537563
research-article2014
SRIXXX10.1177/1553350614537563Surgical InnovationFalkinger et al
Innovative Technologies
Design of a Test System for the Development of Advanced Video Chips and Software Algorithms
Surgical Innovation 2015, Vol. 22(2) 155–162 © The Author(s) 2014 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1553350614537563 sri.sagepub.com
Marita Falkinger, Dipl Ing1, Michael Kranzfelder, MD1,2, Dirk Wilhelm, MD1,2, Verena Stemp, BSc1, Susanne Koepf, BEng1, Judith Jakob, MSc3, Andreas Hille, Dipl Ing3, Wolfgang Endress, Dipl Ing3, Hubertus Feussner, MD1,2, and Armin Schneider, Dr Ing1
Abstract Background. Visual deterioration is a crucial point in minimally invasive surgery impeding surgical performance. Modern image processing technologies appear to be promising approaches for further image optimization by digital elimination of disturbing particles. To make them mature for clinical application, an experimental test environment for evaluation of possible image interferences would be most helpful. Methods. After a comprehensive review of the literature (MEDLINE, IEEE, Google Scholar), a test bed for generation of artificial surgical smoke and mist was evolved. Smoke was generated by a fog machine and mist produced by a nebulizer. The size of resulting droplets was measured microscopically and compared with biological smoke (electrocautery) and mist (ultrasound dissection) emerging during minimally invasive surgical procedures. Results. The particles resulting from artificial generation are in the range of the size of biological droplets. For surgical smoke, the droplet dimension produced by the fog machine was 4.19 µm compared with 4.65 µm generated by electrocautery during a surgical procedure. The size of artificial mist produced by the nebulizer ranged between 45.38 and 48.04 µm compared with the range between 30.80 and 56.27 µm that was generated during minimally invasive ultrasonic dissection. Conclusion. A suitable test bed for artificial smoke and mist generation was developed revealing almost identical droplet characteristics as produced during minimally invasive surgical procedures. The possibility to generate image interferences comparable to those occurring during laparoscopy (electrocautery and ultrasound dissection) provides a basis for the future development of image processing technologies for clinical applications. Keywords surgical smoke and mist, artificial generation, minimally invasive surgery, digital image processing
Introduction A clear and sharp field of vision is an important aspect in minimally invasive surgery, since poor visibility impairs both identification of structures and quality of surgical manipulations. Beyond the fact that laparoscope lenses often get fogged due to direct tissue contact, the use of electrosurgical devices and ultrasound dissectors additionally disturb the surgeon’s field of vision. Electrocautery (diathermy) is one method of energized dissection. The patient is part of the electrical circuit by use of an inactive electrode for direct current return and the surgical instrument operates as a handheld active electrode. The applied current generates heat of high temperature which brings intracellular water to seethe and desiccates cellular material with a subsequent dissection or coagulation of the manipulated tissue.1 This process of
burning produces smoke plumes. Ultrasonic dissectors are mostly used during large resections to reduce tissue injury and duration of surgery. Cell structures are dissected by the coagulating and dissecting effect of high-frequency (HF) ultrasound vibration.2 Thus, tissue particles—in 1
Research Group MITI (Minimally invasive Interdisciplinary Therapeutical Intervention), Klinikum rechts der Isar, Technical University Munich, Munich, Germany 2 Department of Surgery, Klinikum rechts der Isar, Technical University Munich, Munich, Germany 3 C.R.S. iiMotion GmbH, Villingen-Schwenningen, Germany Corresponding Author: Michael Kranzfelder, Department of Surgery, Klinikum rechts der Isar, Technical University Munich, Ismaningerstraße 22, Munich, D-81675, Germany. Email: [email protected]
Downloaded from sri.sagepub.com at University of Sydney on November 15, 2015
156
Surgical Innovation 22(2)
Figure 1. Fusion of 3 images with different exposure times (left) to 1 high dynamic range (HDR) image with consistent illumination (right).
particular fat—are mobilized and accelerated causing vaporization mist. Both smoke plumes and fat bubbles are floating through the insufflated abdominal cavity, adhering to the lens of the laparoscope3 and deteriorating visualization significantly. Up to now, the only way to reestablish a clear laparoscopic view is to exchange the intra-abdominal gas.2,3 Recently, an additional solution was introduced to remove smoke particles by electrostatic precipitation.4 However, either method is more or less time consuming, results in high gas consumption and requires the procurement of additional devices. Modern image processing technologies could offer new alternatives of image optimization by digital elimination of disturbing particles. In particular, high dynamic range (HDR) cameras possess a promising potential to overcome visualization problems by virtual elimination of fog and mist, exposure optimization, and reflection suppression. However, considerable research and development efforts are still needed. As conventional laparoscopic camera systems (standard dynamic range [SDR] cameras) only achieve a dynamic range of up to 60 dB, precise recording of the naturally high differences of brightness is not possible, since the maximum ratio between the darkest and the brightest point is only 1000:1. In contrast, HDR cameras are able to record a dynamic range up to 120 dB, which means a contrast ratio of 1 000 000:1. To achieve such high dynamic ranges, HDR camera systems take simultaneously 3 or more pictures with
different exposure times.5 By digital fusion of the single pictures to 1 image,6 illumination can be optimized (Figure 1). Since the HDR algorithms have to be adapted individually to ambient conditions, a precise adjustment to medical requirements in laparoscopic surgery (eg, small optics) is necessary. One precondition for making HDR mature for clinical application is therefore a suitable test bed that reproduces surgical smoke and mist reliably, thus allowing optimizing technical solutions in an iterative way. The creation of a test bed has to be based on detailed information about these disturbing effects. It was the aim of this study to develop an experimental setup to produce artificial smoke and mist formations similar to those that are generated in minimally invasive surgical procedures and to evaluate first applications of image improvements of HDR technology.
Methods Systematic Review of the Literature First a comprehensive search of the published literature on properties and conditions of surgical smoke and mist as well as methods of its artificial generation was conducted in the MEDLINE, IEEE, and Google Scholar databases using the medical heading term electrocautery, surgical smoke, ultrasonic dissection, mist production, and minimally invasive surgery. All retrieved abstracts were reviewed to obtain full-text articles of potentially
Downloaded from sri.sagepub.com at University of Sydney on November 15, 2015
157
Falkinger et al
Figure 2. Test chamber for generation of video sequences of deteriorating effects. The box includes 1 trocar port (A) for inserting a laparoscope connected with a laparoscopic video system and 2 inflow tubes (B) for connecting the fog machine or the nebulizer. The optical light conductor (C) in the middle of the chamber is controlled by a LabView routine.
relevant reports, studies, case series, and reviews. Liberati et al7 give an overview of data extraction according to the PRISMA statement. It was our purpose to compile all available knowledge on the features, quality, and generation of smoke and mist in minimally invasive surgery.
