A method for robotic grasping based on improved Gaussian mixture model.

The present research envisages a method for the robotic grasping based on the improved Gaussian mixture model. The improved Gaussian mixture model is a method proposed by incorporating Bayesian ideas into the Gaussian model. It will use the Gaussian model to perform grasping training in a certain area which we called trained area. The improved Gaussian models utilized the trained Gaussian models as prior models. The proposed method improved the cumulative updates and the evaluation results of the improved models to make robots more adaptable to grasp in the untrained areas. The self-taught learning ability of the robot about grasping was semi-supervised. Firstly, the observable variables of objects were determined by a camera. Then, we dragged the robot to grasp object. The relationship between the variables and robot's joint angles were mapped. We obtained new samples in the close untrained area to improve the Gaussian model. With these new observable variables, the robot grasped it successfully. Finally, the effectiveness of the method was verified by experiments and comparative tests on grasping of real objects and grasping simulation of the improved Gaussian models through the virtual robot experimentation platform.

A shape memory alloy-actuated surgical instrument with compact volume.

BACKGROUND: Surgical robotic systems have been proven to be accurate and dexterous during minimally invasive surgery (MIS). However, single incision laparoscopic surgery (SILS) requires more compact robotic arms with sufficient dexterity and output. METHODS: A shape memory alloy (SMA)-actuated instrument was adopted to assist operations. The instrument fixed at the slave site had three degrees of freedom (DOFs) and was driven by SMA wires, pitch, yaw, and grip, with an 8 mm diameter; a passive hydraulic support mechanism was placed in vitro. RESULTS: In vitro, the maximum force, velocity and accuracy were proven to be sufficient for medical application. In an animal study, the system was inserted into the abdomen; three more generic constraints were tested: medical gesture, safety, and implementation. CONCLUSIONS: The robot was feasible and safe for multi-angle monitoring, grasping, and holding an organ. The SMA-actuated system was accurate and dexterous with minimal system volume.

[Percutaneous pelvic fracture stabilization using CT-based 3D navigation software combined with targeting mechanical arm: a cadaver model trial].

OBJECTIVES: To investigate a new targeting mechanical arm for CT-based navigated percutaneous fixation of pelvic fractures, and to evaluate the safety and efficiency of the procedures. METHODS: Using CT-based 3D navigation software combined with targeting mechanical arm, percutaneous insertion of pelvic models (3 dry human cadaver pelvic skeletons and 5 plastic Sybone pelvic models) were performed, 8 pelvic models allowed percutaneous cannulated screw insertion of both S-I joint (2 S-I screws placement for each side, total 32 screws in this experiment) and both superior ramus (1 ramus medullary screw placement for each side, total 16 screws in this experiment). Percutaneous insertion of pelvic models (4 dry human cadaver pelvic skeletons and 4 plastic Sybone pelvic models, 1 S-I screws and 1 ramus medullary scre placement for each side, 32 screws in this experiment) were performed using fluoro-navigation system (Stryker, USA). Time necessary for every screw insertion were recorded. Accuracy of screw placement was assessed using C-arm imaging and direct eyes inspecting. The time and accuracy of the two methods were compared. RESULTS: The time required for the CT-based 3D navigation procedure (3.6 ± 1.2) min was significantly less than using the targeting mechanical arm compared to drilling freehand with navigation (9.1 ± 0.8) min (t = 2.50, P < 0.01). There was no significant difference in accuracy between the two methods. CONCLUSION: CT-based 3D navigation software combined with targeting mechanical arm should be potential to apply percutaneous sacroiliac screwing for pelvic fractures with more accurate and more reliable.

An accurately controlled antagonistic shape memory alloy actuator with self-sensing.

