniques.

FIG. 1.

General procedures of robot-assisted puncture interventional surgery, including (a) preoperative CT scanning, (b) intraoperative navigation, and (c) needle insertion, reprinted with permission from https://www.hanglok-tech.cn/ (last accessed January, 2024). Copyright 2024 Hanglok-Tech Co. Ltd.

FIG. 1.

General procedures of robot-assisted puncture interventional surgery, including (a) preoperative CT scanning, (b) intraoperative navigation, and (c) needle insertion, reprinted with permission from https://www.hanglok-tech.cn/ (last accessed January, 2024). Copyright 2024 Hanglok-Tech Co. Ltd.

Close modal

Generally, the primary tasks of PISR involve path planning, needle pose adjusting, needle advancing, etc.10–12 However, the corresponding technologies for these tasks are still immature and need to be further progressed, such as spatial registration, positioning and navigation, real-time feedback, respiratory gating, and motion control. In this context, this paper systematically summarized the evolution of commercial PISR systems with a focus on the thoracoabdominal region. Then, the aforementioned technology development are discussed, including spatial registration, positioning navigation, and puncture guidance feedback technology.

The organization of this paper is as follows: Sec. I introduces the research background. Sec. II gives the classification of the representative PISR systems according to the level of automation. Then, the research progress of spatial registration and positioning navigation, puncture guidance feedback, respiratory motion compensation, and motion control are analyzed in Secs. III and IV, followed by the future direction of PISR systems in Sec. V.

At present, a series of PISR systems have appeared in clinical applications,13 which are introduced sequentially. Considering the coherence and clarity of this paper, the level of automation in positioning and needle insertion is adopted as the classification criterion for different parts in this section. Specifically, the level of automation in positioning and needle insertion is roughly divided into three categories in this paper: manual positioning and manual needle insertion; automatic positioning and manual needle insertion; and automatic positioning and teleoperation for needle insertion. It is noteworthy that, in this paper, tracking refers to the process of moving the end-tip guiding sleeve according to the planned path and needle placement refers to the process of advancing the needle.

The robot mentioned in this subsection could plan the puncture needle path, while the path tracking and needle advancing both are manually controlled by the clinicians, as shown in Table I. In this paper, the puncture accuracy is defined as the maximum among the stereotactic accuracy, puncture navigation accuracy, and lesion targeting accuracy. This kind of robot registers the current setup mainly based on the preoperative image information as well as the position information of optical or electromagnetic sensing, before the clinicians manually locating the end-tip guiding sleeve to the entry point assisted by 3D navigation software. Recently, two distinct skeleton graph-based non-rigid registration approaches were proposed using sparse key points14 and dense graphs,15 respectively, to precisely map preplanned trajectories in intercostal space from preoperative to the current setup.

TABLE I.

Representative products for PISR.

Year System RD institutions Country Imaging equipment DOF Level of automation Clinical cases
2014  CAS-One IR  CAScination AG.  Switzerland (CH)  CT/MRI  Manual Tracking and Manual Needle Placement  Biopsy 
Ablation 
2017  IQQA-Guide MII  EDDA Tech Inc.  America(US)  CT/MRI  Biopsy 
Ablation 
2021  Savior-L  Accumed Tech Ltd.  China (CN)  CT  Biopsy 
Ablation 
2021  SGNNO  Hanglok-Tech Co. Ltd.  China (CN)  CT  Biopsy 
Ablation 
2019  Needle Placement System (NPS)  DEMCON Adcanced Mcchatronics  Netherland (NL)  CT  Automatic Tracking and Manual Needle Placement  Biopsy 
Ablation 
2016  Micromate  Interventional Systems  Austria (AT)  CT  Biopsy 
Ablation 
2013  ROBIO EX  Perfint healthcare  India (IN)  CT  Biopsy 
Ablation 
2014  MAXIO  Perfint healthcare  India (IN)  CT  Biopsy 
Ablation 
2020  TH-S1  True Health (Guangdong Hengqin) Medical Technology Co. Ltd.   China (CN)  CT  Biopsy 
Ablation 
2021  Epione  Quantum Surgical  France (FR)  CT  Biopsy 
Ablation 
2022  SGNIO  Hanglok-Tech Co. Ltd.  China (CN)  CT  Biopsy 
Ablation 
2021  XACT ACETM  XACT Robotics  Israel (IL)  CT  Automatic Tracking and Automatic Needle Placement  Biopsy 
Ablation 
2022  VINCENT  Hanglok-Tech Co. Ltd.  China (CN)  CT  Seeds implantation 
Year System RD institutions Country Imaging equipment DOF Level of automation Clinical cases
2014  CAS-One IR  CAScination AG.  Switzerland (CH)  CT/MRI  Manual Tracking and Manual Needle Placement  Biopsy 
Ablation 
2017  IQQA-Guide MII  EDDA Tech Inc.  America(US)  CT/MRI  Biopsy 
Ablation 
2021  Savior-L  Accumed Tech Ltd.  China (CN)  CT  Biopsy 
Ablation 
2021  SGNNO  Hanglok-Tech Co. Ltd.  China (CN)  CT  Biopsy 
Ablation 
2019  Needle Placement System (NPS)  DEMCON Adcanced Mcchatronics  Netherland (NL)  CT  Automatic Tracking and Manual Needle Placement  Biopsy 
Ablation 
2016  Micromate  Interventional Systems  Austria (AT)  CT  Biopsy 
Ablation 
2013  ROBIO EX  Perfint healthcare  India (IN)  CT  Biopsy 
Ablation 
2014  MAXIO  Perfint healthcare  India (IN)  CT  Biopsy 
Ablation 
2020  TH-S1  True Health (Guangdong Hengqin) Medical Technology Co. Ltd.   China (CN)  CT  Biopsy 
Ablation 
2021  Epione  Quantum Surgical  France (FR)  CT  Biopsy 
Ablation 
2022  SGNIO  Hanglok-Tech Co. Ltd.  China (CN)  CT  Biopsy 
Ablation 
2021  XACT ACETM  XACT Robotics  Israel (IL)  CT  Automatic Tracking and Automatic Needle Placement  Biopsy 
Ablation 
2022  VINCENT  Hanglok-Tech Co. Ltd.  China (CN)  CT  Seeds implantation 

In the navigation framework combined with optical sensing, the CAS-One IR abdominal puncture surgery navigation system of CAScination AG, Switzerland, is as a commercial representative. With the assistance of CAS-One IR, the clinicians manually match the end-tip guiding sleeve to the entry point of the needle based on the planned trajectory. During the operation, the clinicians manually hold the needle, the accuracy of which is impacted by the issues of hand tremor and fatigue. In addition, due to the lack of real-time positioning information of the end of the needle in the human body, the system cannot accurately get the precise position intraoperatively. As a solution to this problem, the navigation framework with electromagnetic sensing could be considered, enabling the end of the needle to position by sensing devices in real time. However, the physical device makes it inconvenient to be integrated with the multi-degree of freedom manipulators and is extremely susceptible to the influence of magnetic fields. This causes that the corresponding systems are unable to accurately locate the end of the needle in vivo, such as IQQA-Guide MII, Savior-L, and SGNNO16 (as shown in Fig. 2), created by EDDA Technology Inc., ACCU Med Inc., Hanglok-Tech Co. Ltd., respectively.

FIG. 2.

SGNNO, reprinted with permission from https://www.hanglok-tech.cn/ (last accessed January, 2024). Copyright 2024 Hanglok-Tech Co. Ltd.

FIG. 2.

SGNNO, reprinted with permission from https://www.hanglok-tech.cn/ (last accessed January, 2024). Copyright 2024 Hanglok-Tech Co. Ltd.

Close modal

For this kind of robot systems, they are not highly automated and rely heavily on the previous experience of clinicians. At the same time, for the operation that lasts for several hours, the surgeons need to bear physical fatigue since an operation always lasts for several hours, having negative effect on surgery to certain extent.

The discussion of robot systems in this subsection could plan a puncture needle path, automatically align the insertion site according to the planned path, and manually advance the needle by a clinician, as shown in Table I. The following content is discussed according to the degrees of freedom (DOF) of the manipulator incorporating non-6-DOF and 6-DOF.

1. Systems with non-6-DOF manipulator

This part mainly introduces the non-6-DOF PISR systems, especially for the robot systems with lower than 6-DOF, including Needle Placement System (NPS), Micromate, ROBIO EX and MAXIO from DEMCON Advanced Mechatronics, Interventional Systems, and Perfint Healthcare, respectively.

Both NPS and Micromate are CT-guided navigation systems for needle positioning. Specifically, NPS is composed of a 2-DOF orientation module and a mechanical device, where the orientation module consists of two stacked rotatable segments and is connected to the CT table by a passive locking module. The insertion angle and depth are planned according to the target and entry point, which is completed manually by the operator. Micromate is a needle positioning and navigation system with 4-DOF, which is previously known as iSYS117,18 and B-Rob.19,20 The orientation module of iSYS includes X–Y axis positioning module and needle angle determination module. Based on these modules, the robot moves to the appropriate position with the proper angle according to the planned path. In the same way, the insertion of the needle is also finished manually by the operator.

ROBIO EX and MAXIO are both CT-guided ground-fixed stereotactic puncture positioning navigation systems. Namely, the appropriate needle entry point is recommended by calculating the spatial relationship between the systems of operating table and robot. The relative position of these two coordinate systems is kept unchanged. ROBIO EX is a 4-DOF surgical positioning assistance system, which plans the trajectory of the needle based on the axial relationship between CT images and entry point. Then, the clinician inserts the needle to the expected depth through the needle holder. MAXIO is an upgraded version of ROBIO EX, which introduces a 5-DOF manipulator. However, the insertion of needle is still completed by the clinician. Intraoperatively, multiple CT/MR scans are required to confirm whether the needle has reached the target. This method is with the attributes of long operation time and inability to deal with lesion movement, bleeding, pneumothorax, etc., as well as the huge exposure of radiation for patients.

The non-6-DOF PISR systems with automatic tracking function and manual placement of needle free the operators to some degree and greatly improve the accuracy of the needle positioning. However, the limitation of the puncturing scope makes it difficult to complete complex surgery and the probability of puncturing through the crucial tissues, like lung and stomach, will rise up, increasing surgery risk. Meanwhile, the respiratory motion of the patient would affect the actual position of the soft tissues, resulting in the tissue shifting and deformation as well as the loss of surgical accuracy.

2. Systems with 6-DOF manipulator

This part mainly introduces the 6-DOF automatic positioning robot systems, which assist the surgeons to automatically locate to the entry point of the complex body surface without the need to learn the complicated operation methods of the robot system.

TH-S1 of True Health (Guangdong hengqin) Medical Technology Co. Ltd. includes the following modules: intelligent image analysis and puncture planning, optical navigation, 6-DOF manipulator, and respiratory tracking. The respiratory tracking module collects respiratory information through surface marker signals to finish accurate punctures. Epione developed by Quantum Surgical is a CT-guided 6-DOF PISR system for liver ablation. With the assistance of this system, operators first need to manually move the manipulator to a position close to the surface of the human body. Subsequently, the manipulator can autonomously move to the entry point, following the planned puncture path. Then, the operator manually inserts the needle into the end-tip puncture needle sleeve according to the indicated depth.21,22 Epione follows the architecture of ROSA One robots series developed by ZIMMER BIOMET. This series of 6-DOF robots carry out surgical planning through 3D images, tracking the position of the manipulator and surgical instruments in real time.23,24 Specifically, the series of ROSA One robot includes the ROSA One Brain neurosurgical surgical robot, ROSA One Spine, and ROSA Keen orthopedic surgical robots.17,18,23,25,26 These robots integrate the Polaris VEGA optical sensing module from NDI27 and the 6-DOF collaborative robot from KUKA LBR Med (KUKA, Augsburg, DE).28 Notably, the manual control could directly offer the tactile cues and force feedback to operators. The robotic assisted radio frequency ablation was compared against the standard manual approach and the study showed that the robotic assisted treatment of liver tumors provided an improvement in terms of the procedure time, procedure accuracy, physician radiation exposure, and patient radiation exposure.28 Both the autonomous and the hands-on (i.e., cooperative) control modes for the robotic system were considered to assess the targeting accuracy and no significant difference was observed between the considered control modes.29 

The 6-DOF PISR has greatly boosted the mobility of the needle and the level of automation of the robots. However, it cannot achieve all the optimal clinical insertion angles. Additionally, several scans are required to verify the needle position, increasing the amount of exposed radiation.30,31 The loss of accuracy caused by respiratory movement is still not well solved, either. In the light of these issues to be optimized, the SGNIO shown in Fig. 3 was created by Hanglok-Tech Co. Ltd. with the consideration of the optical positioning, real-time ultrasound imaging, and patients' respiratory behavior.