Experimental Setting The development of new software algorithms for digital improvement of minimally invasive imaging implies a standardized test setting that facilitates video recordings without interferences as well as with adjustable disturbances in a defined and reproducible manner. In addition to sufficient illumination, it should allow to distinguish between surgical smoke produced by electrocautery and mist caused by ultrasonic dissection.
Artificial Smoke and Mist Our developed experimental test bed (Figure 2) consists of a transparent, conical sealed chamber with a volume of 23.5 L (top 220 × 330 × 280 mm, bottom 260 × 370 × 280mm) with 1 trocar port for insertion of the laparoscope and 2 tube tunnels for inflow of smoke and particles (side and top position of the chamber). Within the chamber, an optical light measurement system (optoCONTROL CLS-K, Micro-Epsilon, Ortenburg, Germany) was implemented for regulation and measurement of the fog, controlled by a purpose built LabView routine (National Instruments, Munich, Germany). The
optical measurement system consists of an optical transmitter generating infrared and ultraviolet light and a receiverm which delivers the corresponding output voltage signal. The light signal is altered by smoke and mist leading to a reduced output voltage, which is then captured by the LabView routine. To imitate the surgical smoke, a commercially available fog machine (N10, eurolite, Waldbüttelbrunn, Germany) was used with a water and glycol- or glycerine-based fluid. The solution is pumped into a heat exchanger where the fluid vaporizes at high temperature. The fluid expands by transition from liquid in gaseous aggregation state. The increased pressure then pushes the aerosol through a nozzle into the test chamber. The smoke effect, which is important to simulate smoke from electrocautery, appears when the aerosol comes in contact with ambient air. Varying effects (stability time and smoke intensity) depend on fluid composition and inflow duration; the latter was controlled by the LabView routine. A smoke inflow duration of 1, 3, 5, and 7 seconds was evaluated and compared with that of a real surgical procedure. Additionally, the smoke retention time within the chamber and the suitability of the inflow tubes (side and top position of the chamber) were tested. To generate the most realistic surgical smoke, further smoke generators besides the fog machine (smoke pens, pills, and powders) were examined. To produce image interferences similar to the mist generated during ultrasonic dissection 2 jet nebulizers originally developed for inhalation therapies (Turboboy
Downloaded from sri.sagepub.com at University of Sydney on November 15, 2015
158
Surgical Innovation 22(2)
SX and Pari Sole N, PariTec Starnberg, Germany) were used. The fog fluids were mixed with pressurized air produced by an integrated compressor and then sputtered by a nozzle adapter. Thereby, droplets of different sizes are generated. Whereas larger droplets condensed on the walls of an additionally created chamber, smaller droplets with a predefined size (filter plate) remained and left the jet nebulizer through another nozzle connected to the test chamber inflow tubes. The jet nebulizers were tested with saline solution, purified water and vegetable oil to find an optimal solution ratio for mist production equal to that generated by ultrasound dissectors.
Biological Smoke and Mist To ensure a simulation similar to real surgical conditions, the size of artificial smoke and mist particles was quantified by a specially developed method facilitating comparison to biological droplets generated by surgical devices. Therefore, biological droplets were generated ex vivo with the HF device ERBOTOM T400C (ERBE, Tübingen, Germany) and the ultrasonic dissector SonoSurg G2 (Olympus, Tokyo, Japan) using lard, porcine, and human fat from 11 patients of different age and body mass index (BMI). Prior to examination, ethical approval of our local ethics committee was obtained. All patients had given written informed consent to participate in our study.
Data Analysis To ensure accurate particle analysis, both artificial and biological droplets were collected on microscope slides fixed 5 cm above the manipulated tissue and flattened by cover glasses after 5 seconds of dissection. Subsequently, the 2-dimensional (2D) expansion of the generated droplets was measured with a light microscope Axioskop 40 (Zeiss, Munich, Germany) with 10× and 20× magnification. For each tissue sample, the 2D diameter of n = 30 droplets was measured in 3 axes (horizontal, vertical, diagonal) and the mean diameter (d) ± standard deviation (SD) calculated.
Results Systematic Review of the Literature The MEDLINE search with the medical heading term “surgery AND electrocautery AND ultrasonic” retrieved 2918 results. To obtain more detailed data, the articles were filtered using the heading term “surgery AND electrocautery AND ultrasonic AND (smoke OR mist OR fog OR aerosol OR particle).” The majority of retrieved articles (n = 14) referred to operation techniques and health risks
Figure 3. Flowchart of the literature review process according to Liberati et al.7 The first search in MEDLINE with the medical heading term “surgery AND electrocautery AND ultrasonic” revealed 2918 results and was filtered by “surgery AND electrocautery AND ultrasonic AND (smoke OR mist OR fog OR aerosol OR particle).”
due to the use of electrocautery for the operating team. Two MEDLINE articles were included into the review. Additional search of the IEEE database, Google Scholar, and reference indices of retrieved articles eventually revealed a total of 15 articles reporting on composition and droplet size of surgical smoke and mist (Figure 3). Finally, 7 full-text articles were identified with detailed particle characteristics (size, composition, volume), which are shown in Table 1. Plumes of aerosols mainly consist of aliphatic and aromatic hydrocarbons, aldehydes, phenols, nitriles, and fatty acids, as well as carbon dioxide, carbon monoxide, ammonia, and hydrogen cyanide.8 Several articles verified these components describing droplet diameters between d = 0.05 and 25 µm.9-12 According to Hensman et al,13 the type of insufflated gas has no influence on the structure and chemical constituents of particles. Surgical smoke generated in air, CO2, and helium showed similar consistency and particles.13 Ultrasonic dissection produces approximately one fourth of smoke concentration compared to electrocautery application.14 For vapor during ultrasonic dissection, the measured particle sizes ranged between d = 0.35 and 6.5 µm.10,15 In most research projects, mist and smoke production was achieved by combination of ultrasonic dissectors with HF-devices, revealing 2 different particle populations: small spherical homogeneous particles and larger irregular particles with cellular components.3 Corabajo-Rodríguez et al11 proved that the plumes consist
Downloaded from sri.sagepub.com at University of Sydney on November 15, 2015
159
Falkinger et al Table 1. Review of the Literature (MEDLINE, IEEE, Google Scholar) According to Liberati et al7: 7 Full-Text Articles Were Identified With Detailed Particle Characteristics (Size, Composition, Volume). Author(s) 10
Alp et al
Instruments Used Electrocautery
Particle Size (µm)
Composition
Particle Volume
0.07
Ultrasonic scalpels
0.35-6.5
Liquid aerosols
Carbajo-Rodríguez et al11
Electrocautery Ultrasonic scalpel
0.07-25 n.s.
