HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Bob B. Kiaii Thomas Walther Friedrich W. Mohr Volkmar Falk Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Jacobs, S. Articles by Falk, V. Search for Related Content PubMed PubMed Citation Articles by Jacobs, S. Articles by Falk, V. Related Collections Minimally invasive surgery

Ann Thorac Surg 2003;76:2029-2035
© 2003 The Society of Thoracic Surgeons

---

Original article: cardiovascular

Limitations for manual and telemanipulator-assisted motion tracking—implications for endoscopic beating-heart surgery

^a Department of Cardiac Surgery, Heartcenter, University of Leipzig, Leipzig, Germany

Accepted for publication May 14, 2003.

^* Address reprint requests to Dr Jacobs, Klinik für Herzchirurgie, Universität Leipzig Herzzentrum, Strümpellstr 39, 04289 Leipzig, Germany
e-mail: stjacobs{at}aol.com

	Abstract

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
BACKGROUND: Surgical performance is limited by human factors. Beating-heart surgery requires full dexterity and motion tracking. Currently techniques for total endoscopic beating-heart bypass grafting using telemanipulation systems are being developed. The aim of this study was to assess the limitations for manual and telemanipulator-assisted motion tracking using the da Vinci telemanipulator system.

METHODS: To simulate beating-heart conditions an endoscopic trainer was developed. Twenty subjects were asked to touch targets manually and with telemanipulator assistance with different patterns of increasing index of difficulty (resting model, unstabilized, and stabilized model with a frequency of 35, 60, and 90 beats per minute). In addition one task was performed using different scaling ratios on a resting model. The times between hits as well as errors were electronically recorded.

RESULTS: There was no significant impact of various frequencies and amplitudes for manual tracking. The average values for the delay (k_m[ms]) and information-processing (c_m [ms/bit]) constants for the manual tasks were 201 ms and 86 ms/bit respectively. Both the delay constant (k_t = 630 ms; p < 0.0005) and the information-processing constant (c_t = 250 ms/bit; p < 0.0005) were increased for the telemanipulator-assisted tasks at rest. When working on moving targets telemanipulator-assisted tracking required significantly more time and led to more errors. At a frequency of 90 beats per minute telemanipulator-assisted tracking became more difficult.

CONCLUSIONS: Endoscopic beating-heart bypass grafting requires optimal stabilization to avoid inaccuracies due to incomplete motion tracking. At higher frequencies telemanipulator-assisted tracking became more difficult, demonstrating the technical limits of current telemanipulator technology.

	Introduction

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
Beating-heart coronary artery bypass grafting has become a routine procedure [1–3]. With the introduction of surgical telemanipulators total endoscopic coronary artery bypass grafting on the arrested heart became feasible [4, 5]. Computer-enhanced instrumentation systems allow dexterous manipulations in confined spaces and have helped to overcome the limitations of conventional endoscopic instruments. Despite improved technology and increasing experience with off-pump procedures endoscopic beating-heart surgery is still in its infancy [6–8]. That can be attributed to the additional difficulty of working on a moving target in a closed chest environment.

The information processing model for manual control and tracking includes perception, cognition, reaction, and feedback of the operator. Faced with tracking tasks the operator acts directly on the target. While the operator-related human factors are well studied, some basic information on the capability of the currently used telemanipulation systems such as absolute geometric accuracy, bandwidth, and system response time are lacking [9]. The aim of this study was to evaluate the impact of a telemanipulator on simple control and tracking tasks mimicking a beating heart scenario.

	Material and methods

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
Fitts’ law
In 1954 Fitts described the relation of the speed of a controlled movement and the difficulty of the response under resting conditions (nonmoving object). Fitt's law states that the movement time (T_M) varies with the distance and width of the target. This follows from the general formula relating task difficulty and movement time that has come to be known as Fitt's law:

where A is the target distance, W is the target width and k and c are constants (k [s] is a delay constant and c [s/bit] a measure of information processing). The typical values for hand movements are k = 0.177 s and c = 0.1 s/bit). Target width and distance thus describe the difficulty to achieve the task. The index of difficulty (ID) is expressed in bits:

Fitt's law can therefore be expressed as:

By varying target distance (or width or both) it is possible to change the informational demands of a psychomotor task [9, 10]. Fitt's law predicts that a decrease of the index of difficulty of an one-dimensional task (by decreasing either the distance between targets or increasing the target size) will decrease the time to perform the movement (Fig 1).