FIG. 3.

SGNIO, reprinted with permission from https://www.hanglok-tech.cn/ (last accessed January, 2024). Copyright 2024 Hanglok-Tech Co. Ltd.

FIG. 3.

SGNIO, reprinted with permission from https://www.hanglok-tech.cn/ (last accessed January, 2024). Copyright 2024 Hanglok-Tech Co. Ltd.

Close modal

The robotic systems mentioned in this part with the features of planning the puncture needle path, automatically aligning the entry point and complete the advancement of the puncture needle by teleoperation, as shown in Table I. Compared to the robots where clinicians manually advance the onsite puncture needle, teleoperated needle-placement robots have received extensive attention due to their more flexible operation.

XACT Robotics launches a 5-DOF puncture needle navigation system named XACT ACETM Robotic system. This system is fixed to the patient's body with CT guidance to confirm the entry point and puncture path. The forward and stop of the needle are controlled by a foot pedal to process the insertion of needle. If the needle deviated from the planned path, the robot would correct and adjust the needle according to the current CT image information without repositioning the patient or reinserting the needle. This system also incorporates with a respiratory sensor to compensate for the deviation of puncture accuracy.32 

The robot systems with remote needle placement are first able to reduce the radiation exposure. Furthermore, they also alleviate the errors caused by manual needle placement, such as the breathing, cough, and natural shaking of the operators. Hanglok-Tech Co. Ltd. is now working on a teleoperated multi-functional robot system for seeds implantation—VINCENT, shown in Fig. 4, which fuses the information force, position, and image to provide better sense of immersion. The VINCENT is a master–slave teleoperated multi-functional particle implantation robot system based on a force-position-image hybrid navigation system.

FIG. 4.

VINCENT, reprinted with permission from https://www.hanglok-tech.cn/ (last accessed January, 2024). Copyright 2024 Hanglok-Tech Co. Ltd.

FIG. 4.

VINCENT, reprinted with permission from https://www.hanglok-tech.cn/ (last accessed January, 2024). Copyright 2024 Hanglok-Tech Co. Ltd.

Close modal

The above-mentioned three types of PISR systems have their own advantages and their level of automation has been gradually raised. However, there are many remained challenges demanding further investigations to enhance the puncture accuracy. Especially when it comes to the patient's respiratory movement, the positioning of needle's end tip, the deformation of the needle, and the uneven force of the needle in the soft tissues, the puncture accuracy cannot always be guaranteed. Actually, none of the current robotic systems can solve these issues well. Therefore, for promoting the subsequent evolution of PISR systems, this paper analyzes and discusses the research status from the view of involved technology, such as spatial registration and positioning navigation, puncture guidance feedback technology, respiratory motion compensation technology, and motion control technology.

In traditional interventional surgeries, clinicians should determine the shape, size, location, and surrounding tissues of the lesion by observing the medical images to decide the best entry point of needle and puncture path. This manual operation process requires rich experience of clinicians. Meanwhile, the effect of the puncture surgery mainly depends on the fitting between the actual puncture path and the planned path, since there are several factors that affect surgery precision and treatment results including the positioning of the needle tip, the alignment of needle pose, the instability of manual operations, etc.33 Compared to manual operation, robotic-assisted puncture surgery has absolute advantages in ensuring the accuracy and stability. It could assist clinicians in performing surgical operations and compensate the deficiencies of manual operations. The general process of robotic-assisted puncture interventional surgery is as follows: In the preoperative step, segmenting the target and tissues based on medical images, such as lung, stomach, and brain. Then, the puncture path can be automatically planned. Intraoperatively, tissues are rigidly or non-rigidly registered for accurately positioning the target. The surgical plan is accordingly adjusted to guide the robot to reach the planned entry point.

In order to plan paths in the surgery, it is necessary to obtain anatomical imaging information of the human body through images. The 3D visualization of medical images refers to reconstructing 3D models by segmenting and processing preoperative medical image data, like MRI or CT, to obtain digital anatomical models of the patient's lesions. This facilitates surgeons to determine the location of lesions and provides information support for subsequent surgical planning and intraoperative guidance.34 Based on the reconstruction of 3D models, a robot-assisted path is planned by integrating preoperative and intraoperative multimodal imaging information, interventional surgical operation requirements, and anatomical features. To derive the optimal path, a two-step strategy was proposed to optimize the most critical settings of minimally invasive interventional surgeries.35 An optimization-based approach was developed by formulating constant curvature needle path planning as a nonlinear problem.36 

Furthermore, the preoperative medical images should be registered to the patient's intraoperative coordinate system. As mentioned before, medical image registration generally includes rigid and non-rigid registration. The methods of rigid registration aim to the rigid organ, which are not influenced by the respiratory motion and widely used in neurosurgery.37,38 However, flexible tissues (e.g., chest and abdomen) always suffer motion or deformation because of respiratory behavior. For this issue, the common method is to assume that flexible tissues are rigid and the preoperative positions and gestures of these tissues could be maintained. Inevitably, this would lead to inexact registration results and cannot meet the accuracy requirements. Therefore, the deformation of flexible tissues should be considered and then register to preoperative images through non-rigid medical image registration methods. These methods could be classified into two categories: methods based on physical model and methods based on spatial transformation. Specifically, methods based on physical models treat the image as a physical model and image differences are shown through the physical deformation of model.39,40 This kind of method could maintain stable registration effects under ideal conditions, while the physical model is usually a simplification of the real physical world. Namely, the performance of these methods would be degraded when the ideal assumptions are removed and the applicability of these methods is limited. Methods based on spatial transformation are a parameterized method originating from interpolation and approximation theory.41 Basic functions are used to represent image deformations as well as combining interpolation algorithms to determine the mapping relationship between images.42,43

Deep learning is often adopted in non-rigid medical image registration since it has the ability to perform nonlinear fitting and could achieve higher registration accuracy.44 A learning-based image registration framework was proposed to discover compact and highly discriminative features.45 In many clinical applications, the fusion of multi-modal images is complex, and a deep similarity learning method was given that trained a binary classifier to learn the correspondence of two image patches.46 Traditional registration methods iteratively optimize the objective function of each pair of images to solve the spatial transformation, which has the problems of long registration time and large amounts of calculation. Due to this problem, a bidirectional diffeomorphic registration network was used to improve the accuracy of registration while ensuring the topology-preserving registration, providing a solution to the scarcity of bidirectional registration in traditional convolutional neural networks and local smoothness in the deformation field.47 The aforementioned registration methods based on deep learning are still in the experimental stage as they are established on a small amount of sample data and the scope of application is not clear. Hence, it remains to be further explored for registration methods to have both high registration accuracy and reliability.

The navigation module is the core of a PISR system, which connects the physical and virtual objects in the surgery. With the help of the navigation system, clinicians could have surgical operations, including intuitively estimating the location of lesions, setting entry point, locating target area, and guiding the manipulator to move.48,49 The accuracy of positioning the target area plays a decisive role for the effect of surgery. According to the positioning principle, it can be divided into mechanical positioning, electromagnetic positioning, and optical positioning. Among them, the most widely used are electromagnetic positioning and optical positioning, and the latter has highest precision.50,51

To be more specific, electromagnetic positioning devices have the advantages of small size, easy measurement, convenient operation, strong penetrability, and high accuracy. A noninvasive magnetic localization method of a helical robot was implemented with a diameter of 1 mm for clearing superficial blood clots, where the localization accuracy is characterized using visual feedback with a position tracking error of 2.35 ± 0.4 mm.52 However, the way of electromagnetic positioning is easily affected by metals and magnets. External electromagnetic waves generated by anesthesia machines, high-frequency electric knives, and other equipment will interfere with the positioning effect. In contrast, optical positioning has wider applications due to its ease of use and no electromagnetic interference.53,54 For example, based on the CAS-One IR navigation system, Medtronic added a custom needle holding module to locate the needle tip in real time and developed two surgical robot systems, Mazor X55,56 and StealthStationTM S8.57,58 Mazor X introduced an optical navigation system and a 6-DOF serial manipulator fixed on the operating table. It is compatible with preoperative CT and intraoperative 3D scanning planning modes. Also, it processes 2D perspective projections from the standard C-arm and interacts with 3D imaging to improve accuracy. In order to quantify the accuracy, the actual distance between each coordinate in the space and the vital structure could be introduced, to obtain the minimum distance between each potential path and the vital one.59 In addition, StealthStation S8 introduced a 3D camera for tracking instruments and electromagnetic sensors with the ability to fuse different imaging data sources. The function of 360° augmented reality (AR) is added to render the real-time navigation images, allowing anatomy to be viewed from different angles and synchronously enhancing surgical immersion. Likewise, in the optical navigation positioning process, Remebot integrates visual positioning algorithms to automatically complete registration by identifying and locating markers using three cameras and completes surgery manually.60 In particular, optical navigation systems are mainly provided by NDI in Canada except for self-development.61 

However, due to achieve “hand–eye coordination” operation in the robot systems ( “hand” refers to the manipulator and “eye” refers to the navigation system), hand–eye calibration is an essential task.62 Hand–eye calibration of robot refers to achieving mapping from optical imaging space to manipulator space by calculating the precise coordinate relationship between “hand” and “eye,” thereby controlling the end effector of manipulator to reach the target position.63 The distance error was measured between the real tip of the external probe and the calculated tip to estimate the orientation error.64 The angle error between the regressed motion and the real motion was used to evaluate the accuracy of motion direction estimation.65 To achieve a relatively ideal accuracy in tracking and positioning, combining MR and ultrasound is a useful approach for medical registration.66 Nevertheless, ultrasound guidance limits the accessible area and cannot be used for lung-directed cases because the lungs are full of air. Then, some studies were carried out based on CT. For example, a robotic needle insertion system was proposed using CT images and visual servoing, where the skin entry point and the target to be reached were defined in the CT image and the needle target was then expressed in the robot frame thanks to a 3D registration.67 A system composed of CT, a stereo camera, and a robotic system was designed to percutaneously position a needle, where the needle were detected in the triangular region of interest using the Hough transform to identify needle borders.68 For providing a high accuracy of the needle position and orientation in real-time to surgeons rather than relying on their individual experience, a method was proposed through making full use of edge points from two sides of the needle in image and creating center points through which an accurate axis was extracted.69 

In PISR systems, the patient may be harmed if the robot control is accurate. In the process of traditional puncture, medical staff often perceives the position of the needle tip by their previous experience. In order for the robot to complete the puncture process automatically, there should be a perception feedback module, i.e., a real-time information feedback module. This module is generally responsible for collecting data from imaging equipment and sensors, and this collected information is then feedback to the controller to further guide the puncture needle.54,70

PISR systems are mostly based on visual feedback, utilizing ultrasound devices or camera equipment for navigation guidance during the puncture process. Depending on the imaging method, ultrasound devices provide feedback based on ultrasound imaging, while most camera devices offer feedback using near-infrared optics.

The merits of ultrasound guidance are that it can obtain real-time information with low cost and no radiation. While the drawbacks are the narrow imaging range and the potential needle deviation from the field of view.71,72 Amit and ClearGuide both fuse ultrasound images to take advantage of ultrasound images' mobility, providing clinicians with more intuitive image references.73–75 Accordingly, the advantages of near-infrared image feedback are its wide imaging range and the entire motion process within the camera's field of view. The disadvantages are the difficulty in obtaining accurate depth information, which can easily lead to the puncture failure.76,77 Taking the aforementioned systems as an example, Epione, Mazor X, and StealthStationTM S8 all enhance the positioning and navigation capabilities of the puncture system through the principles of near-infrared light feedback.