Weld et al3
Bipolar macroforceps Harmonic scalpel Floating ball Monopolar shears Electrocautery Ultrasonic scalpel
95% water and 5% chemical products: sodium, magnesium, calcium, iron, carbon, oxygen aldehyde, toluene, ethylbenzene, xylene, carbon monoxide n.s.
Barrett and Garber15
Brüske-Hohlfeld et al17
DesCôteaux et al12
Electrocautery
Schneider et al16
Different types of ultrasound shears
0.07-0.89 0.07-0.10 0.07-1.08 0.10-0.92 0.07-0.31 0.35-6.5
0.1-0.5 2-24 n.s.
of 95% water vapour and 5% chemical compounds/cell remnants. Evaporation of liquids creates the group of small homogeneous spheres, composed mainly of water and small amounts of sodium, potassium, magnesium, calcium, and iron. In contrast, low-temperature vaporization contains tissue fragments, which can also carry infectious or viable cells and present the latter population.10 According to Schneider et al,16 different ultrasonic shears induce varieties in mist generation. The authors demonstrated that the amount of vapor production depends on the technical design of ultrasonic shears. Brüske-Hohlfeld et al17 performed particle measurements with a condensation particle counter in several operation rooms during different surgical procedures, revealing a particle diameter range of d = 10 nm to 1 µm. Weld et al3 used the same setup in a hermetic chamber utilizing an aerodynamic particle sizer (time-of-flight technique) and electrostatic classifier (electrical mobility principle) for the calculation of the mean particle size. In 1988, Vanderpool anf Rubow18 developed a method to
Hydrocarbons, acrylonitrile, fatty acids, phenols
Sodium, chloride, magnesium, calcium, potassium, carbon, oxygen n.s.
n.s. Four times more compared with dissection of similar amount of tissue by ultrasonic scalpel Fatty tissue generates 17-23 times more than lean tissue Influenced by the destroyed type of tissue Particle number concentration in an condensation particle counter Depending on surgical procedure, tissue, intensity of energy; measurements in an condensation particle counter n.s. Measured with infrared light transmission
generate monodisperse aerosols in a diameter range of d = 1 to 20 µm and demonstrated that both the generation method and the physical/chemical aerosol characteristics have an influence on the particle size.
Experimental Setting Artificial Smoke and Mist Our purpose built test chamber allowed us to produce artificial smoke and mist close to real surgical conditions. The best results of the irritating effect of smoke plumes deteriorating the surgeon’s view of the operation field were achieved by application of the commercially available fog machine N10 (eurolite, Waldbüttelbrunn, Germany). The other tested smoke/fog generators, smoke pens, pills, and powders produced a nonrealistic smoke formation as the smoke was either too dense and intensive or the duration until the smoke disperses was too short. The most realistic smoke plumes (comparable to
Downloaded from sri.sagepub.com at University of Sydney on November 15, 2015
160
Surgical Innovation 22(2)
Figure 4. Mean particle diameter (smoke and mist in micrometers) subject to its generation (artificial, porcine, human). Electrocautery on human or porcine fat (left) shows comparable results with a mean diameter of dHFh = 4.65 µm and dHFp = 4.14 µm to the artificial generation with the fog machine (df = 4.19 µm). Fatty droplets of ultrasonic dissection are well emulated by a nebulizer (d = 46.71 µm). The comparison with human (dbF = 45.08 µm) and porcine (dbL = 45.45 µm) droplets on (right) side does not show any significant difference.
those of biological electrocautery) were achieved with a smoke inflow of 3 and 5 seconds (smoke inflow 1 second, smoke too light; 7 seconds, smoke too dense). Additionally, the inflow tube positioned on the side of the test chamber yielded a more convenient smoke distribution (steady dispersal) than the tube positioned on the top of the box (deepening dispersal). The smoke retention time in the test chamber was 14 to 15 minutes. The microscopic examination of the particle sizes generated by the fog machine revealed comparable results to those of electrocautery aerosols with a mean diameter of df = 4.19 µm. For the artificial generation of fatty droplets, 2 jet nebulizers were tested with 3 different fluids. The vaporizer Turboboy SX (PariTec, Starnberg, Germany) created too small droplets (dn1 = 10.65 µm) and was therefore excluded. Saline solution and purified water dried too fast that only downsized or even no droplets were visible under the microscope. The best results were achieved by utilizing the Nebulizer Sole N (PariTec, Starnberg, Germany) with vegetable oil generating particles with a mean diameter range of dn2 = 45.38 ± 6.16 to 48.04 ± 8.50 µm.
Biological Smoke and Mist The artificially generated particles were compared to droplets produced by HF electrocautery and ultrasonic dissection applied on lard, porcine and human visceral fat (Figure 4). The mean droplet diameter produced by the HF device Erbotom T400C (Erbe, Tübingen, Germany) was dHFh = 4.65 µm for human fat and dHFp = 4.14 µm for porcine fat.
Figure 5. Particle diameter during ultrasonic dissection depends on body mass index (BMI) and age of the patients (increase of particle size with increasing BMI).
For ultrasound dissection, a mean diameter size of dUSl = 45.45 ± 12.33 µm (lard) and dUSh = 30.80 ± 4.01 to 56.27 ± 17.01 µm (human fat) was noted (Figure 4). Between the different patients, a high droplet size variation was observed with a tendency that the droplet diameter increases with BMI and age (Figure 5).