View larger version (18K): [in this window] [in a new window]   Fig 1. Description of Fitt's law. The operator is asked to move from a starting point to a target. The width (W) of the target in which the operator wishes to terminate the movement and the distance (A) between the two points determine the difficulty of the task (a). By increasing the width of the target the index of difficulty (ID) decreases (b). Similarily, by decreasing the distance towards the target, the ID is decreased (c). Fitt's law predicts that an increase in the index of difficulty results in an increase of movement time.

Endoscopic trainer
To simulate beating-heart conditions an endoscopic trainer consisting of a moving plate with a fixed task pad was developed (Fig 2). The plate was driven by three servomotors. A single chip computer controlled the servomotors and recorded the touches on the surface of the platform automatically. Twelve metallic contacts were arranged as targets in the middle of the moving plate. Each contact had a diameter of 1 mm and the horizontal and vertical distance between the contacts was 2.5 mm; for the diagonal it was 3.6 mm. The single chip computer allowed for automatic registration of the time between contacts and thus calculation of the speed of each motion [12]. Based on experimental real time measurements of beating-heart motion with and without stabilization using ultrasonic crystals [13] different surface motion were chosen to simulate a realistic beating-heart situation. The amplitude and frequency of target plate motion could be changed independently to allow for different motion patterns that resembled stabilized and unstabilized beating-heart conditions.

View larger version (109K): [in this window] [in a new window]   Fig 2. Endoscopic trainer consisting of a moving plate with a fixed task pad. Design of two point tapping task in a vertical, diagonal, and horizontal direction and the meander figure are displayed.

Study protocol
Twenty subjects inexperienced in robotic and laparoscopic surgery were asked to touch targets on the plate in different patterns with increasing index of difficulty. The task was to move an instrument from one point to another and by intention avoided any complex endoscopic manipulation and decision making. By intention the tasks did not require a deeper understanding of telemanipulator technology. None of the features offered by the system (camera motion, clutching) were used and all tasks were single handed to avoid bimanual manipulation. The time between hits (target touch signal) as well as the errors (ground plate touch) was electronically recorded. First, the tasks were performed manually using a light pen with a tip of 0.2 mm in diameter that was electronically connected with the single chip computer. Second, all tasks were repeated with the da Vinci telemanipulation system (Intuitive Surgical, Sunnyvale, CA). The da Vinci system is a surgical telemanipulation system with six degrees of freedom. The system has been described in detail elsewhere [14]. Upon contact of the instruments with the target an electronic loop was closed that encoded start and stop signals that were recorded automatically. Whenever the instrument failed to touch a target and the surrounding plate was touched instead, an error signal was automatically recorded. The number of errors for all tapping tasks (horizontal, vertical, diagonal) was averaged at rest and for the different frequencies and amplitudes. The two point tapping tasks were performed first, followed by a meander task. All subjects completed each task, which was performed eight times by each subject. The different patterns consisted of one-dimensional movements between two points (two point tapping task) of varying distance and angle. A second exercise required following a continous motion in a meander pattern (Fig 2). To assess the impact of scaling, a two point tapping task was performed at rest at different scaling levels. Scaling refers to the relation of motion performed at the master console and the resulting motion of the endoscopic instruments. Three scaling levels are available with the da Vinci system: 1:1, 3:1, and 5:1. In the 5:1 scaling mode the outside motion at the master is decreased at a ratio of 5:1 in relation to the motion of the instrument.