To improve the quality of the acquired ultrasound images, the ultrasound confidence map was adopted in an ultrasound-based visual servoing framework.78 As an extension of this work, a framework was designed that embedded the confidence-driven control and the ultrasound probe based on the image quality signal.79 Also, in order to the high-quality sound propagation within the tissue, a method was presented that aligned the central axis of the US probe to the tissue's surface normal at the point of contact.80 To achieve real-time robotic insertion of bevel-tipped flexible needles, a teleoperation framework was considered that enabled the operator to directly and intuitively control the trajectory of the needle tip via a haptic interface.81 However, due to the inherent characteristics of ultrasound imaging, a certain contact force between the probe and human is required to optimize acoustic coupling. For manually operating clinicians, maintaining a constant force during ultrasound scanning is challenging. Meanwhile, varying forces would lead to deformation of ultrasound images. As a result, specialized force controllers are needed to maintain contact force during scanning and such controllers are also crucial for preventing excessive force to ensure the safety of patients.82,83 It is worth noting that there is only little work that integrates the ultrasound and near-infrared light, which could provide multi-angle navigation for the insertion process and improves surgical accuracy.

To achieve more accurate control of the PISR, researchers have begun to add sensors for real-time puncture feedback force detection and explore a variety of feedback mechanisms to improve the control effect. In traditional puncture surgeries, medical staff needs to judge whether the needle tip is aligned with the corresponding target according to their intuition or experience for adjusting the needle insertion angle and speed correspondingly. At present, most of the existing PISR systems only provide visual feedback without force feedback. Therefore, some work introduced force sensors in the perceptual feedback module to obtain more perceptual information to guide the robot to manipulate the puncture needle more accurately.

As an example, a way was proposed to autonomously localize the epigastric region as the starting position for the automatic ultrasound scan with a six axes force sensor fixed to the end effector of the manipulator.84 A robotic catheter manipulating system was developed to reduce the performance error and radiation, as well as a torque sensor was applied to measure the torque information during the operation.85 In order to stop the scanning process when the pressure on the skin exceeded a preset threshold, two force sensors were attached to the front side of the probe for force measurement.86 As an extension, a robotic system with 6-DoF was developed attaching two force sensors to the left and right sides of the front face of the ultrasound probe.87 Similarly, an autonomous robotic system for US-guided biopsy of breast lesions was proposed, in which the applied force was measured by a force sensor located in the flange of the robot.88 There are some research with the robot's built-in force and torque sensors, further allowing ultrasound acquisitions with optimal skin force and providing repeatable 3D ultrasound volumetric datasets.89–92 Although the reported systems generated a contact force by pushing the ultrasound probe onto the body surface directly with actuators and sensors, it is of concern that an excessive contact force can be accidentally applied through control error or because of the failure of actuators and sensors.93 

Recently, fiber optic sensors have attracted widespread attention due to their characteristics of resistance to electromagnetic interference, excellent biocompatibility, lower cross-sensitivity to temperature, and miniaturized size. This kind of sensor could overcome most of the limitations of electrical sensitive components and achieve precise measurement of surgical instrument deformation in confined space.94,95

In order to provide shape sensing of the continuum manipulator, a nitinol wire with an integrated array of 12 Fiber Bragg Grating (FBG) sensors was introduced into the manipulator.96 A triaxial distal force sensor for haptic feedback was developed through a novel high-precision FBG based on freedom and constraint topology.97 Phase modulated fiber optic force sensing is another significant approach to measure the force.98 To be specific, Fabry–Pérot Interferometry (FPI) based fiber optic sensors are convenient for force measurement during MRI guided biopsy procedures because they are MRI compatible and safe for clinical use due to their dielectric nature, chemical inertness, and nontoxicity.99 

In response to the issue of puncture needles easily undergoing deformation, an approach was developed to reconstruct the needle shape using the strain data obtained from the FBG sensors and then it was as the feedback information for the steering algorithm.100 Similarly, to determine the shape of a needle that is being inserted, a method was considered based on the elastic rod theory and Lie group theoretic approach, together with the curvature measurement data obtained by FBG sensors inside the needle.101 A reconstruction technique that provides the pose of fiber was presented by converting the measurement from the Bragg gratings to strain and the deduced curvature.102 Then, a temperature insensitive calibration model for calibrating needles with FBG sensors was considered to accurately estimate its deformation and a device was designed for simultaneously calibrating the loading direction and shape.103 A three-axial force sensor was proposed to measure the pulling force, the steering force, and the lifting force in a flexible endoscopic surgical robot with FBGs embedded in the sensing structure.104 Notably, the higher the degree of polynomial fitting, the greater the accuracy required for the position information. Otherwise, any deviation of position information could lead to significant changes in the reconstruction results. Hence, the aforementioned reconstruction method is not suitable for complex deformation reconstruction.

In percutaneous interventional surgeries, the most commonly used instrument is the puncture needle, which is mainly divided into rigid and flexible types. Specifically, rigid needles perform punctures in a straight line, but the linear path of needle insertion is restricted. Due to the rigid characteristics of the needle, it is easy to puncture through important blood vessels or nerves during the process. Needle deformations and the patient's breathing can also cause deviations in needle positioning, leading to an inability to reach the target location, severely affecting the effectiveness of surgical treatment.105 In contrast to rigid puncture needles, asymmetric bevel-tipped flexible needles experience bending in the body due to forces like tissue-generated action and friction. This allows them to execute curved movements to reach targets within tissues, avoiding vital organs, nerves, and blood vessels.106,107 Flexible needles could achieve curved puncture paths since they have lower needle body stiffness, executing more refined interventional operations and reducing surgical injuries. Inevitably, interventional techniques with flexible needles have become a research hotspot recently. They are categorized into passive and active types according to the driving methods. Compared to rigid needles, asymmetric bevel-tipped flexible needles tend to bend as the generated force, such as friction and pressure. This means flexible needles easily bypass vital organs and reach the target during puncture process. Owing to the lower stiffness, flexible needles can follow curved puncture paths, allowing for more sophisticated interventional operations and reducing surgical injuries. Depending on the driving method, flexible needles are classified as passive and active. However, the interactions between flexible needles and soft tissue are complex and related challenges need to be further investigation, such as friction and torsional control as well as the tip pose estimation.108,109

On the one hand, passive flexible needles are curved and their puncture path can be altered by changing the base direction or varying the interaction force between the needle tip and tissues. How to achieve low stiffness suitable for flexible organizations is a technical challenge for passive flexible needles. Flexible needle steering techniques offer improved control over implant placement, but often require complex closed-loop control for accurate implantation.110 On the other hand, the bending of active flexible needles requires external actuators, such as rope-driven, piezoelectric ceramic-driven, and shape memory alloy-driven. Introducing active flexible needles can add active degrees of freedom to the movement of needle tip. An active needle was proposed for the development of MRI-guided percutaneous procedures, where the needle used a low-transition-temperature shape memory alloy wire actuator to produce bending in the distal section of the needle.111 Also, a 3D path planning method to steer flexible needles along curved paths was proposed on the basis of a rapidly exploring random tree strategy.112 In order to improve the accuracy of needle placement, the kinesthetic and vibrotactile haptic feedback were combined and a friction estimation algorithm was designed to extract salient information about the cutting force at the needle tip from a force sensor placed at the needle base.113 Furthermore, based on the needle shaft radial cross-sectional centroid coordinates, two methods were proposed to estimate the needle tip pose in real-time, which were based on Rodrigues' rotation formula, and Fourier expansion with three-terms and linear interpolation algorithm, respectively.114 To further address the issue of needle tip's positioning, ultrasound technology115 should be integrated and deployed to plan and execute interventions accurately because of its better reproducibility of measurements,116 such as the motion-aware robotic ultrasound imaging technologies,117,118 ultrasound deformation correction approaches,119 and dual-robot tracking algorithms.91 

Although a large number of puncture robots and systems have been developed, there are still many problems to be investigated for thoracoabdominal PISR systems. The main reason is that the chest and abdomen undergo nonlinear and non-rigid deformation because of respiration. Also, the registration of medical image space and patient space is generally based on rigid matching, which makes it hard to achieve real-time positioning of the target position in vivo. As well, there would be a relative movement between needle and tissues due to the respiration. Consequently, it is vital to research respiratory motion compensation technology in robotic surgery.120,121

In percutaneous puncture surgery, the bearing condition of the puncture needle is intricate as the complex structure of human skin tissues. To simulate the clinical puncture procedure conducted by clinicians as closely as possible and to compensate for the rigidity shortcomings of robots, some work focuses on establishing respiratory motion models to explore the interaction forces between puncture needle and tissue in soft tissues.122,123 To further clarify, the interaction forces mainly comprise of normal force, frictional force, and cutting force.124 Depending on the various stages of the puncturing process, the needle bears different forces. When the needle tip punctures the tissue, it is primarily subjected to the normal force. As it pierces through the tissue nearing the target, the main bearing forces are the cutting force from the needle tip passing through the tissue and the frictional force along the direction of the needle shaft. During the needle withdrawal phase, it is primarily subjected to frictional force.125 

According to the puncture forces during different stages, some work designs patients' respiratory motion models. Instantly, the correlation of the external signal to the motion of the liver in a porcine study using ϵ-support vector regression was carried out.126 Some work correlates respiratory signal data with internal tumor movement data aiming to predict the real-time position of tumors on the basis of deep learning.48,127,128 Yet, these methods were proposed with the assumption that there was a consistent correlation between respiratory signals and the movement of tumors, which might not always be valid from the view of respiratory motion's physiology.129,130 On the basis of recent lung motion measurements and the physiologic functioning of the lungs, the motion of lung and lung tumor tissues was modeled with the position of the tissues at a user-specified reference breathing phase, tidal volume, and its temporal derivative airflow (tidal volume phase space).131 A robotic needle biopsy technique was introduced, adapting to the patient respiratory pattern and using a robot manipulator to drive the needle toward a moving lung nodule. Then, needle placement was planned to follow an optimal timing and path and was triggered based on the respiratory phase tracking.132 To insert accurately needle without stopping the breath, the intermittent insertion control method was proposed, which contributed to decreasing the discomfort and the amount of radiation exposure. Also, a recursive image registration network under deformable respiratory motion was proposed based on ordinary differential equations, where the network learned to estimate time varying voxel velocities to model the deformation in 4D image data.133 

In addition to establishing respiratory motion models, some work proposed strategies to estimate the motion patterns of the target under the influence of respiration in real-time, aiming to reduce the accuracy impact brought by the respiration in the surgery. Currently, multimodality imaging techniques, such as fluoroscopy, ultrasound, and real-time MRI, have been adopted to alleviate this burden. The True Beam system, developed by Varian based on respiratory gating technology,134 has been successfully applied in the field of radiation oncology. However, this system is not suitable for thoracoabdominal puncture interventions. Therefore, how to reasonably apply the motion patterns of targets under the influence of respiration to thoracoabdominal PISR systems is still one of the challenges in this field.

A fast CT-CTF deformable registration algorithm was proposed that warped the preprocedural inhale CT onto the intraprocedural CTF for guidance in 3D with incorporating a respiratory motion compensation framework for accurate registration,135 where the inhale image, exhale image, and an intermediate phase CT were used for each subject. The shape motion model was designed to enable an unsupervised decomposition of respiration induced high-dimensional body surface displacement fields into a low-dimensional representation encoding thoracic and abdominal breathing.136 Furthermore, a real-time matching algorithm to compensate for respiratory motion between a 2D fluoroscopic image and 3D CT images of the lung was proposed, regardless of cardiac motion, based on a newly improved matching measure. This algorithm can improve the accuracy of a guiding system by providing 3D images precisely registered to 2D fluoroscopic images in real-time, without time-consuming respiratory-gated or cardiac-gated CT images.137 However, the extraction of real-time image targets and the prediction of moving target areas require high computational costs, which are not conducive to the real-time motion control of robots. Meanwhile, CT imaging takes a long time, making real-time scanning hard to achieve. Also, the diaphragmatic movement caused by breathing has a certain periodicity under CT guidance, while the respiratory cycle is not symmetrical. The exhalation phase within a cycle is generally longer than the inhalation phase, and the movement direction is mainly in the up and down direction rather than the expansion or contraction.