Discussion The improvement of visualization is an essential precondition for the future development of minimally invasive interventions and further reduction of the surgical trauma by novel intervention techniques. Several research groups are working on miniaturized cameras to make Monoport or natural orifice translumenal endoscopic surgery (NOTES) mature for routine clinical application.19,20 New developments in medical imaging techniques now offer the chance to overcome existing obstacles such as complex and defective intra-abdominal camera fixation. A promising technology is the use of HDR imaging. HDR cameras grab 2 or more images simultaneously with different exposure levels and combine these to 1 image with optimal brightness. Thus, small optics placed almost anywhere on the laparoscopic instrument or solitarily introduced through tiny ports can be used for advanced visualization of the abdominal cavity. Additionally, HDR video chips offer the opportunity to digitally reduce interfering effects like electrocauterization smoke or mist from ultrasound dissection by computing algorithms in the visualization chain. However, for a sufficient reduction of disturbing particles by computer algorithms directly on the video processor, detailed knowledge about the particle size and retention time inside is essential. We therefore developed
Downloaded from sri.sagepub.com at University of Sydney on November 15, 2015
161
Falkinger et al an experimental test bed for particle size and composition analysis that can be used with both organic and artificially generated particle samples to facilitate testing of different camera systems and developing digital image correction algorithms in a standardized and reproducible setup. Within a specially developed test chamber, artificial smoke was produced by a fog machine revealing almost identical particle diameters (df = 4.19 µm) as compared with smoke generated by surgical electrocautery (dHFh = 4.65 µm [human fat] and dHFp = 4.14 µm [porcine fat]). Mist, which is typically evolving during ultrasound tissue dissection, was generated by a nebulizer. Artificial droplet size ranged between dn2 = 45.38 and 48.04 µm and biological droplet size between dUSh = 30.80 and 56.27 µm. A tendency of enlarged particle diameter with increased BMI and age was noted. Our results are in line with those published by other research groups.11,12 For generation of artificial smoke, droplet diameters can vary due to the composition of fog fluid, which is typically based on distilled water and propylene glycol.18 The fog retention time depends on additives for stabilisation like glycerol. In our experimental setting, the smoke was evenly spread after 20 seconds in the test chamber and remained for about 15 minutes until dissolving. For development of smoke reduction algorithms, especially the first 20 seconds of smoke spreading are important, as video sequences of shifting smoke plumes are needed for digital image processing. In addition to simulating and quantitative measuring surgical smoke and mist, the presented test bed was also partly used for developing image processing algorithms and to evaluate first applications of image improvements of HDR technology. Therefore, reflections on abdominal structures and over- or underexposed (shady) images needed to be generated. For this, the shutter of a fixed laparoscopic camera used in our test chamber was varied from 1.28 to 1/10 000 second and consecutive sequences of standardized test images of the USC-SIPI (University of Southern California Signal and Image Processing Institute) Image Database21 were recorded. Thereby, we achieved the generation of identical deteriorated images in different exposure times as the basis for future testing and optimizing of a first prototypic HDR camera system. In conclusion, we developed a suitable test bed for artificial smoke and mist generation revealing almost identical droplet characteristics as produced during minimally invasive procedures (electrocautery and ultrasound dissection). The possibility to generate image interferences comparable to those occurring during laparoscopy provides a basis for the future development of image processing technologies for clinical applications.
Declaration of Conflicting Interests The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported in part by AiF Projekt GmbH.
References 1. Fitzgerald JE, Malik M, Ahmed I. A single-blind controlled study of electrocautery and ultrasonic scalpel smoke plumes in laparoscopic surgery. Surg Endosc. 2012;26:337-342. 2. Gossot D, Buess G, Cuschieri A, et al. Ultrasonic dissection for endoscopic surgery. Surg Endosc. 1999;13: 412-417. 3. Weld KJ, Dryer S, Ames CD, et al. Analysis of surgical smoke produced by various energy-based instruments and effect on laparoscopic visibility. J Endourol. 2007;21: 347-351. 4. Ansell J, Warren N, Sibbons P, Torkington J. The innervision surgical smoke removal system. Int J Surg. 2012;10(8):S43. 5. Pece F, Kautz J, eds. Bitmap Movement Detection: HDR for Dynamic Scenes. Proceedings of the IEEE Conference on Visual Media Production (CVMP) 2010. Piscataway, NJ: IEEE. 6. Debevec PE, Malik J, eds. Recovering High Dynamic Range Radiance Maps From Photographs. ACM SIGGRAPH 2008 Classes. New York, NY: ACM; 2008. 7. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. Ann Intern Med. 2009;151:W-65-W-94. 8. Wu Y-C, Tang C-S, Huang H-Y, et al. Chemical production in electrocautery smoke by a novel predictive model. Eur Surg Res. 2011;46:102-107. 9. Al Sahaf O, Vega-Carrascal I, Cunningham F, McGrath J, Bloomfield F. Chemical composition of smoke produced by high-frequency electrosurgery. Irish J Med Sci. 2007;176:229-232. 10. Alp E, Bijl D, Bleichrodt R, Hansson B, Voss A. Surgical smoke and infection control. J Hosp Infect. 2006;62:1-5. 11. Carbajo-Rodríguez H, Aguayo-Albasini JL, Soria-Aledo V, García-López C. Surgical smoke: risks and preventive measures. Cir Esp (Engl Ed). 2009;85:274-279. 12. DesCôteaux J-G, Picard P, Poulin E, Baril M. Preliminary study of electrocautery smoke particles produced in vitro and during laparoscopic procedures. Surg Endosc. 1996;10:152-158. 13. Hensman C, Baty D, Willis R, Cuschieri A. Chemical composition of smoke produced by high-frequency electrosurgery in a closed gaseous environment. Surg Endosc. 1998;12:1017-1019. 14. Ott DE, Moss E, Martinez K. Aerosol exposure from an ultrasonically activated (harmonic) device. J Am Assoc Gynecol Laparosc. 1998;5:29-32.
Downloaded from sri.sagepub.com at University of Sydney on November 15, 2015
162
Surgical Innovation 22(2)
15. Barrett WL, Garber SM. Surgical smoke: a review of the literature. Surg Endosc. 2003;17:979-987. 16. Schneider A, Doundoulakis E, Can S, Fiolka A, Wilhelm D, Feußner H. Evaluation of mist production and tissue dissection efficiency using different types of ultrasound shears. Surg Endosc. 2009;23:2822-2826. 17. Brüske-Hohlfeld I, Preissler G, Jauch K-W, et al. Surgical smoke and ultrafine particles. J Occup Med Toxicol. 2008;3:31. 18. Vanderpool R, Rubow K. Generation of large, solid, monodisperse calibration aerosols. Aerosol Sci Technol. 1988;9:65-69.
19. Abu Gazala M, Shussman N, Abu Gazala S, et al. Miniature camera for enhanced visualization for single-port surgery and NOTES. J Laparoendosc Adv Surg Tech. 2012;22: 984-988. 20. Chang VC, Tang S-J, Swain CP, et al. A randomized comparison of laparoscopic, flexible endoscopic, and wired and wireless magnetic cameras on ex vivo and in vivo NOTES surgical performance. Surg Innov. 2013;20:395-402. 21. Weber A. The USC-SIPI image database. Signal and Image Processing Institute of the University of Southern California. http://sipi.usc.edu/services/database. Accessed May 19, 2014.
