|
From June 1997 Issue
David Allen with Jon
Bowersox, M.D. (Telepresence Surgery) and G. Garfield Jones (Telementoring)
Author David Allen is a 4th year medical student at
the U. of Kansas Medical Center. This overview is a result of the clinical
rotation David did in 1996 with Dr. Bowersox at Stanford Research Institute.
David accompanied Dr. Bowersox and sat in the cockpit of the telepresence
surgery unit at SRI, reporting, "From the moment I sat down at the console and
manipulated the control arms of the system, most movements felt natural. The
visual environment was very realistic -- at one point I tried to reach into the
empty workstation to adjust placement of an errant needle."
Dr. Jon Bowersox is Director of Trauma Service,
Stanford U. School of Medicine, and a pioneer in telepresence surgery
(bowersox_j@hosp.stanford.edu)
G. Garfield Jones is Director of Member Services, US
TeleMedx Corp., a telemedicine service bureau providing technical, training,
telecommunications and clinical support on-line (GarfieldPA@aol.com)
Telesurgery is the provision of surgical care over a
distance, with direct, real-time visualization of the operative field. It may
be categorized as follows:
1) Telepresence surgery. Uses a computerized
interface to transmit the surgeons actions at a surgical workstation to the
operative site at the remote surgical unit, with haptic (force feedback) input
to transmit to the surgeon the tactile environment of the operative field
2) Telerobotics. Remote control with a robotic
arm, usually in conjunction with a laparoscope, without haptic feedback
3) Telementoring. An experienced surgeon acts
as a preceptor for a remote inexperienced surgeon by observing the surgeon via
interactive video. Teleproctoring is an extension of telementoring,
referring to documentation of performance for privileging purposes.
Related fields, such as virtual reality and computer
generated educational and training scenarios, will not be discussed.
Telepresence Surgery
Telepresence was spawned in the 1950s by the
atomic energy industrys unique need for the remote handling of dangerous
isotopes. These applications, termed "teleoperation", allowed hazardous
materials to be handled safely at a distance while being monitored by
television. Modern "telepresence" adds multisensory input that recreates the
remote visual, auditory, and tactile environment. This allows the operator to
feel physically present at the remote site, in terms of sensory input and the
ability to manipulate objects. The perception of distance is erased, and the
operator can act effectively in a locally recreated environment that is in fact
an illusion.
The Equipment
The prototype telepresence system was developed by
electrical engineer Dr. Philip Green at Stanford Research Institute (SRI; Menlo
Park, CA). The technology has gone through several generations since the first
one-handed model was constructed in 1993. The current two-handed system allows
the surgeon to manipulate robotic arms that exactly reproduce the
surgeons hand motions from the surgical workstation to the remote
surgical unit. This is accomplished as the surgeons fingers are placed
into the rings of actual surgical instruments (SEE PHOTO). However, rather than
having normal instrument tips, the rings are directly connected to motors,
gears, and belts. These precisely translate the surgeons hand and finger
motions into digital signals that are transmitted through a computer and
telecommunications link to robotic arms at the remote surgical station. The
system allows movements similar to those that could be performed by the human
hand and arm and moveable elbow -- but with a fused wrist. This range of motion
corresponds to 4 degrees of freedom (DOF), or number of ways that a robotic arm
can move. Note that the human arm has 7 DOF, the hand over 20 DOF. Current work
is directed at increasing the DOF of robotic arms to 7 or more DOF.
There has been talk for years of a "virtual reality
glove" that would receive the full tactile range, and project the full
manipulative capabilities of, the human hand. There is no evidence that such a
device will be encountered in clinical practice in the foreseeable future.
Haptic feedback
The surgeon also receives haptic feedback ("force
reflection"; graduated tactile input) back through the robotic arms being
manipulated. This feedback is accomplished through small servo motors at the
near and remote sites, controlled by complex mathematical algorithms. Simply
stated, as resistance is encountered by the motors at the remote surgical unit
(e.g., the tip of a suture needle passes into tissue more dense than the
initial subcutaneous tissue), electronic signals are instantly fed back to the
surgical workstation. These cause the motors in the instrument arms there to
recreate the same amount of resistance being encountered in the remote surgical
unit, even though the instrument handles the surgeon is controlling are in fact
suspended in mid air, attached only to electronic cables.
Visual input
Visual input is stereoscopic and three-dimensional
(3-D). Two digital color cameras transmit images at a combined rate of 30
frames/second (15 fps / eye) and can be controlled to zoom and pan into and
away from the operative field. The 3-D aspect of visual input requires the
surgeon to wear optically treated 3-D glasses. The visual environment in the
surgical workstation corresponds exactly to that in the remote surgical suite.