Applying markers or sensors on the skin to track the patient's respiratory motion is a common practice because there is a strong correlation between the trajectory of tumor movement and the fluctuation of the chest and abdominal skin surface during breathing. Then, the robot's operation could be adjusted correspondingly.130,138–141

The relationship between internal and external marker positions was continuously accounted for and was regularly checked and updated, where a polynomial model was studied to describe the relationship between internal tumor motion and external (chest/abdominal) marker motion.142 As a commercial representative, three external markers were used by CyberKnife, developed by Accuray, to capture the patient's respiratory characteristics, establish the correlation function between markers and tumor locations, and then estimate the location of the tumor.143 Also, a method was proposed to the 4D active tracking and dynamic delivery incorporating the tumor motion prediction technique, which was applied to the two commercially available robotic treatment couches.144 Furthermore, a marker-based depth frame registration technique was introduced to limit the measuring area into an anatomically consistent region, handling the patient's movement during the treatment.145 The approach to extract highly representative surface motion features based on selecting moving surface areas was considered. Similarly, the surface area was segmented into several regions before measuring their corresponding correlations with internal tumor motion according to high Pearson's correlation coefficients.146 It is noted that the estimation of the tumor position requires extracting the respiratory motion characteristics of the chest and establishing correlation models manually, which is always affected by the placement and number of markers. As a result, point clouds of the chest and abdominal skin surface were employed and a hierarchical network was built to extract the features of the skin and map those features to the location of tumors.147 

Surgical robots have complex dynamics, strong coupling between degrees of freedom, and nonlinear characteristics, such as hysteresis, which need to rely on advanced control technology to achieve their precise operation. At the same time, master–slave control is a common control method of surgical robots, which can not only give full play to the advantages of precise control of the robot but also fully combine the clinical experience of the surgeon. This section focuses on the analysis and discussion of the manipulator motion control and needle motion control involved in the puncture surgery robot.

To achieve precise control of the manipulator, innovative algorithms based on classical motion control theory provide a fundamental guarantee for the flexible movement and reliability. For instance, a haptic teleoperation system with force feedback and a bilateral controller method was suggested, where the master control used impedance control and the slave control used a sliding mode control.148 For collaborative operations under force-position hybrid control, a robot trajectory planning for edge detection and a collaborative control strategy with constant force were proposed.149 As a matter of fact, real robotic systems are complex to model owing to the nonlinear and time-varying property. To deal with this issue, an active-model-enhanced control scheme was developed to enhance a PID controller, and the modeling errors was eliminated by integrating an active modeling algorithm into this controller.150 A way of controlling the position of the active arms of a Da Vinci-type surgical robot was considered and a camera was placed on the slave system to provide these positions information, then the presence of medical staff to handle these arms manually is no longer necessary.151 Also, a perception and control framework were achieved that used real-time tactile feedback to accomplish the task of following a dangling cable, in which an linear-quadratic regulator controller based on a learned linear model of the cable sliding dynamics was combined.152 

In the context of master–slave control, a dual-loop control system was considered, including closed-loop inverse kinematics and adaptive control with optimal control parameters, for the slave robot to track a random trajectory upon updating the dynamic parameters of the slave robot and this system could be potentially utilized for actual teleoperated surgery and surgical training.153 Then, the lack of maneuverability for physicians when operating master–slave robots is also a significant challenge.154 Therefore, a control scheme of master–slave manipulators that provided the ideal kinesthetic coupling such that the operators can maneuver the system as though they were directly manipulating the remote object themselves.155 On the basis of requirement analysis of manual operations, a master–slave robotic system mainly consists of a 3-DOF slave manipulator and a 3-DOF master manipulator to achieve intuitive motion mapping and proper ergonomics.156 To provide the operators with high-fidelity force feedback, a thorough modeling approach based on successive isolation of system components was investigated.157 Similarly, a force feedback hand controller system was shown, including a 3-DOF translational and 3-DOF rotational hand controllers, respectively, to implement position and posture teleoperation of the robot end effector.158 Considering the stability issues, a type of the force reflection algorithms was studied on the stability and performance of a dual-arm haptic-enabled teleoperator system both with negligible as well as non-negligible communication delays.159 A forward and inverse kinematics model of the slave hand and a master–slave motion control algorithm were designed based on the screw theory and motion model to boost the hand–eye coordination consistency of the surgical robot system, by integrating consistency motion control, relative position control and proportional control.160 

Master–slave robots have certain advantages in improving surgical accuracy, while the careful attention is required to avoid dangers from inappropriate operations.161 To have fast tracking performance and strong robustness, a variable universe fuzzy PID method was adopted to improve fuzzy PID method's control precision without reducing the speed of response.162 Furthermore, for enhancing the precision and comfort of operations, a physical human-robot interaction control scheme of the haptic master manipulator was used with a torque observer based on generalized momentum.163 To learn the finger motions and to determine when to lift an object, an algorithm was trained by combing fingertip tactile sensing, joint torques, and proprioception as well as a multimodal agent through reinforcement learning.164 As for the aforementioned haptic feedback, it has been investigated by many researchers. This technology could not only help surgeons minimize potential side effects but also reduce errors in remote operations, tissue damage, and operation time.165–167 

During the whole puncture operation, the key issue is how to control the puncture needle to reach the target point while avoiding important tissues, so the motion control of the puncture needle is particularly important. Based on the basic concepts of puncture needles (see Sec. IV D), this subsection first discusses the motion control of flexible puncture needles. There are two main modes of motion control for flexible puncture needles: teleoperation control and autonomous control.168 

Teleoperation control mainly involves manipulating the robot's master arm to control the slave arm located at a distant end, thereby accomplishing the advancement and rotation of the needle. Traditional control algorithms have the advantages of simplicity, robustness and ease of control and a PID algorithm was proposed to control the needle's motion.169 While these traditional methods cannot satisfactorily meet the puncture performance since the needle is always subjected to the force of human tissue. In the presence of tissue deformation and uncertainty, a feedback controller was designed to steer a needle along 3D helical paths and vary the helix radius to correct for perturbations.170 Then, an approach keeping the physician in the loop to control insertion speed was proposed, where the tip was automatically guided to a planar slice of the tissue during the insertion process.171 To maintain the needle tip in a desired plane in 2D control, three estimation algorithms for estimating the states of the needle tip were employed to measure the position and the posture of the needle in the process of puncture.172 

Autonomous robots primarily rely on real-time imaging to independently perform the puncture surgery. Since this scenario relates to safety and ethics, there are many issues to be resolved before it can be practically implemented. The early puncture robots are mainly single DOF, which could only be used as an experimental platform. For achieving higher precision and smoother actions as well as completing more complex tasks, robots should have more DOF, for example, 3-DOF puncture mechanisms,173 4-DOF manipulators operating needle to complete insertion actions,174 5-DOF,175,176 and 7-DOF robots.177 To accurately guide the needle toward inner body targets, a model-based nonlinear observer was adopted for partially estimating the needle tip orientation by image-based position measurements.178 However, this method can only obtain the pitch angle and rotation angle of the needle, and the absence of yaw angle is a limitation. Benefit of flexible echogenic needles for robotized guidance under 3D ultrasound was quantified for improving needle visibility and detection robustness, where echogenic term refers to the etching of microstructures on the needle shaft.179 Similarly, a control law using task functions was proposed to reduce targeting error and forces applied to the tissue, steering via base manipulation and tip-based steering.180 Then, a path planning and tracking control method for flexible needle steering in dynamic environment was proposed based on the artificial potential field algorithm and curve fitting algorithm.181 

Robotic systems are renowned for their precision in spatial positioning, high stability, dexterous maneuverability, and operational endurance. The integration of robotics with medical technology holds significant clinical value and represents a current direction of development in the field. In this review, we highlight commercialized PISR systems that are at the cutting edge of technology and advancements. Despite the remarkable capabilities of these robotic systems, they are not without their limitations and challenges, particularly in addressing the complexities of soft tissue deformation and the intricacies of patient respiratory movements. To pave the way for the next-generation of PISR systems, a concerted effort toward evolutionary advancements is necessary and several key directions could be considered: 1. Enhanced multi-dimensional perception: future PISR systems should harness information across a spectrum of dimensions, not limited to but including optical information, force feedback, positional insight, and visual cues. An integration approach to holistic information is essential for achieving a comprehensive understanding of the surgical environment. 2. Advanced information acquisition: the incorporation of ultrasound technology into navigational systems, when synergized with AI or advanced LLM techniques, promises a leap in the robots' perceptual capabilities. This combination is poised to unlock a new level of precision and awareness within robotic systems. 3. Hardware innovation: addressing the need for minimizing patient trauma, especially in procedures requiring multiple puncture needles, calls for the innovation of robot hardware structures, to broaden the applicability of robots in complex puncture procedures.

The work was supported by the National Key Research and Development Program (Nos. 2018YFA0704100 and 2018YFA0704104), the National Natural Science Foundation of China (Nos. 81827805, 82130060, 61821002, and 92148205), the Zhuhai Industry-University-Research Collaboration Program (No. ZH22017002210011PWC), the Jiangsu Provincial Medical Innovation Center (No. CXZX202219), the Collaborative Innovation Center of Radiation Medicine of Jiangsu Higher Education Institutions, Nanjing Life Health Science and Technology Project (No. 202205045), and the Natural Science Foundation of Guangdong Province of China (Nos. 2021A1515220183 and 2022B1515020010). The funding sources had no role in the writing of the report or decision to submit the paper for publication.

The authors have no conflicts to disclose.

Ethics approval is not required.

Cheng Wang and Li Guo contributed equally to this work.

Cheng Wang: Conceptualization (equal); Funding acquisition (equal); Resources (equal); Writing – original draft (equal). Li Guo: Conceptualization (equal); Investigation (equal); Resources (equal); Writing – original draft (equal). Jianjun Zhu: Resources (equal); Writing – review & editing (equal). Lifeng Zhu: Supervision (equal); Writing – review & editing (equal). Chichi Li: Investigation (equal); Writing – review & editing (equal). Haidong Zhu: Supervision (equal). Aiguo Song: Supervision (equal); Writing – review & editing (equal). Ligong Lu: Supervision (equal); Writing – review & editing (equal). Gao-Jun Teng: Funding acquisition (equal); Supervision (equal); Writing – review & editing (equal). Nassir Navab: Supervision (equal); Writing – review & editing (equal). Zhongliang Jiang: Writing – review & editing (equal).

Data sharing is not applicable to this article as no new data were created or analyzed in this study.