Downloaded from sri.sagepub.com at University of Sydney on November 15, 2015
research-article2014
SRIXXX10.1177/1553350614537563Surgical InnovationFalkinger et al
Innovative Technologies
Design of a Test System for the Development of Advanced Video Chips and Software Algorithms
Surgical Innovation 2015, Vol. 22(2) 155–162 © The Author(s) 2014 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1553350614537563 sri.sagepub.com
Marita Falkinger, Dipl Ing1, Michael Kranzfelder, MD1,2, Dirk Wilhelm, MD1,2, Verena Stemp, BSc1, Susanne Koepf, BEng1, Judith Jakob, MSc3, Andreas Hille, Dipl Ing3, Wolfgang Endress, Dipl Ing3, Hubertus Feussner, MD1,2, and Armin Schneider, Dr Ing1
Abstract Background. Visual deterioration is a crucial point in minimally invasive surgery impeding surgical performance. Modern image processing technologies appear to be promising approaches for further image optimization by digital elimination of disturbing particles. To make them mature for clinical application, an experimental test environment for evaluation of possible image interferences would be most helpful. Methods. After a comprehensive review of the literature (MEDLINE, IEEE, Google Scholar), a test bed for generation of artificial surgical smoke and mist was evolved. Smoke was generated by a fog machine and mist produced by a nebulizer. The size of resulting droplets was measured microscopically and compared with biological smoke (electrocautery) and mist (ultrasound dissection) emerging during minimally invasive surgical procedures. Results. The particles resulting from artificial generation are in the range of the size of biological droplets. For surgical smoke, the droplet dimension produced by the fog machine was 4.19 µm compared with 4.65 µm generated by electrocautery during a surgical procedure. The size of artificial mist produced by the nebulizer ranged between 45.38 and 48.04 µm compared with the range between 30.80 and 56.27 µm that was generated during minimally invasive ultrasonic dissection. Conclusion. A suitable test bed for artificial smoke and mist generation was developed revealing almost identical droplet characteristics as produced during minimally invasive surgical procedures. The possibility to generate image interferences comparable to those occurring during laparoscopy (electrocautery and ultrasound dissection) provides a basis for the future development of image processing technologies for clinical applications. Keywords surgical smoke and mist, artificial generation, minimally invasive surgery, digital image processing
Introduction A clear and sharp field of vision is an important aspect in minimally invasive surgery, since poor visibility impairs both identification of structures and quality of surgical manipulations. Beyond the fact that laparoscope lenses often get fogged due to direct tissue contact, the use of electrosurgical devices and ultrasound dissectors additionally disturb the surgeon’s field of vision. Electrocautery (diathermy) is one method of energized dissection. The patient is part of the electrical circuit by use of an inactive electrode for direct current return and the surgical instrument operates as a handheld active electrode. The applied current generates heat of high temperature which brings intracellular water to seethe and desiccates cellular material with a subsequent dissection or coagulation of the manipulated tissue.1 This process of
burning produces smoke plumes. Ultrasonic dissectors are mostly used during large resections to reduce tissue injury and duration of surgery. Cell structures are dissected by the coagulating and dissecting effect of high-frequency (HF) ultrasound vibration.2 Thus, tissue particles—in 1
Research Group MITI (Minimally invasive Interdisciplinary Therapeutical Intervention), Klinikum rechts der Isar, Technical University Munich, Munich, Germany 2 Department of Surgery, Klinikum rechts der Isar, Technical University Munich, Munich, Germany 3 C.R.S. iiMotion GmbH, Villingen-Schwenningen, Germany Corresponding Author: Michael Kranzfelder, Department of Surgery, Klinikum rechts der Isar, Technical University Munich, Ismaningerstraße 22, Munich, D-81675, Germany. Email: [email protected]
Downloaded from sri.sagepub.com at University of Sydney on November 15, 2015
156
Surgical Innovation 22(2)
Figure 1. Fusion of 3 images with different exposure times (left) to 1 high dynamic range (HDR) image with consistent illumination (right).
particular fat—are mobilized and accelerated causing vaporization mist. Both smoke plumes and fat bubbles are floating through the insufflated abdominal cavity, adhering to the lens of the laparoscope3 and deteriorating visualization significantly. Up to now, the only way to reestablish a clear laparoscopic view is to exchange the intra-abdominal gas.2,3 Recently, an additional solution was introduced to remove smoke particles by electrostatic precipitation.4 However, either method is more or less time consuming, results in high gas consumption and requires the procurement of additional devices. Modern image processing technologies could offer new alternatives of image optimization by digital elimination of disturbing particles. In particular, high dynamic range (HDR) cameras possess a promising potential to overcome visualization problems by virtual elimination of fog and mist, exposure optimization, and reflection suppression. However, considerable research and development efforts are still needed. As conventional laparoscopic camera systems (standard dynamic range [SDR] cameras) only achieve a dynamic range of up to 60 dB, precise recording of the naturally high differences of brightness is not possible, since the maximum ratio between the darkest and the brightest point is only 1000:1. In contrast, HDR cameras are able to record a dynamic range up to 120 dB, which means a contrast ratio of 1 000 000:1. To achieve such high dynamic ranges, HDR camera systems take simultaneously 3 or more pictures with
different exposure times.5 By digital fusion of the single pictures to 1 image,6 illumination can be optimized (Figure 1). Since the HDR algorithms have to be adapted individually to ambient conditions, a precise adjustment to medical requirements in laparoscopic surgery (eg, small optics) is necessary. One precondition for making HDR mature for clinical application is therefore a suitable test bed that reproduces surgical smoke and mist reliably, thus allowing optimizing technical solutions in an iterative way. The creation of a test bed has to be based on detailed information about these disturbing effects. It was the aim of this study to develop an experimental setup to produce artificial smoke and mist formations similar to those that are generated in minimally invasive surgical procedures and to evaluate first applications of image improvements of HDR technology.
Methods Systematic Review of the Literature First a comprehensive search of the published literature on properties and conditions of surgical smoke and mist as well as methods of its artificial generation was conducted in the MEDLINE, IEEE, and Google Scholar databases using the medical heading term electrocautery, surgical smoke, ultrasonic dissection, mist production, and minimally invasive surgery. All retrieved abstracts were reviewed to obtain full-text articles of potentially
Downloaded from sri.sagepub.com at University of Sydney on November 15, 2015
157
Falkinger et al
Figure 2. Test chamber for generation of video sequences of deteriorating effects. The box includes 1 trocar port (A) for inserting a laparoscope connected with a laparoscopic video system and 2 inflow tubes (B) for connecting the fog machine or the nebulizer. The optical light conductor (C) in the middle of the chamber is controlled by a LabView routine.
relevant reports, studies, case series, and reviews. Liberati et al7 give an overview of data extraction according to the PRISMA statement. It was our purpose to compile all available knowledge on the features, quality, and generation of smoke and mist in minimally invasive surgery.