With the progress of miniaturization, shape memory alloy (SMA) actuators exhibit high energy density, self-sensing ability and ease of fabrication, which make them well suited for practical applications. This paper presents a self-sensing controlled actuator drive that was designed using antagonistic pairs of SMA wires. Under a certain pre-strain and duty cycle, the stress between two wires becomes constant. Meanwhile, the strain to resistance curve can minimize the hysteresis gap between the heating and the cooling paths. The curves of both wires are then modeled by fitting polynomials such that the measured resistance can be used directly to determine the difference between the testing values and the target strain. The hysteresis model of strains to duty cycle difference has been used as compensation. Accurate control is demonstrated through step response and sinusoidal tracking. The experimental results show that, under a combination control program, the root-mean-square error can be reduced to 1.093%. The limited bandwidth of the frequency is estimated to be 0.15 Hz. Two sets of instruments with three degrees of freedom are illustrated to show how this type actuator could be potentially implemented.

Application study of medical robots in vascular intervention.

BACKGROUND: Based on the background of minimally invasive surgery and applications of medical robots, a vascular interventional robotic system has been developed that can be used in the field of vascular intervention. METHODS: The robotic system comprises a propulsion system, an image navigation system and a virtual surgery training system. Integration of the three systems constitutes a vascular intervention prototype robotic system used to carry out in vitro vascular intervention and animal experiments. RESULTS: On a transparent glass vascular model, a catheter was shown to enter an arbitrary branch of the vascular model with catheter motion meeting the requirements of clinical vascular intervention surgery (VIS); i.e. error band of catheter motion < 0.5 mm. In the animal experiments, 1.33-2.00 mm (4F-6F) diameter catheters were selectively inserted successfully into predefined targets in the animal, such as the renal, cardiovascular and cerebrovascular artery. Compared with conventional manual surgery, the time for robotic surgery is a little longer. There were no operative complications in the animal experiments. CONCLUSIONS: These simulation and animal study results demonstrate that this vascular interventional robotic system allows doctors to perform angiography remotely and prevents them from radiation exposure. The system may be the basis for further clinical applications of vascular intervention.

Computer-assisted navigation systems for insertion of cannulated screws in femoral neck fractures: a comparison of bi-planar robot navigation with optoelectronic navigation in a Synbone hip model trial.

BACKGROUND: Computer-assisted procedures have recently been introduced for navigated femoral neck screw placement. Currently there is little information available regarding accuracy and efficiency of the different navigated procedures. The aim of this study was to compare two fluoroscopic navigation tracking technologies, a novel bi-planar robot navigation and standardized optoelectronic navigation, versus standard freehand fluoroscopic insertion in a Synbone hip model. METHODS: Eighteen fixed Synbone hip models were divided into 3 groups. C-arm navigated cannulated screws (AO-ASIF, diameter 7.3 mm) were inserted using freehand targeting (control group). A novel bi-planar robot system (TINAV, GD2000) and an optoelectronic system (Stryker OTS Navigation System) were used for the navigated procedures (robot group and optoelectronic group). Accuracy was measured using radiographic evaluation including the measurement of screw parallelism and decentralization, and joint penetration. To evaluate the efficiency, the number of guidewire passes, operative time and fluoroscopic images taken were noted. RESULTS: The two computer-assisted systems provided significantly improved accuracy compared to the freehand technique. Each of the parameters, including guidewire passes and number of fluoroscopy images, was significantly lower when using the computer-assisted systems than for freehand-unguided insertion (P <0.05), but operative time was significantly shorter when using freehand-unguided insertion than for the computer-assisted systems (P <0.05). Accuracy, operative time and number of fluoroscopy images taken were similar among the two navigated groups (P >0.05), but guidewire passes in the robot group were significantly less than in the optoelectronic group (P <0.05). CONCLUSIONS: Both bi-planar robot navigation and optoelectronic navigation were similarly accurate and have the potential to improve accuracy and reduce radiation for freehand fluoroscopic targeting for insertion of cannulated screws in femoral neck fractures. Guidewire passes in the robot group were significantly less than in the optoelectronic group. However, both navigated procedures were associated with time-consuming registration and high rates of failed matching procedures.

[A perspective on medical robotics].

Medical robotics has played an increasingly important role in the recent years in robotics field. An overview of the minimally invasive surgery robots, rehabilitative robots and hospital service robots are offered here, and relevant tendencies are also referred to.

[Clinical application of robotic tele-manipulation system in stereotactic surgery].