This is accomplished by mounting a monitor above the surgeons
workstation, which projects the images from the remote cameras onto a mirror
directly over the surgeons hands and control rings. The projected image
displays the instrument tips from the remote site just as if they were at the
end of the rings in the surgeons hands. Image transmission is essentially
instantaneous. Data are transmitted at 90 Mbps (although research suggests that
45 Mbps should suffice). Audiovisual information accounts for nearly all of
this; only a few Mbps of bandwidth are required for haptic information. Thus, a
stable, real-time kinetic environment is recreated. Avoiding lag time in
transmission is extremely important. Lag time can lead to vestibulo-ocular
disruption and "3-D Vertigo" or "simulator sickness". This is frequently
encountered in head-mounted virtual reality displays; over 50% of virtual
reality (VR) system users experience dizziness, nausea, or headaches within 20
minutes of donning VR headgear. This has not been a problem in high-bandwidth
telepresence surgery.
Movement Scaling
Ultimately, one of the most clinically important
refinements of telepresence surgery may be "movement scaling". Using
straightforward mathematical and engineering techniques, gross movements can be
scaled to control much finer movements at the receiving end. Thus, a 1-cm
movement of the surgeons finger could translate into a 1-mm movement of
the remote instrument tip. This would enable a 10x increase in fine motor
control for microvascular and plastic surgery. For example, a future surgeon
could "be" inside a vessel, repairing the intima with a nearly microscopic
needle and suture material. These could be manipulated to tolerances far beyond
the physiologic capability of unaided human hand-eye kinetics. Allied
engineering also enables the dampening of hand tremors.
Current and Future Uses
Currently, telepresence surgery is experimental and is
not a realistic alternative in a modern hospital. Arteriotomies involving
incisions and various suturing techniques have been done successfully on swine;
wound closure takes about twice as long as standard on-site techniques.
However, the precision of the telepresence and on-site techniques has been the
same.
Current telepresence surgery systems have operated
intra-enterprise, typically with fewer than 1,000 feet separating the
surgeons console and the remote surgical unit. Concern has been expressed
about time lag introduced by further separation -- not so much from compression
delay, which is minimal in these broad-band systems, but by the speed of
electricity (the speed of light). Preliminary studies have shown that a
transmission delay of 100 msec or more leads to problems with surgical accuracy
and precision. This should not be a problem in images transmitted terrestrially
across a continent, using microwave or fiberoptic links. It is a potential
problem in circumglobal or satellite uplinks, where timeliness of transmission
is constrained by distance and the speed of light.
Telepresence surgery may have its earliest clinical
impact in management of combat trauma. 90% of combat-related deaths occur from
exsanguination in the combat zone, prior to evacuation. Many of these lives
could be saved if surgical treatment were available within the first "Golden
Hour" of injury. Field exercises have tested the feasibility of military
telesurgery, and at least one prototype armored vehicle has been outfitted with
telepresence surgery equipment. This highly experimental military deployment is
augmenting other combat uses of telemedicine, including digital Personnel
Status Monitors for locating soldiers and assessing their vital signs, and
store-and-forward and interactive video access to worldwide medical
expertise.
Application within the space program is also being
considered.
Telementoring
As noted, this is the use of interactive video to
provide "looking over the shoulder" expertise by a remote expert to a local
surgeon inexperienced in a particular technique. It was first reported in 1965
by Dr. Michael DeBakey, transmitting over broadband satellite. More recent
efforts have been at bringing the remote surgical expert more into the
operative environment through the incorporation of telerobotics.
Telementoring addresses the serious problems of
disseminating new operative and especially minimally invasive surgery (MIS)
techniques. Currently, over 90% of cholecystectomies (gall bladder removals)
are done laparoscopically, using MIS. Other common MIS procedures are
herniorraphies, appendectomy, fundoplications, and arthroscopy. Typical
training involves time consuming, expensive on-site coursework with an
experienced instructor. Telementoring, especially when augmented by
telerobotics, helps eliminate the need for travel and adjusting schedules for
training and credentialing.
Early MIS was done under direct observation using
mirrors and optical gadgets. Today, it is done using fiberoptic lighting and
image acquisition that transmits the image to a video monitor, where the
surgeon views the operative field indirectly. Since the image is on video, it
can be readily digitized, compressed, and transmitted to a remote viewing
station within the institution or miles away. Usually images are transmitted at
384 768 Kbps (ºT1 T1) using the H.320 compression
standard. Multipoint conferencing allows participation from multiple sites, so
that a single master surgeon can be precepting at several remote sites at once.