Experimental Setting The development of new software algorithms for digital improvement of minimally invasive imaging implies a standardized test setting that facilitates video recordings without interferences as well as with adjustable disturbances in a defined and reproducible manner. In addition to sufficient illumination, it should allow to distinguish between surgical smoke produced by electrocautery and mist caused by ultrasonic dissection.
Artificial Smoke and Mist Our developed experimental test bed (Figure 2) consists of a transparent, conical sealed chamber with a volume of 23.5 L (top 220 × 330 × 280 mm, bottom 260 × 370 × 280mm) with 1 trocar port for insertion of the laparoscope and 2 tube tunnels for inflow of smoke and particles (side and top position of the chamber). Within the chamber, an optical light measurement system (optoCONTROL CLS-K, Micro-Epsilon, Ortenburg, Germany) was implemented for regulation and measurement of the fog, controlled by a purpose built LabView routine (National Instruments, Munich, Germany). The
optical measurement system consists of an optical transmitter generating infrared and ultraviolet light and a receiverm which delivers the corresponding output voltage signal. The light signal is altered by smoke and mist leading to a reduced output voltage, which is then captured by the LabView routine. To imitate the surgical smoke, a commercially available fog machine (N10, eurolite, Waldbüttelbrunn, Germany) was used with a water and glycol- or glycerine-based fluid. The solution is pumped into a heat exchanger where the fluid vaporizes at high temperature. The fluid expands by transition from liquid in gaseous aggregation state. The increased pressure then pushes the aerosol through a nozzle into the test chamber. The smoke effect, which is important to simulate smoke from electrocautery, appears when the aerosol comes in contact with ambient air. Varying effects (stability time and smoke intensity) depend on fluid composition and inflow duration; the latter was controlled by the LabView routine. A smoke inflow duration of 1, 3, 5, and 7 seconds was evaluated and compared with that of a real surgical procedure. Additionally, the smoke retention time within the chamber and the suitability of the inflow tubes (side and top position of the chamber) were tested. To generate the most realistic surgical smoke, further smoke generators besides the fog machine (smoke pens, pills, and powders) were examined. To produce image interferences similar to the mist generated during ultrasonic dissection 2 jet nebulizers originally developed for inhalation therapies (Turboboy
Downloaded from sri.sagepub.com at University of Sydney on November 15, 2015
158
Surgical Innovation 22(2)
SX and Pari Sole N, PariTec Starnberg, Germany) were used. The fog fluids were mixed with pressurized air produced by an integrated compressor and then sputtered by a nozzle adapter. Thereby, droplets of different sizes are generated. Whereas larger droplets condensed on the walls of an additionally created chamber, smaller droplets with a predefined size (filter plate) remained and left the jet nebulizer through another nozzle connected to the test chamber inflow tubes. The jet nebulizers were tested with saline solution, purified water and vegetable oil to find an optimal solution ratio for mist production equal to that generated by ultrasound dissectors.
Biological Smoke and Mist To ensure a simulation similar to real surgical conditions, the size of artificial smoke and mist particles was quantified by a specially developed method facilitating comparison to biological droplets generated by surgical devices. Therefore, biological droplets were generated ex vivo with the HF device ERBOTOM T400C (ERBE, Tübingen, Germany) and the ultrasonic dissector SonoSurg G2 (Olympus, Tokyo, Japan) using lard, porcine, and human fat from 11 patients of different age and body mass index (BMI). Prior to examination, ethical approval of our local ethics committee was obtained. All patients had given written informed consent to participate in our study.
Data Analysis To ensure accurate particle analysis, both artificial and biological droplets were collected on microscope slides fixed 5 cm above the manipulated tissue and flattened by cover glasses after 5 seconds of dissection. Subsequently, the 2-dimensional (2D) expansion of the generated droplets was measured with a light microscope Axioskop 40 (Zeiss, Munich, Germany) with 10× and 20× magnification. For each tissue sample, the 2D diameter of n = 30 droplets was measured in 3 axes (horizontal, vertical, diagonal) and the mean diameter (d) ± standard deviation (SD) calculated.
Results Systematic Review of the Literature The MEDLINE search with the medical heading term “surgery AND electrocautery AND ultrasonic” retrieved 2918 results. To obtain more detailed data, the articles were filtered using the heading term “surgery AND electrocautery AND ultrasonic AND (smoke OR mist OR fog OR aerosol OR particle).” The majority of retrieved articles (n = 14) referred to operation techniques and health risks
Figure 3. Flowchart of the literature review process according to Liberati et al.7 The first search in MEDLINE with the medical heading term “surgery AND electrocautery AND ultrasonic” revealed 2918 results and was filtered by “surgery AND electrocautery AND ultrasonic AND (smoke OR mist OR fog OR aerosol OR particle).”
due to the use of electrocautery for the operating team. Two MEDLINE articles were included into the review. Additional search of the IEEE database, Google Scholar, and reference indices of retrieved articles eventually revealed a total of 15 articles reporting on composition and droplet size of surgical smoke and mist (Figure 3). Finally, 7 full-text articles were identified with detailed particle characteristics (size, composition, volume), which are shown in Table 1. Plumes of aerosols mainly consist of aliphatic and aromatic hydrocarbons, aldehydes, phenols, nitriles, and fatty acids, as well as carbon dioxide, carbon monoxide, ammonia, and hydrogen cyanide.8 Several articles verified these components describing droplet diameters between d = 0.05 and 25 µm.9-12 According to Hensman et al,13 the type of insufflated gas has no influence on the structure and chemical constituents of particles. Surgical smoke generated in air, CO2, and helium showed similar consistency and particles.13 Ultrasonic dissection produces approximately one fourth of smoke concentration compared to electrocautery application.14 For vapor during ultrasonic dissection, the measured particle sizes ranged between d = 0.35 and 6.5 µm.10,15 In most research projects, mist and smoke production was achieved by combination of ultrasonic dissectors with HF-devices, revealing 2 different particle populations: small spherical homogeneous particles and larger irregular particles with cellular components.3 Corabajo-Rodríguez et al11 proved that the plumes consist
Downloaded from sri.sagepub.com at University of Sydney on November 15, 2015
159
Falkinger et al Table 1. Review of the Literature (MEDLINE, IEEE, Google Scholar) According to Liberati et al7: 7 Full-Text Articles Were Identified With Detailed Particle Characteristics (Size, Composition, Volume). Author(s) 10
Alp et al
Instruments Used Electrocautery
Particle Size (µm)
Composition
Particle Volume
0.07
Ultrasonic scalpels
0.35-6.5
Liquid aerosols
Carbajo-Rodríguez et al11
Electrocautery Ultrasonic scalpel
0.07-25 n.s.