1.
Global Burden of Disease 2019 Cancer Collaboration
, “
Cancer incidence, mortality, years of life lost, years lived with disability, and disability-adjusted life years for 29 cancer groups from 2010 to 2019: A systematic analysis for the global burden of disease study 2019
,”
JAMA Oncol.
8
,
420
444
(
2022
).
2.
Y.
Guan
,
M.
Ren
,
D.
Guo
, and
Y.
He
, “
Research progress on lung cancer screening
,”
Chin. J. Lung Cancer
23
,
954
960
(
2020
).
3.
R. A.
Charalel
,
G.
McGinty
,
M.
Brant-Zawadzki
,
S. C.
Goodwin
,
N. M.
Khilnani
,
A. H.
Matsumoto
,
R. J.
Min
, and
G. M.
Soares
, “
Interventional radiology delivers high-value health care and is an imaging 3.0 vanguard
,”
J. Am. Coll. Radiol.
12
,
501
506
(
2015
).
4.
L.
Li
,
X.
Li
,
B.
Ouyang
,
H.
Mo
,
H.
Ren
, and
S.
Yang
, “
Three-dimensional collision avoidance method for robot-assisted minimally invasive surgery
,”
Cyborg Bionic Syst.
4
,
0042
(
2023
).
5.
Z.
Song
,
W.
Zhang
,
W.
Zhang
, and
D.
Paolo
, “
A novel biopsy capsule robot based on high-speed cutting tissue
,”
Cyborg Bionic Syst.
2022
,
9783517
.
6.
R. R.
Feng
and
L.
Tang
, “
Path planning system of lung puncture based on CT image
,”
Robot
44
,
694
707
(
2022
).
7.
G. J.
Teng
,
C.
Wang
, and
J.
Lu
, “
Current status and future of interventional robotic surgical system
,”
Chin. J. Dig. Surg.
20
,
479
486
(
2021
).
8.
M.
Yang
,
H.
Ding
,
J.
Kang
,
L.
Zhu
, and
G.
Wang
, “
Subject-specific real-time respiratory liver motion compensation method for ultrasound/MRI/CT fusion imaging
,”
Int. J. Comput. Assisted Radiol. Surg.
10
,
517
529
(
2015
).
9.
N. K.
Wang
,
M. Z.
Luo
, and
L. J.
Lu
, “
Application progress of puncture robot in minimally invasive surgery
,”
Chin. J. Pain Med.
26
,
376
380
(
2020
).
10.
S.
Volpi
,
G.
Tsoumakidou
,
A.
Loriaud
,
A.
Hocquelet
,
R.
Duran
, and
A.
Denys
, “
Electromagnetic navigation system combined with high-frequency-jet-ventilation for CT-guided hepatic ablation of small us-undetectable and difficult to access lesions
,”
Int. J. Hyperthermia
36
,
1051
1057
(
2019
).
11.
Z.
Zhang
,
G.
Shao
,
J.
Zheng
,
S.
Wen
,
H.
Zeng
,
W.
Hao
,
J.
Luo
, and
L.
Guo
, “
Electromagnetic navigation to assist with computed tomography-guided thermal ablation of liver tumors
,”
Minim Invasive Ther. Allied Technol.
29
,
275
282
(
2020
).
12.
T. D.
Baere
,
C.
Roux
,
G.
Noel
,
A.
Delpla
,
F.
Deschamps
,
E.
Varin
, and
L.
Tselikas
, “
Electromagnetic navigation to assist with computed tomography-guided thermal ablation of liver tumors
,”
Eur. Radiol. Exp.
6
,
13
(
2022
).
13.
F.
Ficuciello
,
G.
Tamburrini
,
A.
Arezzo
,
L.
Villani
, and
B.
Siciliano
, “
Autonomy in surgical robots and its meaningful human control
,”
Paladyn J. Behav. Rob.
10
,
30
43
(
2019
).
14.
Z.
Jiang
,
X.
Li
,
C.
Zhang
,
Y.
Bi
,
W.
Stechele
, and
N.
Navab
, “
Skeleton graph-based ultrasound-CT non-rigid registration
,”
IEEE Rob. Autom. Lett.
8
,
4394
4401
(
2023
).
15.
Z.
Jiang
,
C.
Li
,
X.
Lil
, and
N.
Navab
, “
Thoracic cartilage ultrasound-ct registration using dense skeleton graph
,” arXiv:2307.03800 (
2023
).
16.
See https://www.hanglok-tech.cn/ for information about PISR systems (last accessed January, 2024).
17.
S. S. S.
Khalsa
,
P. V.
Mummaneni
,
D.
Chou
, and
P.
Park
, “
Present and future spinal robotic and enabling technologies
,”
Oper. Neurosurg.
21
,
48
56
(
2021
).
18.
M.
Huang
,
T. A.
Tetreault
,
A.
Vaishnav
,
P. J.
York
, and
B. N.
Staub
, “
The current state of navigation in robotic spine surgery
,”
Ann. Transl. Med.
9
,
86
(
2021
).
19.
R. M.
Martinez
,
W.
Ptacek
,
W.
Schweitzer
,
G.
Kronreif
,
M.
Fürst
,
M. J.
Thali
, and
L. C.
Ebert
, “
CT-guided and minimally invasive and postmortem needle biopsy using the B-Rob II needle-positioning robot
,”
J. Forensic. Sci.
59
,
517
521
(
2014
).
20.
J.
Kettenbach
,
G.
Kronreif
,
M.
Figl
,
M.
Fürst
,
W.
Birkfellner
,
R.
Hanel
, and
H.
Bergmann
, “
Robot-assisted biopsy using ultrasound guidance: Initial results from in vitro tests
,”
Eur. Radiol.
15
,
765
771
(
2005
).
21.
T.
de Baère
,
C.
Roux
,
F.
Deschamps
,
L.
Tselikas
, and
B.
Guiu
, “
Evaluation of a new CT-guided robotic system for percutaneous needle insertion for thermal ablation of liver tumors: A prospective pilot study
,”
Cardiovasc. Intervent. Radiol.
45
,
1701
1709
(
2022
).
22.
B.
Guiu
,
T. D.
Baère
,
G.
Noel
, and
M.
Ronot
, “
Author correction: Feasibility and safety and accuracy of a CT-guided robotic assistance for percutaneous needle placement in a swine liver model
,”
Sci. Rep.
11
,
5218
(
2021
).
23.
M.
Lefranc
and
J.
Peltier
, “
Evaluation of the ROSATM spine robot for minimally invasive surgical procedures
,”
Expert Rev. Med. Devices
13
,
899
906
(
2016
).
24.
N.
Lonjon
,
E.
Chan-Seng
,
V.
Costalat
,
B.
Bonnafoux
,
M.
Vassal
, and
J.
Boetto
, “
Robot-assisted spine surgery: Feasibility study through a prospective case-matched analysis
,”
Eur. Spine. J.
25
,
947
955
(
2016
).
25.
J.
Gonzalez-Martinez
,
S.
Vadera
,
J.
Mullin
,
R.
Enatsu
,
A. V.
Alexopoulos
,
R.
Patwardhan
,
W.
Bingaman
, and
I.
Najm
, “
Robot-assisted stereotactic laser ablation in medically intractable epilepsy
,”
Neurosurgery
10
,
167
173
(
2014
).
26.
M.
Lefranc
and
J.
Peltier
, “
Accuracy of thoracolumbar transpedicular and vertebral body percutaneous screw placement: Coupling the ROSA® spine robot with intraoperative flat-panel CT guidance—A cadaver study
,”
J. Rob. Surg.
9
,
331
338
(
2015
).
27.
G.
Fattori
,
A. J.
Lomax
,
D. C.
Weber
, and
S.
Safai
, “
Technical assessment of the NDI polaris Vega optical tracking system
,”
Radiat. Oncol.
16
,
87
(
2021
).
28.
G.
Fichtinger
,
J.
Troccaz
, and
T.
Haidegger
, “
Image-guided interventional robotics: Lost in translation?
,”
Proc. IEEE
110
,
932
950
(
2022
).
29.
M.
Niccolini
,
V.
Castelli
,
C.
Divers
,
B.
Kang
,
F.
Mussa
, and
E.
Sinibaldi
, “
Development and preliminary assessment of a robotic platform for neuroendoscopy based on a lightweight robot
,”
Int. J. Med. Rob. Comput. Assisted Surg.
12
,
4
17
(
2016
).
30.
A.
Patriciu
,
M.
Awad
,
S. B.
Solomon
,
M.
Choti
,
D.
Mazilu
,
L.
Kavoussi
, and
D.
Stoianovici
, “
Robotic assisted radio-frequency ablation of liver tumors—Randomized patient study
,” in
Medical Image Computing and Computer Assisted Intervention
,
2005
.
31.
R.
Pollock
,
P.
Mozer
,
T. J.
Guzzo
,
J.
Marx
,
B.
Matlaga
,
D.
Petrisor
,
B.
Vigaru
,
S.
Badaan
,
D.
Stoianovici
, and
M. E.
Allaf
, “
Prospects in percutaneous ablative targeting: Comparison of a computer-assisted navigation system and the acubot robotic system
,”
J. Endourol.
24
,
1269
1272
(
2010
).
32.
S.
Levy
,
S. N.
Goldberg
,
I.
Roth
,
M.
Shochat
,
J.
Sosna
,
I.
Leichter
, and
S.
Flacke
, “
Clinical evaluation of a robotic system for precise CT-guided percutaneous procedures
,”
Abdom. Radiol.
46
,
5007
5016
(
2021
).
33.
A.
Guo
,
S.
Pomenti
, and
J.
Wattacheril
, “
Health disparities in screening and diagnosis and treatment of hepatocellular carcinoma
,”
Clin. Liver Dis.
17
,
353
358
(
2021
).
34.
X. Z.
Qi
,
J.
Meng
,
M.
Li
,
Y. Y.
Yang
,
Y.
Hu
,
B.
Li
,
J. W.
Zhang
, and
W.
Tian
, “
An automatic path planning method of pedicle screw placement based on preoperative CT images
,”
IEEE Trans. Med. Rob. Bionics
4
,
403
413
(
2022
).
35.
L.
Adhami
and
E.
Coste-Maniere
, “
Optimal planning for minimally invasive surgical robots
,”
IEEE Trans. Rob.
19
,
854
863
(
2003
).
36.
S. A.
Pedram
,
P.
Ferguson
,
J.
Ma
,
E.
Dutson
, and
J.
Rosen
, “
Autonomous suturing via surgical robot: An algorithm for optimal selection of needle diameter and shape and path
,” in
Proceedings of International Conference on Robotics and Automation
(ICRA,
2017
), pp.
2391
2398
.
37.
R. R.
Shamir
,
L.
Joskowicz
, and
Y.
Shoshan
, “
Fiducial optimization for minimal target registration error in image-guided neurosurgery
,”
IEEE Trans. Med. Imaging
31
,
725
737
(
2012
).
38.
M.
Wang
and
Z.
Song
, “
Improving target registration accuracy in image-guided neurosurgery by optimizing the distribution of fiducial points
,”
Int. J. Med. Rob.
5
,
26
31
(
2009
).
39.
S. P.
Singh
,
L.
Wang
,
S.
Gupta
,
H.
Goli
,
P.
Padmanabhan
, and
B.
Gulyás
, “
3d deep learning on medical images: A review
,”
Sensors
20
,
5097
(
2020
).
40.
G.
Haskins
,
U.
Kruger
, and
P. K.
Yan
, “
Deep learning in medical image registration: A survey
,”
Mach. Vis. Appl.
31
,
118
(
2020
).
41.
J. L.
Leng
,
G. L.
Xu
, and
Y. J.
Zhang
, “
Medical image interpolation based on multi-resolution registration
,”
Comput. Math. Appl.
66
,
1
18
(
2013
).
42.
H. R.
Boveiri
,
R.
Khayami
,
R.
Javidan
, and
A.
Mehdizadeh
, “
Medical image registration using deep neural networks: A comprehensive review
,”
Comput. Electr. Eng.
87
,
106767
(
2020
).
43.
Y.
Fu
,
Y.
Lei
,
T.
Wang
,
W. J.
Curran
,
T.
Liu
, and
X.
Yang
, “
Deep learning in medical image registration: A review
,”
Phys. Med. Biol.
65
,
20TR01
(
2020
).
44.
R. W. K.
So
and
A. C. S.
Chung
, “
Learning-based non-rigid image registration using prior joint intensity distributions with graph-cuts
,” in
IEEE ICIP
(IEEE,
2011
), pp.
709
712
.
45.
G.
Wu
,
M.
Kim
,
Q.
Wang
,
B. C.
Munsell
, and
D.
Shen
, “
Scalable high performance image registration framework by unsupervised deep feature representations learning
,”
IEEE Trans. Biomed Eng.
63
,
1505
1516
(
2016
).
46.
X.
Cheng
,
L.
Zhang
, and
Y. F.
Zheng
, “
Deep similarity learning for multimodal medical images
,”
