1. In-Vivo and In-Situ Compressive Properties of Porcine Abdominal Soft Tissues Jeffrey D. Brown MMVR2003 biorobotics laboratory university of washington 23-january-2003

2. 23-Jan-03 2 Introduction Surgery training ¤ Expensive! » Financial ($1M) costs; ethical, societal concerns ¤ Different modalities VR surgical simulation technology ¤ Reduced use of cadavers/animals/patients ¤ No risk of error to patient ¤ Allows training for difficult and/or rare scenarios

3. 23-Jan-03 3 Problem Statement Accurate tissue models are essential ¤ For realistic haptic feedback ¤ Need to represent in-vivo behavior ¤ Current simulators lack in-vivo models Hypothesis: There is a significant difference in mechanical properties in-vivo vs. postmortem ¤ No quantitative information on time course of change

4. 23-Jan-03 4 Background — Soft Tissues Highly nonlinear behavior ¤ Nonlinear stress-strain » not F = k*x! ¤ Anisotropic ¤ Heterogeneous ¤ Viscoelastic » Rate-dependent stress, creep, relaxation ¤ Hysteresis ¤ Strain history dependence » Preconditioning ¤ Behavior depends on many factors » Age, gender, pH, tissue health…

5. 23-Jan-03 5 Other Devices Ottensmeyer (2001, 2002) ¤ TeMPeST 1-D » in-vivo compression (indentation) Ottensmeyer (2001) Carter et al. (2001) Rosen et al. (1999) Carter et al. (2001) ¤ in-vivo indentation ¤ human data! Rosen et al. (1999)

6. 23-Jan-03 6 Preliminary Results — Grasping Measured grasping force applied during 3 surgical tasks for 5 expert surgeons ¤ 97.1% of all grasps (both hands) were held less than 10 sec ¤ Maximum force observed was 40 N (rarely) ¤ Majority of frequency content was below 3 Hz

7. 23-Jan-03 7 the MEG Cyclic compressions up to 3 Hz > 40 N grasping force Weighs about 0.7 kg (1.5 lbs)

8. 23-Jan-03 8 Methodology Animal testing ¤ pig, female, avg. weight 39.5 kg ¤ standard laparoscopic setup (11 mmHg) ¤ in-vivo: 9 pigs; in-situ: 4 pigs; ex-vivo: 1 cow liver » 2 pigs done both in-vivo and in-situ ¤ 7 organs tested » gallbladder » liver » small bowel » large bowel » spleen » stomach » urinary bladder

9. 23-Jan-03 9 Methodology — Cyclic Loading Haversine and constant velocity ¤ 0.25 - 3 Hz No preconditioning! ¤ We want 1st-squeeze behavior and squeeze history ¤ New site for each test

10. 23-Jan-03 10 Data Collection Data synchronized with video from camera

11. 23-Jan-03 11 Results — Cyclic Testing 1 Liver 1 Hz haversine in-vivo vs. in-situ 10 squeezes Same organ, different locations Remarks: ¤ More inter-squeeze variability in-vivo ¤ Beating heart, ventilator motion, re-perfusion Stress [MPa] Strain 0.5 0 0 0.02 0 0 in-vivo in-situ

12. 23-Jan-03 12 Results — Cyclic Testing 3 small bowels in-vivo Remarks: ¤ Low initial slope followed by sudden increase ¤ Squeezing contents (air, fluid, solid) then walls contacting Stress [MPa] Strain 0.5 0 0 0.04 0 0

13. 23-Jan-03 13 Results — Cyclic Testing 8 livers 8 spleens 3 small bowels in-vivo 1 Hz 1st squeezes only Remarks: ¤ Large variability Stress [MPa] Strain 0.5 0 0 0.04 0 0

14. 23-Jan-03 14 Methodology — Step Loading Measure stress relaxation ¤ Single steps; held for 60 sec ¤ Periodic steps » 10 sec hold » 2.5, 5, 10, 20, 30 sec off

15. 23-Jan-03 15 Results — Stress Relaxation 1 Liver in-vivo vs. in-situ Single step, 60 sec hold time Strains 21% - 41% Remarks: ¤ Higher percent decay in-situ ¤ Steady-state not reached in 60 sec, especially in-situ Normalized Stress Time [sec] 60 0 0 100% 50% in-vivo in-situ

16. 23-Jan-03 16 Results — Stress Relaxation 1 Liver in-vivo vs. in-situ Periodic steps ¤ 10 sec on / 2.5 sec off Strains 17% - 40% Remarks: ¤ Higher percent decay in-situ ¤ Behavior is similar to single step (little recovery between squeezes) Normalized Stress Time [sec] 35 0 0 100% 50% in-vivo in-situ

17. 23-Jan-03 17 Results — Stress Relaxation 1 Liver in-vivo vs. in-situ Periodic steps ¤ 10 sec on / 30 sec off Strains 21% - 51% Note: ¤ More recovery between squeezes than 10/2.5 ¤ More recovery in-vivo » Returns to 100% » Some >100% – Swelling? ¤ Some recovery in-situ Normalized Stress Time [sec] 90 0 0 100% 50% in-vivo in-situ

18. 23-Jan-03 18 Limitations Difficult to control in-vivo conditions ¤ Heartbeat, respiration, intra-abdominal pressure, anesthesia ¤ Hollow organs » Contents may be different in each location ¤ We want these effects included! Boundary conditions ¤ Organs may slide and rigidly translate instead of deform ¤ Hand tremor may be present in data Compliant mechanism Friction Contact detection Thin samples ¤ Large error in strain

19. 23-Jan-03 19 Conclusions in-vivo compression of soft tissues ¤ Abdominal organs relevant to laparoscopy ¤ Compressive loading in ranges of force and deformation observed in real surgeries Understanding of how tissue mechanical properties change postmortem

20. 23-Jan-03 20 Future Work Further verification of MEG accuracy ¤ Comparison of MEG to MTS Characterization of organ phantoms’ properties Sterilizable MEG for in-vivo human testing

21. 23-Jan-03 21 Acknowledgments Funding: ¤ Whitaker Foundation » Graduate Student Fellowship program ¤ Washington Research Foundation ¤ UW Department of Surgery » Center for Videoendoscopic Surgery Co-authors: ¤ Jacob Rosen, Yoon Sang Kim, Lily Chang, Mika Sinanan, Blake Hannaford

22. 23-Jan-03 22 Thank You! Biorobotics Lab http://brl.ee.washington.edu