Weld et al3
Bipolar macroforceps Harmonic scalpel Floating ball Monopolar shears Electrocautery Ultrasonic scalpel
95% water and 5% chemical products: sodium, magnesium, calcium, iron, carbon, oxygen aldehyde, toluene, ethylbenzene, xylene, carbon monoxide n.s.
Barrett and Garber15
Brüske-Hohlfeld et al17
DesCôteaux et al12
Electrocautery
Schneider et al16
Different types of ultrasound shears
0.07-0.89 0.07-0.10 0.07-1.08 0.10-0.92 0.07-0.31 0.35-6.5
0.1-0.5 2-24 n.s.
of 95% water vapour and 5% chemical compounds/cell remnants. Evaporation of liquids creates the group of small homogeneous spheres, composed mainly of water and small amounts of sodium, potassium, magnesium, calcium, and iron. In contrast, low-temperature vaporization contains tissue fragments, which can also carry infectious or viable cells and present the latter population.10 According to Schneider et al,16 different ultrasonic shears induce varieties in mist generation. The authors demonstrated that the amount of vapor production depends on the technical design of ultrasonic shears. Brüske-Hohlfeld et al17 performed particle measurements with a condensation particle counter in several operation rooms during different surgical procedures, revealing a particle diameter range of d = 10 nm to 1 µm. Weld et al3 used the same setup in a hermetic chamber utilizing an aerodynamic particle sizer (time-of-flight technique) and electrostatic classifier (electrical mobility principle) for the calculation of the mean particle size. In 1988, Vanderpool anf Rubow18 developed a method to
Hydrocarbons, acrylonitrile, fatty acids, phenols
Sodium, chloride, magnesium, calcium, potassium, carbon, oxygen n.s.
n.s. Four times more compared with dissection of similar amount of tissue by ultrasonic scalpel Fatty tissue generates 17-23 times more than lean tissue Influenced by the destroyed type of tissue Particle number concentration in an condensation particle counter Depending on surgical procedure, tissue, intensity of energy; measurements in an condensation particle counter n.s. Measured with infrared light transmission
generate monodisperse aerosols in a diameter range of d = 1 to 20 µm and demonstrated that both the generation method and the physical/chemical aerosol characteristics have an influence on the particle size.
Experimental Setting Artificial Smoke and Mist Our purpose built test chamber allowed us to produce artificial smoke and mist close to real surgical conditions. The best results of the irritating effect of smoke plumes deteriorating the surgeon’s view of the operation field were achieved by application of the commercially available fog machine N10 (eurolite, Waldbüttelbrunn, Germany). The other tested smoke/fog generators, smoke pens, pills, and powders produced a nonrealistic smoke formation as the smoke was either too dense and intensive or the duration until the smoke disperses was too short. The most realistic smoke plumes (comparable to
Downloaded from sri.sagepub.com at University of Sydney on November 15, 2015
160
Surgical Innovation 22(2)
Figure 4. Mean particle diameter (smoke and mist in micrometers) subject to its generation (artificial, porcine, human). Electrocautery on human or porcine fat (left) shows comparable results with a mean diameter of dHFh = 4.65 µm and dHFp = 4.14 µm to the artificial generation with the fog machine (df = 4.19 µm). Fatty droplets of ultrasonic dissection are well emulated by a nebulizer (d = 46.71 µm). The comparison with human (dbF = 45.08 µm) and porcine (dbL = 45.45 µm) droplets on (right) side does not show any significant difference.
those of biological electrocautery) were achieved with a smoke inflow of 3 and 5 seconds (smoke inflow 1 second, smoke too light; 7 seconds, smoke too dense). Additionally, the inflow tube positioned on the side of the test chamber yielded a more convenient smoke distribution (steady dispersal) than the tube positioned on the top of the box (deepening dispersal). The smoke retention time in the test chamber was 14 to 15 minutes. The microscopic examination of the particle sizes generated by the fog machine revealed comparable results to those of electrocautery aerosols with a mean diameter of df = 4.19 µm. For the artificial generation of fatty droplets, 2 jet nebulizers were tested with 3 different fluids. The vaporizer Turboboy SX (PariTec, Starnberg, Germany) created too small droplets (dn1 = 10.65 µm) and was therefore excluded. Saline solution and purified water dried too fast that only downsized or even no droplets were visible under the microscope. The best results were achieved by utilizing the Nebulizer Sole N (PariTec, Starnberg, Germany) with vegetable oil generating particles with a mean diameter range of dn2 = 45.38 ± 6.16 to 48.04 ± 8.50 µm.
Biological Smoke and Mist The artificially generated particles were compared to droplets produced by HF electrocautery and ultrasonic dissection applied on lard, porcine and human visceral fat (Figure 4). The mean droplet diameter produced by the HF device Erbotom T400C (Erbe, Tübingen, Germany) was dHFh = 4.65 µm for human fat and dHFp = 4.14 µm for porcine fat.
Figure 5. Particle diameter during ultrasonic dissection depends on body mass index (BMI) and age of the patients (increase of particle size with increasing BMI).
For ultrasound dissection, a mean diameter size of dUSl = 45.45 ± 12.33 µm (lard) and dUSh = 30.80 ± 4.01 to 56.27 ± 17.01 µm (human fat) was noted (Figure 4). Between the different patients, a high droplet size variation was observed with a tendency that the droplet diameter increases with BMI and age (Figure 5).