OBJECTIVE: To assess the clinical usefulness, accuracy, and safety of tele-manipulation for frameless stereotactic surgery using the CAS-R-5 robot system. METHODS: We prospectively evaluated 32 patients underwent tele-manipulation of frameless stereotactic operations from Sep. 2005 to Sep. 2006. Tele-manipulations were performed via a digital data network by a neurosurgeon in Beijing while the patients were located in Yan'an. The distance is 1300 kilometers away. The accuracy of location and improvement of symptom were observed after operation. The period of follow-up was from 3 to 14 months (the average was 12 months). RESULTS: The surgical operations in 32 cases were successful. Remote fiducial registration was performed with a mean accuracy of 1. 50 mm and the standard difference were 0.32 mm between the planned and actual target. There were no complications. CONCLUSIONS: Diagnosis and treatment for intracranial disease by tele-manipulation frameless stereotactic surgeries are reliable and safe.

[Accurate traction of long bone fracture with full-length planning module of orthopedic robot system: experiments in vitro and in vivo].

OBJECTIVE: To evaluate the precision in location and clinical flexibility of the newly designed full-length planning module of orthopedic robot system in treatment of fractures of long bone. METHODS: Nine plastic tibia models were selected for the image mosaicing. The full length of each tibia model was measured on the constructed panorama and compared with the real model length to record the length deviation and conduct the precision analysis. Fracture of tibia and fibula with shortening and angulation deformity was caused on a cadaver specimen with two lower limbs. Full-length planning was carried out on the entire tibial panorama with the fracture. After the reduction distance was determined quantitatively, automatic close traction procedure was carried out with the tibial reduction frame to analyze the precision and effectiveness of this module. At the same time, the relative length variation between the two bone fragments was monitored utilizing video camera to ensure the safety of the reduction operation. Image mosaicing, surgical planning, and bone traction were performed on a clinical case of tibial fracture to validate the clinical feasibility of the module. RESULTS: An entire tibial panorama could be constructed from 7 - 10 C-arm images collected during the operation. 1.5 min was needed for image collection. The average mosaicing and planning time was 3 min. The mosaicing error was less than 1.5 mm. The average time for the traction frame installation and traction operation was 4 min. Traction resulted in accurate reposition of the fracture ends meeting the requirement of surgery in both the cadaver specimen and the clinical case. CONCLUSION: The newly designed full-length planning module of orthopedic robot system is easy to use and provides effective and accurate traction result in long bone fracture therapy. This module can not only achieve the minimally invasive surgery, but also dramatically decrease the radiation damage to the medical staff.

[Intra-operative stereotactic accuracy of computer-assisted robot orthopaedic trauma surgery planning system].

OBJECTIVE: To investigate the feasibility and accuracy of the software computer-assisted robot orthopaedic trauma surgery planning system (CAOTS). METHODS: The software CAOTS was developed and used on 85 cases of stereotactic operation, 24 model bones, 21 cadaveric extremity bones, and 40 patients. 307 groups of directional targets in these cases were randomly sampled. The distribution of error sources for evaluating system performance was embodied using Monte-Carlo method in order to derive the theoretic guarantees for further optimizing and enhancing the system performance, then the software SPSS 2.0 was used to analyze the errors. RESULTS: The statistical area of deviation number was 0.0408 +/- 0.4578 mm, corresponding to the result by Monte-Carlo method. Punching succeeded at the first run for all 307 cases without wrong locking and other clinical complications. CONCLUSION: Accurate and reliable, CAOTS improves the intra-operative navigation techniques and facilitates the orthopedists to perform operation.

[Effects of medical robot-assisted surgical navigation system in distal locking of femoral intramedullary nails: an experimental study].

