THE INTERNATIONAL JOURNAL OF MEDICAL ROBOTICS AND COMPUTER ASSISTED SURGERY ORIGINAL Int J Med Robotics Comput Assist Surg 2008; 4: 304–309. Published online 22 September 2008 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/rcs.210
ARTICLE
Pre-clinical remote telesurgery trial of a da Vinci telesurgery prototype
Christopher Nguan1 * Brian Miller2 Rajni Patel3 Patrick PW Luke3 Christopher M. Schlachta3 1
Vancouver General Hospital, 6th floor, 2775 Laurel Street, Vancouver, British Columbia V5Z 1M9, Canada 2 Intuitive Surgical Inc., 950 Kifer Road, Sunnyvale, CA 94086, USA
Abstract Background The objective of this study was to perform a pre-clinical remote telesurgery trial of a da Vinci telesurgery prototype on a surgical-grade virtual private network. Methods A da Vinci telesurgery-enabled prototype was used to conduct surgical trials across a 17 MB/s bandwidth VPNe network spanning 2848 km round-trip landline distance from London, Ontario, to Halifax, Nova Scotia, Canada. The outcomes measured during the trial were surgical times and quality of anastomoses.
3
Canadian Surgical Technologies and Advanced Robotics, 339 Windermere Road, London, Ontario N6A 5A5, Canada *Correspondence to: Christopher Nguan, Vancouver General Hospital, 6th floor, 2775 Laurel Street, Vancouver, British Columbia V5Z 1M9, Canada. E-mail:
[email protected]
Results Network configuration resulted in observed latencies of 370 ms with 140 ms due to transport delay. The da Vinci telesurgery prototype performed well subjectively and average porcine pyeloplasty anastomotic times were 20.7 ± 4.7 min. Conclusions This work constitutes an early evaluation of the da Vinci telesurgery prototype for conceptually remote telesurgical operations. This study clearly demonstrated the feasibility of remote telesurgery using the da Vinci platform to perform a complex surgical task. Copyright 2008 John Wiley & Sons, Ltd. Keywords
telesurgery; robotics; pyeloplasty; internet protocol; da Vinci
Introduction
Accepted: 15 July 2008
Copyright 2008 John Wiley & Sons, Ltd.
Prior to World War II, soldiers sustaining serious injuries in the field of battle would have little chance of survival, as they would have to be transported rearward from the front lines for significant distances in order to reach adequate medical care. The rise of mobile army surgical hospitals (MASH), later superseded by US military combat support hospitals (CASH), attempted to address timely delivery of critical care to the wounded soldier in the combat arena. However, as exemplified by events of World War II, in which medics were deemed primary targets by enemy combatants, pushing the delivery of medical care closer to the frontlines significantly increases personal risk to medical personnel. In an effort to mitigate this risk, groups such as the US military [Telemedicine and Advanced Technology Research Center (TATRC), Defense Advanced Research Projects Agency (DARPA)] began research into the area of remote surgery (telesurgery), which would not only limit the exposure of medical personnel to the dangers of armed combat, but would also allow the delivery of specialized care, such as neurosurgery, to the head-
Pre-clinical remote telesurgery trial of da Vinci telesurgery prototype
wounded soldier in a remote location. These concepts represent the impetus for development of telesurgery as a research area. With the recent invention of surgical robots able to consistently perform complex operations came the possibility of applying telesurgical principles to these platforms for telesurgical applications. Marescaux et al. (1) in 2001 performed the first preclinical telesurgical procedure, followed quickly by the first clinical telesurgical case involving a telesurgery (TS)-enabled Zeus (Computermotion, CA, USA) surgical robot used to perform an uncomplicated cholecystectomy between New York, USA, and Strasbourg, France, with a network latency of 155 ms (1,2). Latency is a networking phenomenon encountered during telesurgery, defined as the period of time it takes for the information element to traverse the network from its origin to its destination. Previous work has demonstrated an association between increasing latency and poorer surgical performance, with threshold values of acceptability in the range 500–600 ms (3–8). Other factors, such as jitter (the standard deviation of latency) and data packet loss, were identified as possible independent negative influences on outcome. The overall telesurgical experience has been limited to date; however, our centre has trialled TS Zeus extensively over a number of network configurations, out to round-trip distances of 75 000 km, through geosynchronous satellites (5–7). During previous trials (9,10), we have developed a standardized model of acute porcine pyeloplasty surgery which encompasses significant elements of fine intracorporeal reconstruction with standardized postsurgical and network metrics. Canadian Surgical Technologies and Advanced Robotics (CSTAR), in London, Ontario, is a state-of-the-art facility dedicated to the critical evaluation and development of medical and surgical innovations. Due to our extensive expertise with TS Zeus, our standardized telesurgical protocols and reconfigurable surgical grade network, Intuitive Surgical (Sunnyvale, CA, USA) and CSTAR collaborated to evaluate a new generation of technology in the form of telesurgery-enabled da Vinci (TS da Vinci). We hypothesize that the improvements in the da Vinci surgical system over previous surgical robotic platforms will further improve telesurgical operability.