The operator was working on the target manually at an angle of approximately 45 degrees (angles were not measured) with a direct straight view to allow for ideal hand-eye alignment. The working angle for the instruments of the da Vinci system was similar (approximately 45 degrees) to looking through a 0-degree scope.

Statistics
For evaluation, tasks were evaluated seperately with repect for time and errors. Fitt postulated a linear relation between the time needed to complete the task and the index of difficulty. Correlation coefficients were calculated using linear regression analysis (SPSS 10.0 statistic software) with the index of difficulty being the independent and the time being the dependent variable. Next using Fitt's law, the delay constant (k[s]) and the information handling capacity (c[bits/s]) were calculated. All values are expressed as mean ± SD. For comparison of means between groups the one-way analysis of variance (ANOVA) test was used. Post-hoc analysis was performed using the Tamhane test. Values of less than 0.05 were considered significant.

	Results

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
Resting model
The resting model times were evaluated using Fitt's law. Regression lines were calculated for the manual and the telemanipulator-assisted movements in horizontal, vertical, and diagonal directions (Table 1, Fig 3). The average values for the delay (k_m [ms]) and information-processing (c_m [ms/bit]) constants for the manual tasks were 201 ms and 86 ms/bit respectively. These values were similar compared with the typical values of 177 ms and 100 ms/bit that have been previously reported in the literature [15]. The delay constant for the telemanipulator-assisted tasks was three times longer (k_t = 630 ms; p < 0.0005) and the information-processing constant is increased (c_t = 250 ms/bit; p < 0.0005). An increase in the information processing constant means that an increased amount of time is needed to process a given information (bit) and therefore describes a decrease in information processing.

View larger version (20K): [in this window] [in a new window]   Fig 3. Regression lines and correlation coefficients (r) for manual and telemanipulator (TM)-assisted two point tapping tasks at rest for diagonal, horizontal, and vertical directions. As Fitt's law describes a linear relationship where the delay constant is the intersection of the straight line with the y-axis and the information processing constant represents the slope of the straight line. Heavy dashed line = TM diagonal (r = 0.97); heavy solid line = TM horizontal (r = 0.98); dash-dot line = TM vertical (r = 0.81); light dashed line = manual diagonal (r = 0.99); broken line = manual horizontal (r = 0.89); light solid line = manual vertical (r = 0.98).

For the two point tapping tasks the errors rose from 10.0 ± 7.24 (manual) to 29.2 ± 16.46 (telemanipulator; p = 0.036); for the meander task it was 4.9 ± 4.46 to 14.0 ± 9.43, respectively (p = 0.184).

Moving model
There was a significant difference between the manually and remotely performed tasks. The subjects required 2.9 ± 0.44 (p < 0.00024) times more time using the telemanipulator. Within the manual as well as the telemanipulator groups there were no significant differences but a tendency toward an increase in the time needed with higher frequency (beats per minute) and amplitude (stabilized versus unstabilized) of the movement (Fig 4). At 90 beats per minute operators were in general not able to track the targets. Owing to multiple errors no meaningful data could be obtained for the tracking time. This was especially true for the telemanipulator-assisted group.

View larger version (27K): [in this window] [in a new window]   Fig 4. Time (T) to touch two targets in a vertical line. A is the distance from the starting point to the target (distance the needle has to move). With increasing distance to the target (A3 > A2 > A1) the index of difficulty and thus the time to perform the task increases. The distance between two points is 2.5 mm for vertical direction. Results are shown at rest and with variing frequencies (35 and 60 beats per minute) and amplitudes. Solid bars = A1, 2.5 mm; open bars = A2, 5.0 mm; hatched bars = A3, 7.5 mm.

Within the telemanipulator group the number of errors of the two point tapping tasks also increased from 29.2 ± 16.5 (meander task 14.0 ± 9.4) at rest to 61.0 ± 28.1 (meander task 33.3 ± 15.0) at 35 beats per minute stabilized and peaked at 84.4 ± 35.3 (meander task 46.2 ± 27.0) errors at 60 bpm unstabilized. The differences between the errors made at rest and with any movement were statistically significant (p < 0.001; Fig 5).