Basic telementoring systems at the operative site
include: a room camera showing a panoramic view of the patient and the O.R.
team, and which can be directed at an X-ray on a viewbox; a patient camera
positioned over the patient, usually incorporated into the light gantry above
the patient; at least one local monitor; and a wireless lapel microphone. Video
mixers allow for split screen viewing (simultaneous internal and external
views). Another common feature is telestration (on-screen annotation for
teaching). To better assist the trainee, the telementor must have the ability
to control the positioning and view of the laparoscopic camera within the
patient. This can be done using a remote control robotic arm (Computer Motion,
Goleta, CA, www.ComputerMotion.com see photo). The total cost for
setting up an operative suite for telementoring run from $50 - $100,000.
Transmission costs for fractional T1 service (385 Kbps) are usually about
$60/hour, plus monthly fixed fees.
In 1992 United Medical Network conducted the first
compressed video telementoring session; that company has since gone out of
business. In the following few years numerous telesurgery demonstrations were
done from Doctors Hospitals in Columbus, OH (Dr. James Perez), at the
American College of Surgeons Annual Meeting (1993), Alliance for Continuing
Medical Education Annual Meeting (1994), and from St. Josephs Hospital in
Towson, MD to Ethicon Endo-Surgery in Cincinnati (1994). The modern era of
telementoring began in 1994 when Dr. C.Y. Liu (Chattanooga, TN) conducted the
first private telementoring course on advanced gynecologic laparoscopy. In 1995
the Society of American Gastro-Intestinal Endoscopic Surgeons (SAGES) used
dial-up, compressed video for telesurgical education at their annual meeting.
Another pioneer association has been the Association of Gynecological
Laparoscopists (AAGL). Important use and expansion of the technology has been
promoted by teams at Johns Hopkins University and Yale University Schools of
Medicine, and at the militarys Advanced Research Projects Agency (ARPA).
ARPA has done experiments in which a surgeon has actually performed MIS
surgical procedures from a distance using a Nintendo-like keypad/joystick
apparatus to control cautery, cutting, laser, coagulation, suction etc.
Several medical manufacturers, such as Ethicon
Endo-Surgery, U.S. Surgical, Origin, and Circon-ACMI are now using dial-up,
compressed video telesurgery at their headquarters to introduce more surgeons
to MIS techniques.
While a few studies have documented the efficacy of
telementoring techniques (see references at end), at least one professional
organization, SAGES, has formally expressed concern about "premature
proliferation of such technology before adequate educational, training, ethical
and legal issues have been appropriately deliberated and developed."
Nevertheless, it is clear that as surgeons strive to perform (and as patients
demand) more types of surgery through minimally invasive techniques,
telementoring will provide an important route for widening the circle of
trained, competent practitioners. Meanwhile, the issues liability,
reimbursement, cross-state licensure -- that confront other telemedicine
applications apply as well to telementoring. A special subgroup of the American
Trial Lawyers Association deals with laparoscopy litigation cases; it is likely
that their purview will extend soon to telesurgical litigation. SAGES will
release its "Guidelines for the Surgical Practice of Medicine in Telemedicine"
later this summer, addressing telementoring, teleproctoring, credentialing, and
privileging.
SUGGESTED READING
Bowersox JC, Shah A, Jensen J, Hill J, Cordts PR.
Vascular applications of telepresence surgery: initial feasibility studies in
swine. J Vasc Surg 23:281-7, 1996
Green PS, Hill JW, Jensen JF, Shah A. Telepresence
surgery. IEEE Engineering in Med and Biol May/June 1995, pp. 324-329
Hiatt JR, Shabot MM, Phillips EH, Haines RF, Grant
TL. Telesurgery: acceptability of compressed video for remote surgical
proctoring. Arch Surg 131:396-400, 1996
Moore RG, Adams JB, Partin AW, Docimo SG, Kavoussi
LR. Telementoring of laparoscopic procedures. Initial clinical experience. Surg
Endosc 10:107-110, 1996
Regan EC, Price KR. The frequency of occurrence and
severity of side-effects of immersion virtual reality. Aviat Space Environ Med
65:527-30, 1994
Satava RM. Virtual reality and telepresence for
military medicine. Comput Biol Med 25:229-236, 1995
PROFILES OF WORKING TELESURGERY PROGRAMS
Telemedical Emergency Neurosurgical Network
Developed by neurosurgeon Paul Chodroff of Walnut
Creek, CA, this network links five private practice neurosurgeons at John Muir
Medical Center with Sutter Solano Medical Center in Vallejo, which doesnt
have around-the-clock access to neurosurgeons. From home or office computers at
11 locations in the north Bay Area, Chodroff and colleagues can view digitally
transmitted CT head scans of patients brought to the Sutter emergency
department 30 miles away, and evaluate whether and how urgently -- they
need surgery. Currently 9-10 cases are seen per month; a total of about 300
cases have been seen since program inception in 1994. Nearly 60% have not
required transfer. Most of these would otherwise have been transferred via
helicopter or ambulance at an average cost of $4,500.