Materials In November of 2005, CSTAR acquired the use of a TSenabled da Vinci surgical robot for a series of preclinical trials. The test unit consisted of an additional surgeon console, which would relay telecommunications from the remote unit to the patient-side cart and back again over an internet protocol (TCP/IP) network. The operator console is connected to the network and the signals are sent through the circuit to arrive at a modified network-ready remote console near the operative bedside. The remote console receives the operator commands and controls the hardwired patient-side cart. In addition, the remote Copyright 2008 John Wiley & Sons, Ltd.
305
console receives stereo video input and encodes this to be sent back through the network to the operator console to be displayed in true three dimensions (3D) within the surgeon console. This set-up allows the operator to manipulate all aspects of a typical da Vinci surgical system that are native to the unit. As such, the camera arm and two instrument arms are fully operable; however, cautery control was not enabled, so the operations were done with cold knife dissection only. The robot was enabled for use on the surgical grade network testbed located at CSTAR’s London, Ontario, location, administered by Bell Canada (Toronto, ON, Canada). For the purposes of these trials, a landlinebased network loop was set up and validated internally by systems and network engineers. During these telesurgical trials, data sent from the surgeon console in one room would be sent over the network via landline IP to Halifax, Nova Scotia, and reflected back to an adjacent room, at CSTAR,where TS da Vinci would reconstruct the data into Cartesian coordinated movement of the patient-side cart actuators. Similarly, two streams of high quality video and all of the reciprocal data from the patient-side cart was sent through TS da Vinci back through the telesurgical network to the remote surgeon console, where the video data streams were re-encoded and synchronized into full 3D context, using a pair of Haivision 500 codecs (Figure 1). The network latency and jitter incurred through traversal of the landline network would be measured in real time over the course of the experiments through inline enterprise router interrogation. Additional network parameters, including average and peak bandwidth usage, would also be measured. Qualitative checks were performed of the imaging system throughout the trials. Prediction of network resource allocation and bandwidth utilization was based on extensive TS Zeus studies in both drylab and wetlab experiments (data not shown). The telesurgical network configured for TS da Vinci trials consisted of two redundant 17 Mbps IP virtual private network (VPN) connections at the surgeon console and two redundant 17 Mbps IP VPN connections at the patientside cart, providing highly available wide area network (WAN) access to the Bell Canada IP VPN core network inside CSTAR in London, Ontario. The WAN connections are then looped back at Bell Canada’s central office (CO) in Halifax, which added approximately 3000 km round-trip distance between the surgeon and patient sides (Figure 2). Quality of service (QoS) flags were utilized, with the video streams receiving the highest QoS. As the up- and downlinks were DS-3 connections, each could potentially accommodate up to 45 Mbps. Cisco 7204 routers are used as the internet protocol/multiprotocol label switching (IP/MPLS) customer edge (CE) routers at each side. Circuit path diversity [building fibre entries, CE, provider edge (PE), CO] provided layer 1 and layer 2 redundancy. Within the 17 Mbps bandwidth, there are three channels of service: a 10 Mbps near real-time (NRT) quality of service (QoS) for robot control signals and Int J Med Robotics Comput Assist Surg 2008; 4: 304–309. DOI: 10.1002/rcs
306
C. Nguan et al.
Figure 1. Hardware set-up for telesurgical da Vinci trials. Left hand side of figure outlines surgeon console and associated connections; right hand side illustrates telesurgical da Vinci accessory surgeon console and patient-side cart, with associated connections
Figure 2. CSTAR network configuration for TS da Vinci preclinical trials. Note that the Halifax–PE loop was utilized in this trial (CSTAR MPLS Network Loopback Testbed, in blue) and NetDisturb was used for monitoring purposes
Copyright 2008 John Wiley & Sons, Ltd.
Int J Med Robotics Comput Assist Surg 2008; 4: 304–309. DOI: 10.1002/rcs
307
Pre-clinical remote telesurgery trial of da Vinci telesurgery prototype
real-time patient video images, a 2 Mbps priority QoS for IP-based conferencing and 5 Mbps standard QoS for network management and monitoring traffic services. The IP VPN QoS on-demand feature is proposed to reduce the total WAN cost by enabling the NRT and priority QoS services for only a short period of time when the telesurgery is taking place. The 10 Mbps NRT bandwidth is only required for one direction, i.e. from the patient’s site to the surgeon’s site. There is a minimal NRT bandwidth required (