View larger version (19K): [in this window] [in a new window]   Fig 5. Number of errors for a more complex task (meander) at rest and different frequencies (35 and 60 beats per minute) and amplitudes for manual and telemanipulator-assisted control and motion tracking.

View larger version (12K): [in this window] [in a new window]   Fig 6. Mean number of errors at different scaling rates. The number of errors at the two point tapping task in a vertical, diagonal, and horizontal direction decreased when the scaling rate was changed from 1:1 over 3:1 to 5:1.

	Comment

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

For tracking tasks the object is followed closely by the controlled system. The operator tries to make the controlled system follow the changes in the target system. In the control task the operators responses act directly upon the target system to keep it within tolerance or to move it to a desired position. The target is moved into a position or held in a defined frame [17]. Both tasks have an impact on each other and occur simultaneously during beating heart surgery (Fig 7).

View larger version (31K): [in this window] [in a new window]   Fig 7. Structural model of tracking and control tasks. For tracking tasks (a) the operator follows a moving object. The manual response is preceded by signal reception and a decision making process. The manual movements are transmitted to the target system by the control panels and the effector devices of the telemanipulator. The current state of the target system is perceived by the operator closing the feedback loop. For the control task (b) the internal (wish) or external (command) goal is to change the state of an object. The information and motion flow as well as the feedback are similar as for tracking tasks. The effect on the object is additionally influenced by its behavior and structural texture (inherited dynamics). Both control and tracking tasks occur simultaneously and have impact on each other during beating heart surgery.

View larger version (18K): [in this window] [in a new window]   Fig 8. Feedback-loop for the operator performing a manual task (a). Visual and tactile signals are received. A decision is made followed by a response. The effect on the object is immediately seen or felt and thus fed back to the operator. By using a telemanipulator (b) the response of the operator is further processed. This causes a system related delay and the effect on the object is indirect. Additionally, with most telemanipulators fine tactile feedback is lacking and the operator has to rely mostly on indirect visual signals that are displayed on a video monitor.

The geometric accuracy (accuracy by which a target in a Cartesian coordinate system (x, y, z) can be touched) of humans is limited to some 100 to 200 µm and even under ideal conditions (rested elbows and microscopic view) may not exceed 50 µm [22]. Arguments have been made that a computer-enhanced telemanipulator may increase accuracy but recent studies have demonstrated that robotically assisted suturing of an anstomosis showed a higher degree of endothelial damage [23]. The shape and the maximum diameter of the needle puncture mark were significantly larger as compared with a conventional manual technique.

Working on the beating heart is a complex multidimensional tracking task during which the operator tries to have the controlled system follow the changes of the target system. By adding a telemanipulation system the task becomes even more complex. In addition to the various factors that influence the human motor response the mechnical, optical, and electronical delays of the telemanipulator additionally prolong the response time.

Accordingly this study demonstrates that telemanipulator-assisted performance required more time and led to more errors even in the nonmoving model. The average values for the delay and information processing of the telemanipulator-assisted tasks were higher compared with the manually executed tasks. For the moving model the number of errors increased dramatically when using the telemanipulator, demonstrating impaired accuracy with motion. At higher speeds (equaling a frequency of 90 beats per minute) tracking became almost impossible, demonstrating the technical limits of current telemanipulator technology. From these results it is obvious that endoscopic beating-heart bypass grafting requires optimal stabilization to avoid inaccuracies due to incomplete motion tracking. Endoscopic stabilizers must at least have the same but preferrably better immobilization capacity than that of the state-of-the-art stabilizers for off-pump surgery.