OBJECTIVE: To investigate the effects of medical robot-assisted surgical navigation system based on fluoroscopic images in distal locking of femoral intramedullary nails. METHODS: Using a robot-assisted computer-guided system designed based on modularization and minimization that permitted C-arm alignment assistance and real-time navigation control, provided constant feedback without the need for radiologic updates, thus avoiding constant X-ray exposure. The C-arm was used to collect the orthotopic and lateral X-ray images into the computer so as to calculate the locations of the target points. Nails were locked into 5 plastic femurs (Swiss Sybone, 35 holes), 2 dry human femoral specimens (12 holes), and one leg of fresh human cadaver (6 holes). Radiographs were taken to confirm that screws were positioned correctly, and fluoroscopic time associated with the locking procedure was recorded. RESULTS: All distal holes were locked successfully. In 6 (11.1%) of the 53 holes the drill bit touched the canal of the locking hole, albeit with no damage to the nail. The fluoroscopy time of per screw was 1.83 +/- 0.31 seconds. CONCLUSION: The medical robot-assisted surgical navigation system enables the physicians to precisely navigate surgical instruments throughout the procedure using just a few computer-calibrated radiographic images. The total radiation time per procedure can be significantly reduced because additional X-ray exposure is not required for tool navigation. The idea of a robot-assisted surgical navigation system is practicable.

Computer-assisted auto-frame navigation system for distal locking of tibial intramedullary nails: a preliminary report on clinical application.

OBJECTIVE: To evaluate the clinical feasibility and effect of the computer-assisted auto-frame navigation system for distal locking of tibial intramedullary nails. METHODS: The hardware components of the system included a PC computer with a monitor, auto mechanical stereotactical localization cubic frame, foot holder and localization operative apparatus. Special navigation software can be used for registration of X-ray fluoroscopic images and real-time controlling navigation of tools. Twenty-one cases of close tibial and fibular fractures were treated with closed intramedullary nailing, 6 of which involved in middle third, 12 in middle and lower third, 3 in lower third. C-arm alignment and registration time, fluoroscopic time and drilling time involved in the locking procedure were recorded. The size of unreamed or reamed tibial nails ranged from 8/300-11/330. RESULTS: All distal holes except 1 were locked successfully. In 9 of 41 locked holes (21.95%), the drill bit touched the canal of locking hole without damage of the nail and clinical consequences. The fluoroscopy time per pair of screws was 2.23+/-0.31 s. CONCLUSIONS: The computer-assisted auto-frame navigation system for distal locking is well designed, easy to operate and do not need additional instruments during the procedure. The developed system enables the physician to precisely navigate surgical instruments throughout the anatomy using just a few computer-calibrated radiographic images. The total time of X-ray exposure per procedure can be significantly reduced.

[Design and experimental study of computer-assisted orthopaedic system for distal locking of intramedullary nails].

OBJECTIVE: To test the safety and accuracy of the computer-assisted orthopaedic system for distal locking of intramedullary nails and apply it to internal fixation with intramedullary nails in the lower limb. METHODS: According to the theory of mechanical arms stereotactic localization in computer-assisted orthopaedic surgery (CAOS), we design a CAOS system for distal locking of intramedullary nails. The system comprised 2 independent modules: computer-assisted imaging and registration workstation; mechanical stereotactic framework. Ten plastic tibia models, 20 plastic femur models (Synbone AG, Malans, Switzerland) and 6 human cadaver lower limbs were randomly divided into 2 groups undergoing internal fixation with intramedullary nails (Orthofix, Germany). The first group (CAOS group with 5 plastic tibia models, 10 plastic femur models, 6 human cadaver tibia, 6 human cadaver femur; each nail had 2 holes, and 2 distal locking screws were inserted in each bone, which gave a total number of 54 holes) used a computer-assisted orthopaedic system, the second group (CONTROL GROUP is the same as CAOS group) used Orthofix mechanical targeting device for distal locking. Comparison between 2 groups was made in radiation exposure time, operating time, percentage of correctly placed screws. RESULTS: CAOS group: operating time was (4.44 +/- 2.99) min; radiation exposure time was (1.16 +/- 0.38) min; correctly placed screws rate was (100 +/- 0)%. CONTROL GROUP: operating time was (10.42 +/- 4.18) min; radiation exposure time was (4.71 +/- 3.86) min; correctly placed screws rate was (94.44 +/- 0.36)%. Operating time and radiation exposure time in CAOS group were significantly shorter than those in control group (P < 0.05), no differences were found between 2 groups in relation to the percentage of correctly placed screws. CONCLUSIONS: By using CAOS system for distal locking of intramedullary nails, the locking holes can be drilled accurately and safely. Radiation exposure significantly reduced.