Comput. Methods Biomech. Biomed. Eng.
6
,
248252
(
2018
).
47.
J.
Yu
,
H.
Wei
,
Y.
Sui
, and
B.
Pang
, “
Bidirectional diffeomorphic network for 3d deformable medical image registration
,” in
ICME
(ICME,
2022
) pp.
209
212
.
48.
J.
Zhu
,
H.
Luo
,
C.
Wang
,
J.
Lu
, and
G.
Teng
, “
Intelligent and automatic treatment planning system for low-dose-rate brachytherapy of malignant hepatic tumors
,” in
ICIMH
(ICIMH,
2021
), pp.
47
52
.
49.
J.
Zhu
,
C.
Wang
,
S.
Teng
,
J.
Lu
,
P.
Lyu
,
P.
Zhang
,
J.
Xu
,
L.
Lu
, and
G. J.
Teng
, “
Embedding expertise knowledge into inverse treatment planning for low-dose-rate brachytherapy of hepatic malignancies
,”
Med. Phys.
51
(
1
),
348
362
(
2023
).
50.
W. B.
Luo
,
G. Z.
Wang
, and
H. S.
DING
, “
Computer aided surgery system
,”
Biomed. Eng.
24
,
241
246
(
2001
).
51.
Y. L.
Park
,
K.
Chau
,
R. J.
Black
, and
M. R.
Cutkosky
, “
Force sensing robot fingers using embedded fiber Bragg grating sensors and shape deposition manufacturing
,” in
Proceedings of the IEEE International Conference on Robotics and Automation
(IEEE,
2007
), pp.
1510
1516
.
52.
I. S. M.
Khalil
,
A.
Adel
,
D.
Mahdy
,
M. M.
Micheal
,
M.
Mansour
,
N.
Hamdi
, and
S.
Misra
, “
Magnetic localization and control of helical robots for clearing superficial blood clots
,”
APL Bioeng.
3
,
026104
(
2019
).
53.
C.
Ren
and
P.
Liang
, “
Application of intraoperative navigation in tumor thermal ablation therapy
,”
J. Med. Coll. PLA
37
,
668
670
(
2016
).
54.
G. L.
Wu
,
T. J. W.
Wu
,
S. C.
Wu
, and
Z. H.
Zhou
, “
Research on navigation and positioning method of liver tumor thermal ablation puncture robot
,”
CMD
36
,
22
25
(
2021
).
55.
A.
Khan
,
J. E.
Meyers
,
I.
Siasios
, and
J.
Pollina
, “
Next-generation robotic spine surgery: First report on feasibility and safety and learning curve
,”
Oper. Neurosurg.
17
,
61
69
(
2019
).
56.
F.
Avrumova
,
A.
Sivaganesan
,
R. K.
Alluri
,
A.
Vaishnav
,
S.
Qureshi
, and
D. R.
Lebl
, “
Workflow and efficiency of robotic-assisted navigation in spine surgery
,”
HSS J.
17
,
302
307
(
2021
).
57.
M. E.
Ivan
,
D. G.
Eichberg
,
L.
Di
,
A. H.
Shah
,
E. M.
Luther
,
V. M.
Lu
,
R. J.
Komotar
, and
T. M.
Urakov
, “
Augmented reality head-mounted display-based incision planning in cranial neurosurgery: A prospective pilot study
,”
Neurosurg. Focus
51
,
3
(
2021
).
58.
R.
Altieri
,
G.
Broggi
,
F.
Certo
,
D.
Pacella
,
G.
Cammarata
,
M.
Maione
,
M.
Garozzo
,
D.
Barbagallo
,
M.
Purrello
,
R.
Caltabiano
,
G.
Magro
, and
G.
Barbagallo
, “
Anatomical distribution of cancer stem cells between enhancing nodule and flair hyperintensity in supratentorial glioblastoma: Time to recalibrate the surgical target?
,”
Neurosurg. Rev.
45
,
3709
3716
(
2022
).
59.
Z.
Wang
,
W.
Wu
,
S.
Wu
,
Z.
Zhou
, and
H.
Zhang
, “
An automatic needle puncture path-planning method for thermal ablation of lung tumors
,”
Diagnostics
14
,
215
(
2024
).
60.
S.
Wu
,
J.
Wang
, and
P.
Gao
, “
A comparison of the efficacy and safety and duration of frame-based and remebot robot-assisted frameless stereotactic biopsy
,”
Br. J. Neurosurg.
35
,
319
323
(
2021
).
61.
R.
Elfring
,
M.
Fuente
, and
K.
Radermacher
, “
Assessment of optical localizer accuracy for computer aided surgery systems
,”
Comput. Aided Surg.
15
,
1
12
(
2010
).
62.
J.
Stolzenwald
and
W. W.
Mayol-Cuevas
, “
I can see your aim: Estimating user attention from gaze for handheld robot collaboration
,” in
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
(
IEEE
,
2018
), pp.
3897
3904
.
63.
J.
Sun
,
P.
Wang
,
Z.
Qin
, and
H.
Qiao
, “
Effective self-calibration for camera parameters and hand-eye geometry based on two feature points motions
,”
IEEE/CAA J. Autom. Sin.
4
,
370
380
(
2017
).
64.
Z.
Li
,
X. T.
Liu
,
X.
Xie
,
G. L.
Li
, and
Z. H.
Wang
, “
An optical tracker based registration method using feedback for robot-assisted insertion surgeries
,” in
ISCAS
,
2017
.
65.
X.
Li
,
H.
Zeng
,
C.
Yang
, and
A.
Song
, “
Robot-assisted eye-hand coordination training system by estimating motion direction using smooth-pursuit eye movements
,” in
Proceedings of the IEEE International Conference on Robotics and Automation (ICRA)
(
IEEE
,
2023
), pp.
9852
9857
.
66.
D.
Zhang
,
Z.
Li
,
K.
Chen
,
J.
Xiong
,
X.
Zhang
, and
L.
Wang
, “
An optical tracker based robot registration and servoing method for ultrasound guided percutaneous renal access
,”
Biomed. Eng. Online
12
,
47
(
2013
).
67.
A.
Ayadi
,
B.
Bayle
,
P.
Graebling
, and
J.
Gangloff
, “
An image-guided robot for needle insertion in small animal. accurate needle positioning using visual servoing
,” in
IEEE/RSJ International Conference on Intelligent Robots and Systems
(
IEEE
,
2008
), pp.
1453
1458
.
68.
L.
Goffin
,
G.
Bour
,
F.
Martel
,
S.
Nicolau
,
J.
Gangloff
,
J. M.
Egly
, and
B.
Bayle
, “
Design and in vivo evaluation of a robotized needle insertion system for small animals
,”
IEEE. Trans. Biomed. Eng.
60
,
2193
2204
(
2013
).
69.
Y.
Yang
,
A.
Ahmed
,
S.
Yue
,
X.
Xie
,
H.
Chen
, and
Z.
Wang
, “
An algorithm for accurate needle orientation
,” in
38th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)
,
2016
.
70.
N.
Simaan
,
R.
Yasin
, and
L.
Wang
, “
Medical technologies and challenges of robot-assisted minimally invasive intervention and diagnostics
,”
Annu. Rev. Control, Rob., Auton. Syst.
1
,
465
490
(
2018
).
71.
X.
Bao
,
S.
Wang
,
L.
Zheng
,
R. J.
Housden
,
J.
Hajnal
, and
K.
Rhode
, “
Self-adaptive parallel manipulator with pose and force adjustment for robotic ultrasonography
,”
IEEE Trans. Ind. Electron.
70
,
10333
10343
(
2023
).
72.
Z. L.
Jiang
,
Z.
Li
,
M.
Grimm
,
M.
Zhou
,
M.
Esposito
,
W.
Wein
,
W.
Stechele
,
T.
Wendler
, and
N.
Navab
, “
Autonomous robotic screening of tubular structures based only on real-time ultrasound imaging feedback
,”
IEEE Trans. Ind. Electron.
69
,
7064
7075
(
2022
).
73.
B.
Zhang
,
Y.
Kobayashi
,
T.
Chiba
, and
M. G.
Fujie
, “
Robotic patch-stabilizer using wire driven mechanism for minimally invasive fetal surgery
,” in
Annual International Conference of the IEEE Engineering in Medicine and Biology Society
(IEEE,
2009
), pp.
5076
5079
.
74.
E.
Basafa
,
P.
Foroughi
,
M.
Hoßbach
,
J.
Bhanushali
, and
P.
Stolka
, “
Visual tracking for multi-modality computer-assisted image guidance
,”
Proc. SPIE
10135
,
101352S
(
2017
).
75.
F.
Wanivenhaus
,
C.
Neuhaus
,
F.
Liebmann
,
S.
Roner
,
J. M.
Spirig
, and
M.
Farshad
, “
Augmented reality-assisted rod bending in spinal surgery
,”
Spine J.
19
,
1687
1689
(
2019
).
76.
T.
He
,
C.
Guo
,
H.
Liu
, and
L.
Jiang
, “
A venipuncture robot with decoupled position and attitude guided by near-infrared vision and force feedback
,”
Int. J. Med Rob.
19
(4),
e2512
(
2023
).
77.
T.
He
,
C.
Guo
,
H.
Liu
, and
L.
Jiang
, “
Puncture site decision method for venipuncture robot based on near-infrared vision and multi-objective optimization
,”
Sci. China Technol. Sci.
66
,
13
23
(
2023
).
78.
P.
Chatelain
,
A.
Krupa
, and
N.
Navab
, “
Optimization of ultrasound image quality via visual servoing
,” in
Proceedings of the IEEE International Conference on Robotics and Automation (ICRA)
(
IEEE
,
2015
), pp.
5997
6002
.
79.
P.
Chatelain
,
A.
Krupa
, and
N.
Navab
, “
Confifidence-driven control of an ultrasound probe
,”
IEEE Trans. Rob.
33
,
1410
1424
(
2017
).
80.
Z.
Jiang
,
M.
Grimm
,
M.
Zhou
,
J.
Esteban
,
W.
Simson
,
G.
Zahnd
, and
N.
Navab
, “
Automatic normal positioning of robotic ultrasound probe based only on confifidence map optimization and force measurement
,”
IEEE Rob. Autom. Lett.
5
,
1342
1349
(
2020
).
81.
J.
Chevrie
,
A.
Krupa
, and
M.
Babel
, “
Real-time teleoperation of flexible beveled-tip needle insertion using haptic force feedback and 3D ultrasound guidance
,” in
International Conference on Robotics and Automation (ICRA)
(
IEEE
,
2019
), pp.
2700
2706
.
82.
T.
Haidegger
,
B.
Benyó
,
L.
Kovács
, and
Z.
Benyó
, “
Force sensing and force control for surgical robots
,”
IFAC Proc.
42
,
401
406
(
2009
).
83.
G.
Zeng
and
A.
Hemami
, “
An overview of robot force control
,”
Robotica
15
,
473
482
(
1997
).
84.
A. S. B.
Mustafa
,
T.
Ishii
,
Y.
Matsunaga
,
R.
Nakadate
,
H.
Ishii
,
K.
Ogawa
,
A.
Saito
,
M.
Sugawara
,
K.
Niki
, and
A.
Takanishi
, “
Human abdomen recognition using camera and force sensor in medical robot system for automatic ultrasound scan
,” in
35th IEEE EMBC
(IEEE,
2013
) pp.
4855
4858
.
85.
X.
Ma
and
S.
Guo
, “
Remote operation by a developed robotic catheter manipulating system
,” in
WCICA
,
2014
.
86.
Q.
Huang
,
B.
Wu
,
J.
Lan
, and
X.
Li
, “
Fully automatic three-dimensional ultrasound imaging based on conventional B-scan
,”
IEEE Trans. Biomed. Circuits Syst.
12
,
426
436
(
2018
).
87.
Q.
Huang
,
J.
Lan
, and
X.
Li
, “
Robotic arm based automatic ultrasound scanning for three-dimensional imaging
,”
IEEE Trans. Ind. Inf.
15
,
1173
1182
(
2019
).
88.
S.
Ferrari
,
E.
Tagliabue
,
B. M.
Maris
, and
P.
Fiorini
, “
Autonomous robotic system for breast biopsy with deformation compensation
,”
IEEE Rob. Autom. Lett.
8
,
1215
1222
(
2023
).
89.
C.
Graumann
,
B.
Fuerst
,
C.
Hennersperger
,
F.
Bork
, and
N.
Navab
, “
Robotic ultrasound trajectory planning for volume of interest coverage
,” in
IEEE International Conference on Robotics and Automation (ICRA)
(
IEEE
,
2016
) p.
736
741
.
90.
C.
Hennersperger
,
B.
Fuerst
,
S.
Virga
,
O.
Zettinig
,
B.
Frisch
,
T.
Neff
, and
N.
Navab
, “