One limitation of the study is that only unexperienced operators performed the tasks. It is therefore possible that the results would have been different if an experienced user group would have performed the tasks. However the task was only to move an instrument from one point to another, minimzing the requirements for dexterity. As the tasks did not require a deeper understanding of telemanipulator technology we believe that the results are indeed representative. Telemanipulator-assisted tracking means adding a second information processing system on top of the general information processing model inherent to the human operator (plan, initiate, control, end, and check). That and the added electronic and mechanical delays cause an extended response time as well as a lack in absolute geometric accuracy. For resting targets the system-related delay is not as evident but on the moving object the operator's ability to control and track is affected. These findings may have only limited clinical importance as successful endoscopic bypass grafting has been reported by a number of groups including our own. Nevertheless the development of future surgical telemanipulation systems should aim to offer an increased bandwith and different hardware design that will allow for a faster response. At the same time better quality stabilizers are needed and development of algorithms for virtual immobilization should be encouraged.

	Acknowledgments

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
We thank Chris Ullman for building and programming the endoscopic trainer.

	References

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 

1. Calafiore A.M., DiGiammarco G., Teodori G., et al. Midterm results after minimal invasive coronary surgery (LAST-Operation). J Thorac Cardiovasc Surg 1998;115:763-771.[Abstract/Free Full Text]
2. Diegeler A., Falk V., Matin M., et al. Minimally invasive coronary artery bypass grafting without cardiopulmonary bypass: early experience and follow-up. Ann Thorac Surg 1998;66:1022-1025.[Abstract/Free Full Text]
3. Hart J.C., Spooner T., Pym J., et al. A review of 1582 consecutive Octopus off-pump coronary bypass patients. Ann Thorac Surg 2000;70:1017-1020.[Abstract/Free Full Text]
4. Falk V., Diegeler A., Walther T., et al. Total endoscopic coronary artery bypass grafting. Eur J Cardiothorac Surg 2000;17:38-45.[Abstract/Free Full Text]
5. Kappert U., Schneider J., Cichon R., et al. Development of robotic enhanced endoscopic surgery for the treatment of coronary artery disease. Circulation 2001;104(Suppl 1):102-107.[Abstract/Free Full Text]
6. Diegeler A., Matin M., Falk V., et al. Indication and patient selection in minimally invasive and off-pump coronary artery bypass grafting. Eur J Cardiothorac Surg 1999;16(Suppl 1):79-82.
7. Falk V., Diegeler A., Walther T., et al. Endoscopic coronary artery bypass grafting on the beating heart using a computer enhanced telemanipulation system. Heart Surg Forum 1999;2:199-205.[Medline]
8. Falk V., Diegeler A., Walther T., Jacobs S., Raumans J., Mohr F.W. Total endoscopic off-pump coronary artery bypass grafting. Heart Surg Forum 2000;3:29-31.[Medline]
9. Gibbs C.B. Controller design: interactions of controlling limbs, time lags and gain in positional and velocity systems. Ergonomics 1962;5:385-402.
10. MacKenzie I.S. A note on the informationtheoretic basis for Fitt's law. Motor Behav 1991;21:323-330.
11. Fitts P.M. The information capcity of the human motor system in controlling the amplitude of movement. J Exper Psychol 1954;47:381-383.
12. Falk V. Manual control and tracking—a human factors analysis relevant for beating heart surgery. Ann Thorac Surg 2002;74:624-628.[Abstract/Free Full Text]
13. Koransky M.L., Tavana M.L., Yamaguchi A., Robbins R.C. Quantification of mechanical stabilization for the performance of off-pump coronary artery surgery. Heart Surg Forum 2001;4(Suppl 2):125.
14. Mohr F.W., Falk V., Diegeler A., et al. Computer-enhanced robotic cardiac surgery. Experience in 148 patients. J Thorac Cardiovasc Surg 2001;121:842-853.[Abstract/Free Full Text]
15. Jagacinski R.J., Repperger D.W., Moran M.S., Ward L.B., Glass B. Fitts’ law and the microstructure of rapid discrete movements. J Exper Psychol 1980;6:309-320.
16. Knight JL. Manual control and tracking. In: Salvendy G, ed. Handbook of human factors. New York: John Wiley & Sons, 1987:182–218
17. Bennet C.A. Sampled-data tracking: sampling of the operators output. J Exper Psychol 1956;51:429-438.
18. Damiano R.J., Ehrman W.J., Ducko C.T., et al. Initial United States clinical trial of robotically assisted coronary artery bypass grafting. J Thorac Cardiovasc Surg 2000;119:77-82.[Abstract/Free Full Text]
19. Goertz R.C. Fundamentals of general-purpose remote manipulators. Necleonics 1952;10:36-45.
20. Bejczy A.K., Salisbury J.K. Kinesthetic coupling for remote manipulators. Comp Med Engineer 1983;2:48-62.
21. Detter C., Deuse T., Christ F., Boehm D.H., Reichenspurner H., Reichart B. Comparison of two stabilizer concepts for off-pump coronary artery bypass grafting. Ann Thorac Surg 2002;74:497-501.[Abstract/Free Full Text]
22. Taylor R.H., Jensen M., Whitcomb L., et al. A steady-hand robotic system for microsurgical augmentation. Robot Res 1999;12:1201-1221.
23. Arnold M, Boehm DH, Welsch U, Detter C, Reichart B, Reichenspurner H. Evaluation of acute changes of the coronary artery wall after robotically assisted endoscopic coronary artery bypass grafting. Heart Surg Forum 2002;5:128–31