Since CT images are in digital format to begin with,
no digitization is required. A video interface to Macintosh computers with
boards from MultiAccess Computing expedites image transfer over ISDN at 128
Kbps, using ADTRAN ISU/CSU units. Images are received at Macintosh or PC
computers with monitors resolutions of 1024 x 768. An 18-slice CT scan can be
transmitted at 2:1 lossless compression in four minutes.
The total cash outlay for the project has been about
$43,000; in-kind (pro bono) products and services have been supplied by ADTRAN,
CalREN (a Pacific Bell trust), and BPC Consultants, of which Dr. Chodroff is a
principal.
Dr. Chodroff emphasizes that this is a telemedicine
consultation rather than simply teleradiology. It utilizes simultaneous
interactive audio consultation and, as necessary, text and audio files can be
appended to the radiographic record. This is especially useful for bundling
patient records when soliciting a telemedical "curbside consult." Currently,
there is no evidence that interactive videoconferencing capabilities would add
much value. To date, there has been no case where a significant difference was
noted between the transmitted images and the hard copies sent with patients.
The project has been adjudged fully in compliance with COBRA/EMTALA regulations
for emergency medical consultations by the regional HCFA Health Care Standards
& Quality authority.
This expanding community service project is gradually
adding hospitals and neurosurgeons. There are plans to enable MRI scan
transmission. Contact Dr. Chodroff at PHC_BPCCons@compuserve.com.
Surgical Tele-Education: The Ross Procedure
Classically, diseased aortic valves have been replaced
with artificial or pig valves. These require sometimes hazardous postoperative
care such as anticoagulation. A technique developed by Dr. Donald Ross, a
London-based cardiothoracic surgeon, repairs the diseased portion of the valve
using donor tissue from the patients own heart. This highly complex
surgery can result in a longer lasting repair requiring less postoperative
intervention. However, few cardiothoracic surgeons are trained in the
technique. In early May, 1997 a Ross Procedure symposium was held in
Indianapolis, chaired by Dr. John Brown, chief cardiothoracic surgeon at
Indiana University Medical Center. With Dr. Ross in attendance, Dr. Brown
performed three successive Ross Procedures, and was able to demonstrate the
technique to colleagues in Denver, CO and at a nearby hotel. Over 80 off-site
surgeons were able to observe the surgery with close-up views of the operative
field and more panoramic views of the operating theater, projected onto 10x12
foot screens. Moreover, the sessions were interactive, enabling the observers
to ask questions of (and give advice to) Dr. Brown in real time. Dr. Brown wore
a head-mounted camera intraoperatively, enabling observers to see exactly what
he was seeing. The telemedicine equipment was VTELs F.R.E.D system, which
includes a high resolution 3-CCD camera on an adjustable arm, as well as an
integrated PC-based image digitizer/compressor (CODEC), lavaliere microphone,
and peripheral devices. The high quality of the projected image prompted Dr.
Brown to comment, "Besides the fact that I was the one with the scalpel in
hand
there was nothing that I could see that the audience could not."
Live Surgery Education Over the Internet
Another interactive medical training option: live
Internet broadcasts of surgical operations. This was done recently when an
hour-long cholecystectomy at Wayne State University (Pontiac, MI) was broadcast
over the Internet to physicians in the U.S. and Argentina. The project,
coordinated by Dr. Alejandro Gandsas, was a feasibility study transmitting
images over 28.8 Kbps modems, with image displays of 320x240 pixels. Hardware
included 75 MHz Pentium computers with video-capture cards. While image quality
and motion handling were far from perfect, they were good enough for observing
surgeons to identify pertinent anatomic structures. Also, the interactivity
enabled collegiate commentary and advice. Internet interactions have a powerful
advantage over satellite or other long-distance options: there are no
transmission costs. |