Towards MRI-based autonomous robotic us acquisitions: A first feasibility study
,”
IEEE Trans. Med. Imaging
36
,
538
548
(
2017
).
91.
R.
Kojcev
,
A.
Khakzar
,
B.
Fuerst
,
O.
Zettinig
,
C.
Fahkry
,
R.
DeJong
,
J.
Richmon
,
R.
Taylor
,
E.
Sinibaldi
, and
N.
Navab
, “
On the reproducibility of expert-operated and robotic ultrasound acquisitions
,”
Int. J. Comput. Assisted Radiol. Surg.
12
,
1003
1011
(
2017
).
92.
R.
Tsumura
and
H.
Iwata
, “
Robotic fetal ultrasonography platform with a passive scan mechanism
,”
Int. J. Comput. Assisted Radiol. Surg.
15
,
1323
1333
(
2020
).
93.
Y. X.
Guo
,
Y. H.
Yang
, and
L.
Xiong
, “
Double-layer orthogonal fiber Bragg gratings flexible shape sensing technology
,”
Opt. Precis. Eng.
29
,
2306
2315
(
2021
).
94.
K.
Sung
,
M.
Quo
,
G.
Yan
, and
M.
Song
, “
Bending deformation of optical fiber sensing and shape reconstruction of soft pneumatic driver
,”
Opt. Precis. Eng.
27
,
1052
1059
(
2019
).
95.
R.
Roesthuis
and
S.
Misra
, “
Steering of multisegment continuum manipulators using rigid-link modeling and FBG-based shape sensing
,”
IEEE Trans. Rob.
32
,
372
382
(
2017
).
96.
Z.
Tang
,
S.
Wang
,
M.
Li
, and
C.
Shi
, “
Development of a distal tri-axial force sensor for minimally invasive surgical palpation
,”
IEEE Trans. Med. Rob. Bionics.
4
,
145
155
(
2022
).
97.
D.
Uzun
and
O. Ü.
Kocatürk
, “
Optical force sensor with enhanced resolution for mri guided biopsy
,”
IEEE Sens. J
20
,
9202
9208
(
2020
).
98.
H.
Su
,
I. I.
Iordachita
,
J.
Tokuda
,
N.
Hata
,
X.
Liu
,
R.
Seifabadi
,
S.
Xu
,
B.
Wood
, and
G. S.
Fischer
, “
Fiber-optic force sensors for MRI-guided interventions and rehabilitation: A review
,”
IEEE Sens. J
17
,
1952
1963
(
2017
).
99.
M. K. M.
Abayazid
and
S.
Misra
, “
3D flexible needle steering in soft-tissue phantoms using fiber Bragg grating sensors
,” in
IEEE International Conference on Robotics and Automation
,
2013
.
100.
J. S.
Kim
,
J.
Guo
,
M.
Chatrasingh
,
S.
Kim
, and
I.
Iordachita
, “
Shape determination during needle insertion with curvature measurements
,” in
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
(
IEEE
,
2017
), pp.
201
208
.
101.
F.
Khan
,
A.
Donder
,
S.
Galvan
,
F. R. Y.
Baena
, and
S.
Misra
, “
Pose measurement of flexible medical instruments using fiber Bragg gratings in multi-core fiber
,”
IEEE Sens. J.
20
,
10955
10962
(
2020
).
102.
L.
Zhang
,
C.
Li
,
X.
Zhang
,
G.
Liu
,
Y.
Liu
,
J.
Zhao
,
G.
Ban
, and
Y.
Fan
, “
A new method for fiber Bragg grating based needle shape sensing calibration
,” in
IEEE ROBIO
(IEEE,
2019
), pp.
1953
1958
.
103.
W.
Lai
,
L.
Cao
,
J.
Liu
,
S. C.
Tjin
, and
S.
Phee
, “
A three-axial force sensor based on fiber Bragg gratings for surgical robots
,”
IEEE ASME Trans. Mechatron.
27
,
777
789
(
2022
).
104.
A.
Donder
and
F. R. Y.
Baena
, “
3-D path-following control for steerable needles with fiber Bragg gratings in multi-core fibers
,”
IEEE Trans. Biomed. Eng.
70
,
1072
1085
(
2023
).
105.
A. N.
Aghdam
and
X. P.
Liu
, “
A novel path planner for steerable bevel-tip needles to reach multiple targets with obstacles
,”
IEEE Trans. Instrum. Meas.
69
,
7636
7645
(
2020
).
106.
A.
Hong
,
A. J.
Petruska
,
A.
Zemmar
, and
B. J.
Nelson
, “
Magnetic control of a flexible needle in neurosurgery
,”
IEEE Trans. Biomed. Eng.
68
,
616
627
(
2021
).
107.
B.
Reed
,
M.
Okamura
, and
J.
Cowan
, “
Modeling and control of needles with torsional friction
,”
IEEE Trans. Biomed. Eng.
56
,
2905
2916
(
2009
).
108.
A.
Favaro
,
R.
Secoli
,
F.
Baena
, and
E.
Momi
, “
Model-based robust pose estimation for a multi-segment, programmable bevel-tip steerable needle
,”
IEEE Rob. Autom. Lett.
5
,
6780
6787
(
2020
).
109.
M.
Cotler
,
E.
Rousseau
,
K.
Ramadi
,
J.
Fang
,
A.
Graybiel
,
R.
Langer
, and
M.
Cima
, “
Steerable microinvasive probes for localized drug delivery to deep tissue
,”
Small
15
,
1901459
(
2019
).
110.
S. C.
Ryu
,
Z. F.
Quek
,
J.-S.
Koh
,
P.
Renaud
,
R. J.
Black
,
B.
Moslehi
,
B. L.
Daniel
,
K.-J.
Cho
, and
M. R.
Cutkosky
, “
Design of an optically controlled mr-compatible active needle
,”
IEEE Trans Rob.
31
,
111
(
2015
).
111.
A.
Hong
,
Q.
Boehler
,
R.
Moser
,
A.
Zemmar
,
L.
Stieglitz
, and
B. J.
Nelson
, “
3d path planning for flexible needle steering in neurosurgery
,”
Int. J. Med Robot
15
,
e1998
(
2019
).
112.
M.
Aggravi
,
D. A. L.
Estima
,
A.
Krupa
,
S.
Misra
, and
C.
Pacchierotti
, “
Haptic teleoperation of flexible needles combining 3D ultrasound guidance and needle tip force feedback
,”
IEEE Robot. Autom. Lett.
6
,
4859
4866
(
2021
).
113.
Y. J.
Zhao
,
C.
Wen
,
Y. D.
Zhang
, and
H.
Zhang
, “
Needle tip pose estimation for ultrasound-guided steerable flexible needle with a complicated trajectory in soft tissue
,”
IEEE Rob. Autom. Lett.
7
,
11705
11712
(
2022
).
114.
Y.
Bi
,
Z.
Jiang
,
F.
Duelmer
,
D.
Huang
, and
N.
Navab
, “
Machine learning in robotic ultrasound imaging: Challenges and perspectives
,”
Annu. Rev. Control, Rob., Auton. Syst.
Vol.
7
(
2024
).
115.
R.
Kojcev
,
A.
Khakzar
,
B.
Fuerst
,
O.
Zettinig
,
C.
Fahkry
,
R.
DeJong
,
J.
Richmon
,
R.
Taylor
,
E.
Sinibaldi
, and
N.
Navab
, “
On the reproducibility of expert-operated and robotic ultrasound acquisitions
,”
Int. J. Comput. Assisted Radiol. Surg.
12
,
1003
1011
(
2017
).
116.
Z. L.
Jiang
,
Y.
Gao
,
L.
Xie
, and
N.
Navab
, “
Towards autonomous atlas-based ultrasound acquisitions in presence of articulated motion
,”
IEEE Robot. Autom. Lett.
7
,
7423
7430
(
2022
).
117.
Z. L.
Jiang
,
N.
Danis
,
Y.
Bi
,
M. C.
Zhou
,
M.
Kroenke
,
T.
Wendler
, and
N.
Navab
, “
Precise repositioning of robotic ultrasound: Improving registration-based motion compensation using ultrasound confidence optimization
,”
IEEE Trans. Instrum. Meas.
71
,
5020611
(
2022
).
118.
Z.
Jiang
et al, “
Deformation-aware robotic 3D ultrasound
,”
IEEE Rob. Autom. Lett.
6
,
7675
7682
(
2021
).
119.
R.
Kojcev
,
B.
Fuerst
,
O.
Zettinig
,
J.
Fotouhi
,
S. C.
Lee
,
B.
Frisch
,
R.
Taylor
,
E.
Sinibaldi
, and
N.
Navab
, “
Dual-robot ultrasound-guided needle placement: Closing the planning-imaging-action loop
,”
Int. J. Comput. Assisted Radiol. Surg.
11
,
1173
1181
(
2016
).
120.
G.
Yang
,
R.
Full
,
N.
Jacobstein
,
P.
Fischer
,
J.
Bellingham
,
H.
Choset
,
H.
Christensen
,
P.
Dario
,
B.
Nelson
, and
R.
Taylor
, “
Ten robotics technologies of the year
,”
Sci. Robot.
4, eaaw1826 (2019).
121.
P.
Mountney
,
D.
Stoyanov
, and
G.
Yang
, “
Three-dimensional tissue deformation recovery and tracking
,”
IEEE Signal Process. Mag.
27
,
14
24
(
2010
).
122.
M.
Horvath
,
L.
Hu
,
T.
Mueller
,
J.
Hochstein
,
L.
Rosalia
,
K.
Hibbert
,
C.
Hardin
, and
E.
Roche
, “
An organosynthetic soft robotic respiratory simulator
,”
APL Bioeng.
4
,
026108
(
2020
).
123.
M.
Seregni
,
A.
Pella
,
M.
Riboldi
,
R.
Orecchia
,
P.
Cerveri
, and
G.
Baroni
, “
Real-time tumor tracking with an artificial neural networks-based method: A feasibility study
,”
Phys. Med.
29
,
48
59
(
2013
).
124.
C.
Simone
and
A.
Okamura
, “
Modeling of needle insertion forces for robot-assisted percutaneous therapy
,” in
Proceedings of theIEEE International Conference on Robotics and Automation (Cat. No.02CH37292)
(IEEE,
2002
), Vol.
2
, pp.
2085
2091
.
125.
A.
Okamura
,
C.
Simone
, and
M.
O'Leary
, “
Force modeling for needle insertion into soft tissue
,”
IEEE. Trans. Biomed. Eng.
51
,
1707
1716
(
2004
).
126.
F.
Ernst
,
V.
Martens
,
S.
Schlichting
,
A.
Besirević
,
M.
Kleemann
,
C.
Koch
,
D.
Petersen
, and
A.
Schweikard
, “
Correlating chest surface motion to motion of the liver using epsilon-SVR—A porcine study
,” in
MICCAI
(MICCAI,
2009
), pp.
356
364
.
127.
Y.
Ma
,
Z.
Yang
,
W.
Wu
,
H.
Xie
, and
L.
Gu
, “
Target localization during respiration motion based on lstm: A pilot study on robotic puncture system
,”
MRCAS
17
,
e2247
(
2021
).
128.
D.
Chen
,
H.
Xie
,
L.
Gu
,
W.
Guo
,
L.
Tian
, and
J.
Liu
, “
A statistical weighted sparse-based local lung motion modelling approach for model-driven lung biopsy
,”
Int. J. Comput. Assisted Radiol. Surg.
15
,
1279
1290
(
2020
).
129.
S.
Ahn
,
B.
Yi
,
Y.
Suh
,
J.
Kim
,
S.
Lee
,
S.
Shin
,
S.
Shin
, and
E.
Choi
, “
A feasibility study on the prediction of tumour location in the lung from skin motion
,”
Br. J. Radiol.
77
,
588
596
(
2004
).
130.
J. D. P.
Hoisak
,
K. E.
Sixel
,
R.
Tirona
,
P. C. F.
Cheung
, and
J. P.
Pignol
, “
Correlation of lung tumor motion with external surrogate indicators of respiration
,”
Int. J. Radiat. Oncol., Biol., Phys.
60
,
1298
1306
(
2004
).
131.
D. A.
Low
,
P. J.
Parikh
,
W.
Lu
,
J. F.
Dempsey
,
S. H.
Wahab
,
J. P.
Hubenschmidt
,
M. M.
Nystrom
,
M.
Handoko
, and
J. D.
Bradley
, “
Novel breathing motion model for radiotherapy
,”
Int. J. Radiat. Oncol., Biol., Phys.
63
,
921
929
(
2005
).
132.
Y.
Zhou
,
K.
Thiruvalluvan
,
L.
Krzeminski
,
W.
Moore
,
Z.
Xu
, and
Z.
Liang
, “
CT-guided robotic needle biopsy of lung nodules with respiratory motion—Experimental system and preliminary test
,”
MRCAS
9
,
317
330
(
2013
).
133.
X.
Liang
,
S.
Lin
,
F.
Liu
,
D.
Schreiber
, and
M.
Yip
, “
ORRN: An ODE-based recursive registration network for deformable respiratory motion estimation with lung 4DCT images
,”
IEEE Trans. Biomed. Eng.
1–12
,
3265
3276
(
2023
).
134.
Y.
Tsunashima
,
T.
Sakae
,
Y.
Shioyama
,
K.
Kagei
,
T.
Terunuma
,
A.
Nohtomi
, and
Y.
Akine
, “
Correlation between the respiratory waveform measured using a respiratory sensor and 3D tumor motion in gated radio-therapy