This article has been cited by other articles:

				P. Modi, E. Rodriguez, and W. R. Chitwood Jr. Robot-assisted cardiac surgery Interactive CardioVascular and Thoracic Surgery, September 1, 2009; 9(3): 500 - 505. [Abstract] [Full Text] [PDF]

				U. Kappert, S.-M. Tugtekin, R. Cichon, M. Braun, and K. Matschke Robotic totally endoscopic coronary artery bypass: A word of caution implicated by a five-year follow-up. J. Thorac. Cardiovasc. Surg., April 1, 2008; 135(4): 857 - 862. [Abstract] [Full Text] [PDF]

				R. J. Damiano Jr Robotics in cardiac surgery: The emperor's new clothes J. Thorac. Cardiovasc. Surg., September 1, 2007; 134(3): 559 - 561. [Full Text] [PDF]

				T. Watanabe, S. Omata, M. Odamura, M. Okada, Y. Nakamura, and H. Yokoyama Three-dimensional quantification of cardiac surface motion: a newly developed three-dimensional digital motion-capture and reconstruction system for beating heart surgery. J. Thorac. Cardiovasc. Surg., November 1, 2006; 132(5): 1162 - 1171. [Abstract] [Full Text] [PDF]

				W. Wisser, T. Fleck, D. Hutschala, and E. Wolner The 3rd hand - a simple but useful tool for beating heart total endoscopic coronary bypass grafting (BH-TECAB) Interactive CardioVascular and Thoracic Surgery, October 1, 2006; 5(5): 519 - 520. [Abstract] [Full Text] [PDF]

				C. L. Lake Simulation in Cardiology and Cardiothoracic and Vascular Surgery Seminars in Cardiothoracic and Vascular Anesthesia, December 1, 2005; 9(4): 325 - 333-1. [Abstract] [PDF]

				V. Falk, F. Mourgues, L. Adhami, S. Jacobs, H. Thiele, S. Nitzsche, F. W. Mohr, and E. Coste-Maniere Cardio Navigation: Planning, Simulation, and Augmented Reality in Robotic Assisted Endoscopic Bypass Grafting Ann. Thorac. Surg., June 1, 2005; 79(6): 2040 - 2047. [Abstract] [Full Text] [PDF]

				M. A. Talamini and E. J. Hanly Technology in the Operating Suite JAMA, February 16, 2005; 293(7): 863 - 866. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Bob B. Kiaii Thomas Walther Friedrich W. Mohr Volkmar Falk Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Jacobs, S. Articles by Falk, V. Search for Related Content PubMed PubMed Citation Articles by Jacobs, S. Articles by Falk, V. Related Collections Minimally invasive surgery