Discussion The improvement of visualization is an essential precondition for the future development of minimally invasive interventions and further reduction of the surgical trauma by novel intervention techniques. Several research groups are working on miniaturized cameras to make Monoport or natural orifice translumenal endoscopic surgery (NOTES) mature for routine clinical application.19,20 New developments in medical imaging techniques now offer the chance to overcome existing obstacles such as complex and defective intra-abdominal camera fixation. A promising technology is the use of HDR imaging. HDR cameras grab 2 or more images simultaneously with different exposure levels and combine these to 1 image with optimal brightness. Thus, small optics placed almost anywhere on the laparoscopic instrument or solitarily introduced through tiny ports can be used for advanced visualization of the abdominal cavity. Additionally, HDR video chips offer the opportunity to digitally reduce interfering effects like electrocauterization smoke or mist from ultrasound dissection by computing algorithms in the visualization chain. However, for a sufficient reduction of disturbing particles by computer algorithms directly on the video processor, detailed knowledge about the particle size and retention time inside is essential. We therefore developed
Downloaded from sri.sagepub.com at University of Sydney on November 15, 2015
161
Falkinger et al an experimental test bed for particle size and composition analysis that can be used with both organic and artificially generated particle samples to facilitate testing of different camera systems and developing digital image correction algorithms in a standardized and reproducible setup. Within a specially developed test chamber, artificial smoke was produced by a fog machine revealing almost identical particle diameters (df = 4.19 µm) as compared with smoke generated by surgical electrocautery (dHFh = 4.65 µm [human fat] and dHFp = 4.14 µm [porcine fat]). Mist, which is typically evolving during ultrasound tissue dissection, was generated by a nebulizer. Artificial droplet size ranged between dn2 = 45.38 and 48.04 µm and biological droplet size between dUSh = 30.80 and 56.27 µm. A tendency of enlarged particle diameter with increased BMI and age was noted. Our results are in line with those published by other research groups.11,12 For generation of artificial smoke, droplet diameters can vary due to the composition of fog fluid, which is typically based on distilled water and propylene glycol.18 The fog retention time depends on additives for stabilisation like glycerol. In our experimental setting, the smoke was evenly spread after 20 seconds in the test chamber and remained for about 15 minutes until dissolving. For development of smoke reduction algorithms, especially the first 20 seconds of smoke spreading are important, as video sequences of shifting smoke plumes are needed for digital image processing. In addition to simulating and quantitative measuring surgical smoke and mist, the presented test bed was also partly used for developing image processing algorithms and to evaluate first applications of image improvements of HDR technology. Therefore, reflections on abdominal structures and over- or underexposed (shady) images needed to be generated. For this, the shutter of a fixed laparoscopic camera used in our test chamber was varied from 1.28 to 1/10 000 second and consecutive sequences of standardized test images of the USC-SIPI (University of Southern California Signal and Image Processing Institute) Image Database21 were recorded. Thereby, we achieved the generation of identical deteriorated images in different exposure times as the basis for future testing and optimizing of a first prototypic HDR camera system. In conclusion, we developed a suitable test bed for artificial smoke and mist generation revealing almost identical droplet characteristics as produced during minimally invasive procedures (electrocautery and ultrasound dissection). The possibility to generate image interferences comparable to those occurring during laparoscopy provides a basis for the future development of image processing technologies for clinical applications.
Declaration of Conflicting Interests The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported in part by AiF Projekt GmbH.
References 1. Fitzgerald JE, Malik M, Ahmed I. A single-blind controlled study of electrocautery and ultrasonic scalpel smoke plumes in laparoscopic surgery. Surg Endosc. 2012;26:337-342. 2. Gossot D, Buess G, Cuschieri A, et al. Ultrasonic dissection for endoscopic surgery. Surg Endosc. 1999;13: 412-417. 3. Weld KJ, Dryer S, Ames CD, et al. Analysis of surgical smoke produced by various energy-based instruments and effect on laparoscopic visibility. J Endourol. 2007;21: 347-351. 4. Ansell J, Warren N, Sibbons P, Torkington J. The innervision surgical smoke removal system. Int J Surg. 2012;10(8):S43. 5. Pece F, Kautz J, eds. Bitmap Movement Detection: HDR for Dynamic Scenes. Proceedings of the IEEE Conference on Visual Media Production (CVMP) 2010. Piscataway, NJ: IEEE. 6. Debevec PE, Malik J, eds. Recovering High Dynamic Range Radiance Maps From Photographs. ACM SIGGRAPH 2008 Classes. New York, NY: ACM; 2008. 7. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. Ann Intern Med. 2009;151:W-65-W-94. 8. Wu Y-C, Tang C-S, Huang H-Y, et al. Chemical production in electrocautery smoke by a novel predictive model. Eur Surg Res. 2011;46:102-107. 9. Al Sahaf O, Vega-Carrascal I, Cunningham F, McGrath J, Bloomfield F. Chemical composition of smoke produced by high-frequency electrosurgery. Irish J Med Sci. 2007;176:229-232. 10. Alp E, Bijl D, Bleichrodt R, Hansson B, Voss A. Surgical smoke and infection control. J Hosp Infect. 2006;62:1-5. 11. Carbajo-Rodríguez H, Aguayo-Albasini JL, Soria-Aledo V, García-López C. Surgical smoke: risks and preventive measures. Cir Esp (Engl Ed). 2009;85:274-279. 12. DesCôteaux J-G, Picard P, Poulin E, Baril M. Preliminary study of electrocautery smoke particles produced in vitro and during laparoscopic procedures. Surg Endosc. 1996;10:152-158. 13. Hensman C, Baty D, Willis R, Cuschieri A. Chemical composition of smoke produced by high-frequency electrosurgery in a closed gaseous environment. Surg Endosc. 1998;12:1017-1019. 14. Ott DE, Moss E, Martinez K. Aerosol exposure from an ultrasonically activated (harmonic) device. J Am Assoc Gynecol Laparosc. 1998;5:29-32.
Downloaded from sri.sagepub.com at University of Sydney on November 15, 2015
162
Surgical Innovation 22(2)
15. Barrett WL, Garber SM. Surgical smoke: a review of the literature. Surg Endosc. 2003;17:979-987. 16. Schneider A, Doundoulakis E, Can S, Fiolka A, Wilhelm D, Feußner H. Evaluation of mist production and tissue dissection efficiency using different types of ultrasound shears. Surg Endosc. 2009;23:2822-2826. 17. Brüske-Hohlfeld I, Preissler G, Jauch K-W, et al. Surgical smoke and ultrafine particles. J Occup Med Toxicol. 2008;3:31. 18. Vanderpool R, Rubow K. Generation of large, solid, monodisperse calibration aerosols. Aerosol Sci Technol. 1988;9:65-69.
19. Abu Gazala M, Shussman N, Abu Gazala S, et al. Miniature camera for enhanced visualization for single-port surgery and NOTES. J Laparoendosc Adv Surg Tech. 2012;22: 984-988. 20. Chang VC, Tang S-J, Swain CP, et al. A randomized comparison of laparoscopic, flexible endoscopic, and wired and wireless magnetic cameras on ex vivo and in vivo NOTES surgical performance. Surg Innov. 2013;20:395-402. 21. Weber A. The USC-SIPI image database. Signal and Image Processing Institute of the University of Southern California. http://sipi.usc.edu/services/database. Accessed May 19, 2014.
Downloaded from sri.sagepub.com at University of Sydney on November 15, 2015