,”
Int. J. Radiat. Oncol., Biol., Phys.
60
,
951
958
(
2004
).
135.
P.
Su
,
J.
Yang
,
K.
Lu
,
N.
Yu
,
S.
Wong
, and
Z.
Xue
, “
A fast CT and CT-fluoroscopy registration algorithm with respiratory motion compensation for image-guided lung intervention
,”
IEEE. Trans. Biomed. Eng.
60
,
2034
2041
(
2013
).
136.
J.
Wasza
,
P.
Fischer
,
H.
Leutheuser
,
T.
Oefner
,
C.
Bert
,
A.
Maier
, and
J.
Hornegger
, “
Real-time respiratory motion analysis using 4-D shape priors
,”
IEEE. Trans. Biomed. Eng.
63
,
485
495
(
2016
).
137.
C.
Weon
,
M.
Kim
,
C. M.
Park
, and
J. B.
Ra
, “
Real-time respiratory phase matching between 2D fluoroscopic images and 3D CT images for precise percutaneous lung biopsy
,”
Med. Phys.
44
,
5824
5834
(
2017
).
138.
A.
Schweikard
,
G.
Glosser
,
M.
Bodduluri
,
M. J.
Murphy
, and
J. R.
Adler
, “
Robotic motion compensation for respiratory movement during radiosurgery
,”
Comput. Aided. Surg.
5
,
263
277
(
2000
).
139.
M.
Murphy
,
M.
Isaakson
, and
J.
Jalden
, “
Adaptive filtering to predict lung tumor motion during free breathing
,”in
Computer Assisted Radiology and Surgery
(CARS,
2002
), pp.
539
544
.
140.
I.
Buzurovic
, “
Robotic tumor tracking techniques in radiation therapy
,”
Adv. Rob. Autom.
1
,
e102
(
2012
).
141.
A. L.
Gunderman
,
M.
Musa
,
B. O.
Gunderman
,
F.
Banovac
,
K.
Cleary
,
X.
Yang
, and
Y.
Chen
, “
Autonomous respiratory motion compensated robot for CT-guided abdominal radiofrequency ablations
,”
IEEE Trans. Med. Rob. Bionics
5
,
206
217
(
2023
).
142.
Y.
Seppenwoolde
,
R.
Berbeco
,
S.
Nishioka
,
H.
Shirato
, and
B.
Heijmen
, “
Accuracy of tumor motion compensation algorithm from a robotic respiratory tracking system: A simulation study
,”
Med. Phys.
34
,
2774
2784
(
2007
).
143.
M.
Hoogeman
,
J.
Prévost
,
J.
Nuyttens
,
J.
Pöll
,
P.
Levendag
, and
B.
Heijmen
, “
Clinical accuracy of the respiratory tumor tracking system of the cyberknife: Assessment by analysis of log files
,”
Int. J. Radiat. Oncol., Biol., Phys.
74
,
297
303
(
2009
).
144.
I.
Buzurovic
,
K.
Huang
,
Y.
Yu
, and
T. K.
Podder
, “
A robotic approach to 4D real-time tumor tracking for radiotherapy
,”
Phys. Med. Biol.
56
,
1299
1318
(
2011
).
145.
U.
Wijenayake
and
S. Y.
Park
, “
Real-time external respiratory motion measuring technique using an RGB-D camera and principal component analysis
,”
Sensors
17
,
1840
(
2017
).
146.
J.
Wang
,
R.
Sun
,
S.
Yu
,
F.
Zhang
, and
S.
Lining
, “
An improved correlation model for respiration tracking in robotic radiosurgery using essential skin surface motion
,”
IEEE Rob. Autom. Lett.
6
,
7885
7892
(
2021
).
147.
B.
Li
,
P.
Li
,
R.
Sun
,
S.
Yu
,
H.
Liu
,
L.
Sun
, and
Y.
Liu
, “
Respiratory motion estimation of tumor using point clouds of skin surface
,”
IEEE Trans. Instrum. Meas.
72
,
1
13
(
2023
).
148.
S.
Yoon
,
W.
Kim
, and
M.
Lee
, “
Design of bilateral control for force feedback in surgical robot
,”
Int. J. Controlled Autom. Syst.
13
,
916
925
(
2015
).
149.
C.
Tao
,
L.
Fengming
,
S.
Rui
,
W.
Yanhong
, and
F.
Tianyu
, “
A robotic collaborative sewing system based on visual and force perception
,”
Robot
44
,
352
360
(
2022
) .
150.
J.
Han
,
J.
Chen
,
N.
Yu
, and
Y.
Dang
, “
Active modeling and control of the ring-shaped pneumatic actuator: An experimental study
,”
IEEE/ASME Trans. Mech.
27
,
2918
2929
(
2022
).
151.
C.
Panǎ
,
V.
Rǎdulescu
,
D.
Voicilǎ
,
D.
Pǎtraşcu-Panǎ
, and
I.
Reşceanu
, “
Position control of active arms of da Vinci robotic surgical system
,” in
24th ICCC
,
2023
.
152.
S.
Dong
,
S.
Wang
,
Y.
She
,
N.
Sunil
,
A.
Rodriguez
, and
E.
Adelson
, “
Cable manipulation with a tactile-reactive gripper
,”
Int. J. Rob. Res.
40
,
1385
1401
(
2021
).
153.
S.
Saini
,
M.
Orlando
, and
P.
Pathak
, “
Adaptive control of a master-slave based robotic surgical system with haptic feedback
,”
IEEE Trans. Autom. Sci. Eng.
20
,
1125
1138
(
2023
).
154.
Z.
Lin
,
T.
Zhang
,
Z.
Sun
,
H.
Gao
,
X.
Ai
,
W.
Chen
,
G.
Yang
, and
A.
Gao
, “
Robotic telepresence based on augmented reality and human motion mapping for interventional medicine
,”
IEEE Trans. Med. Rob. Bionics
4
,
935
944
(
2022
).
155.
Y.
Yokokohji
and
T.
Yoshikawa
, “
Bilateral control of master-slave manipulators for ideal kinesthetic coupling-formulation and experiment
,”
IEEE Trans. Rob.
10
,
605
620
(
1994
).
156.
J.
Zhao
,
S.
Wang
,
L.
Cui
,
J.
Li
, and
C.
Shi
, “
Design and experimental validation of a master-slave robotic system for flexible ureteroscopy
,” in
IEEE ICMA
(IEEE,
2020
), pp.
1733
1738
.
157.
K.
Kuchenbecker
and
G.
Niemeyer
, “
Modeling induced master motion in force-reflecting teleoperation
,” in
Proceedings of the IEEE International Conference on Robotics and Automation
(
IEEE
,
2005
), pp.
348
353
.
158.
X.
Gao
,
Y.
Wang
,
S.
Jingzhou
,
Q.
Jia
, and
H.
Sun
, “
Research of a new 6-DOF force feedback hand controller system
,”
J. Robot
2014
,
646574
.
159.
A.
Takhmar
,
I.
Polushin
,
A.
Talasaz
, and
R.
Patel
, “
Cooperative teleoperation with projection-based force reflection for MIS
,”
IEEE Trans. Control Syst. Technol.
23
,
1411
1426
(
2015
).
160.
F.
Yili
,
N.
Guojun
,
P.
Bo
, and
Q.
Cuicui
, “
Master-slave control of the celiac minimally invasive surgical robot
,”
Robot
41
,
551
560
(
2019
).
161.
Y.
Yan
,
H.
Wang
,
H.
Yu
,
F.
Wang
,
J.
Fang
,
J.
Niu
, and
S.
Guo
, “
Machine learning-based surgical state perception and collaborative control for a vascular interventional robot
,”
IEEE Sens. J.
22
,
7106
7118
(
2022
).
162.
J.
Shi
, “
Variable universe fuzzy PID control for minimally invasive operation robotic catheter system
,”
Appl. Mech. Mater
618
,
470
474
(
2014
).
163.
Z.
Du
,
Y.
Liang
,
Z.
Yan
,
L.
Sun
, and
W.
Chen
, “
Human-robot interaction control of a haptic master manipulator used in laparoscopic minimally invasive surgical robot system
,”
Mech. Mach. Theory
156
,
104132
(
2021
).
164.
H.
Liang
,
L.
Cong
,
N.
Hendrich
,
S.
Li
,
F.
Sun
, and
J.
Zhang
, “
Multifingered grasping based on multimodal reinforcement learning
,”
IEEE Robot. Autom. Lett.
7
,
1174
1181
(
2022
).
165.
A.
Saracino
et al., “
Haptic feedback in the da Vinci research kit (dVRK): A user study based on grasping, palpation, and incision tasks
,”
Int. J. Med. Rob. Comput. Assisted Surg.
15
,
e1999
(
2019
).
166.
K.
Zareinia
,
Y.
Maddahi
,
L.
Gan
,
A.
Ghasemloonia
,
S.
Lama
,
T.
Sugiyama
,
W.
Yang
, and
R.
Sutherland
, “
A force-sensing bipolar forceps to quantify tool-tissue interaction forces in microsurgery
,”
IEEE/ASME Trans. Mechatronics
21
,
2365
2377
(
2016
).
167.
W.
Wang
,
J.
Wang
,
Y.
Luo
,
X.
Wang
, and
H.
Son
, “
A survey on force sensing techniques in robot-assisted minimally invasive surgery
,”
IEEE Trans. Haptics
16
,
702
718
(
2023
).
168.
K.
Wu
,
B.
Li
,
Y.
Zhang
, and
X.
Dai
, “
Review of research on path planning and control methods of flexible steerable needle puncture robot
,”
Comput. Aided Surg.
27
,
91
112
(
2022
).
169.
D.
Glozman
and
M.
Shoham
, “
Image-guided robotic flexible needle steering
,”
IEEE Trans Rob.
23
,
459
467
(
2007
).
170.
K.
Hauser
,
R.
Alterovitz
,
N.
Chentanez
,
A.
Okamura
, and
K.
Goldberg
, “
Feedback control for steering needles through 3D deformable tissue using helical paths
,”
Robotics: Science and Systems.
37
, 1–23 (
2009
).
171.
V.
Kallem
and
N.
Cowan
, “
Image guidance of flexible tip-steerable needles
,”
IEEE Trans Rob.
25
,
191
196
(
2009
).
172.
X.
Zhao
,
H.
Guo
,
D.
Ye
, and
B.
Huo
, “
Comparison of estimation and control methods for flexible needle in 2D
,” in
CCDC
(CCDC,
2016
) pp.
5444
5449
.
173.
Y.
Kobayashi
,
A.
Onishi
, and
H.
Watanabe
, “
Development of an integrated needle insertion system with image guidance and deformation simulation
,”
Comput. Med. Imaging Graph.
34
,
9
18
(
2010
).
174.
P.
Carvalho
,
A.
Kesari
,
S.
Weaver
,
P.
Flaherty
, and
G. S.
Fischer
, “
Robotic assistive device for phlebotomy
,” in
ASME IDETC-CIE
,
2015
.
175.
O. S.
O
,
Y. M.
C
, and
B.
O
, “
Design and modeling of a parallel robot for ultrasound guided percutaneous needle interventions
,” in
41st IEEE ICIEA
(IEEE,
2015
).
176.
S.
Chen
,
Y.
Lin
,
Z.
Li
,
F.
Wang
, and
Q.
Cao
, “
Automatic and accurate needle detection in 2D ultrasound during robot-assisted needle insertion process
,”
Int. J. Comput. Assited Rad
17
,
295
303
(
2022
).
177.
M. L.
Balter
,
A. I.
Chen
,
T. J.
Maguire
, and
M. L.
Yarmush
, “
The system design and evaluation of a 7D of image-guided venipuncture robot
,”
IEEE Trans. Rob.
31
,
1044
1053
(
2015
).
178.
B.
Fallahi
,
C.
Rossa
,
R.
Sloboda
,
N.
Usmani
, and
M.
Tavakoli
, “
Partial estimation of needle tip orientation in generalized coordinates in ultrasound image-guided needle insertion
,” in
IEEE AIM
(IEEE,
2016
), pp.
1604
1609
.
179.
P.
Mignon
,
P.
Poignet
, and
J.
Troccaz
, “
Automatic robotic steering of flexible needles from 3d ultrasound images in phantoms and ex vivo biological tissue
,”
Ann. Biomed Eng
46
,
1385
1396
(
2018
).
180.
J.
Chevrie
,
N.
Shahriari
,
M.
Babel
,
A.
Krupa
, and
S.
Misra
, “
Flexible needle steering in moving biological tissue with motion compensation using ultrasound and force feedback
,”
IEEE Rob. Autom. Lett.
3
,
2338
2345
(
2018
).
181.
B.
Zhao
,
S.
Shao
,
L.
Lei
,
X.
Wang
,
X.
Yang
,
Q.
Wang
, and
Y.
Hu
, “
Curve fitting-based dynamic path planning and tracking control for flexible needle insertion
,”
IEEE Trans. Med. Robot. Bionics
4
,
436
447
(
2022
).