Research and development of a new RF-assisted device for bloodless rapid transection of the liver: Computational modeling and in vivo experiments
© Burdío et al; licensee BioMed Central Ltd. 2009
Received: 03 November 2008
Accepted: 18 March 2009
Published: 18 March 2009
Efficient and safe transection of biological tissue in liver surgery is strongly dependent on the ability to address both parenchymal division and hemostasis simultaneously. In addition to the conventional clamp crushing or finger fracture methods other techniques based on radiofrequency (RF) currents have been extensively employed to reduce intraoperative blood loss. In this paper we present our broad research plan for a new RF-assisted device for bloodless, rapid resection of the liver.
Our research plan includes computer modeling and in vivo studies. Computer modeling was based on the Finite Element Method (FEM) and allowed us to estimate the distribution of electrical power deposited in the tissue, along with assessing the effect of the characteristics of the device on the temperature profiles. Studies based on in vivo pig liver models provided a comparison of the performance of the new device with other techniques (saline-linked technology) currently employed in clinical practice. Finally, the plan includes a pilot clinical trial, in which both the new device and the accessory equipment are seen to comply with all safety requirements.
The FEM results showed a high electrical gradient around the tip of the blade, responsible for the maximal increase of temperature at that point, where temperature reached 100°C in only 3.85 s. Other hot points with lower temperatures were located at the proximal edge of the device. Additional simulations with an electrically insulated blade produced more uniform and larger lesions (assessed as the 55°C isotherm) than the electrically conducting blade. The in vivo study, in turn, showed greater transection speed (3 ± 0 and 3 ± 1 cm2/min for the new device in the open and laparoscopic approaches respectively) and also lower blood loss (70 ± 74 and 26 ± 34 mL) during transection of the liver, as compared to saline-linked technology (2 ± 1 cm2/min with P = 0.002, and 527 ± 273 mL with P = 0.001).
A new RF-assisted device for bloodless, rapid liver resection was designed, built and tested. The results demonstrate the potential advantages of this device over others currently employed.
The history of the development of surgical liver resection techniques is largely that of the struggle against hemorrhaging. Before the 1980s, hepatic resection was associated with a mortality rate of 10–20%, mostly due to hemorrhage during liver transection. Nowadays, the hospital mortality rate of liver resection is 5% or lower in most specialized centers . Intraoperative blood loss and perioperative transfusion not only increase the risk of operative morbidity and mortality  but also jeopardize long-term survival, since they actually increase the recurrence rate of the tumor being resected [3–5]. As a result, many techniques have been developed over the last ten years to minimize intraoperative blood loss during this type of operation. In addition to the conventional finger fracture, other techniques such as the ultrasonic dissector , water jet dissector [7, 8], argon beam coagulator  and saline-linked radiofrequency (RF) technology [10, 11] have been intensively employed to reduce intraoperative blood loss.
Several recent approaches, including saline-linked technology, employ RF energy deposited in the liver tissue to achieve a coagulating effect prior to transection. Weber et al  pioneered the use of RF needle electrodes to obtain a 1 or 2 cm wide coagulation band on the resection line before employing the scalpel, thereby facilitating bloodless liver resection. Further applications of the concept of pre-coagulating tissue with RF energy prior to transection have recently been developed with devices like Habib4x  or InLine [14–18], in which a range of RF electrodes are inserted in a monopolar or bipolar configuration. Once the plane of coagulation has been created with these RF-assisted devices, the coagulated tissue is cut bloodlessly with a straight scalpel.
The aim of the present study is to present our mutidisciplinary research, including both bioengineering and experimental approaches, to test the efficiency and safety of the new RF-assisted device.
Description of the RF-assisted device
where ρ, c and k are respectively the density, specific heat and thermal conductivity of the tissue. The term Q p corresponds with the heat loss caused by blood perfusion. Our study modeled an ex vivo transection, and hence we did not consider this term. Finally, the term q is the heat source caused by RF power (Joule loss) which is given by:q = J·E
where J is the current density (A/m2) and E is the electric field intensity (V/m). The values of these two vectors are evaluated using Laplace's equation:∇·σ∇Φ = 0
where Φ is the voltage (V) and σ is the electrical conductivity (S/m).
Characteristics of the elements employed in the computer modeling.
Electrode and conducting blade
7.4 × 106
8 × 103
Insulated handled and insulated blade
where F is the flow of the circulating fluid (0.13 L/min) and A t is the cross-sectional area of the tube = π·(0.002)2 = 1.2566 × 10-5 m2. Flow velocity was therefore ≈0.172 m/s.
The characteristics of the circulating fluid were considered to be those of water at 37°C : k f = 0.63 W/m·K, ρ f = 999.4 Kg/m3, μ = 6.9 × 10-4 N·s/m, and c f = 4174 J/kg·K. We obtained a Reynolds number Re = 3737, and a Prandtl number of Pr = 4.57. As a result, ≈ 67, and h i ≈ 2800 W/m2K. The coolant temperature was considered to be 5°C.
As in every FEM problem, we had to define the boundary conditions. Figure 4 shows the thermal (blue) and electrical (red) boundary conditions. The temperature for surfaces at a distance from the device was assumed to be 36°C (this was also the value for initial temperature). A thermal condition of null thermal flux was used at the symmetry plane. The effect of free heat convection in the tissue-ambient and device-ambient interfaces was taken into account using a thermal transfer coefficient (h e ) of value 20 W/m2K. A value of 21°C was considered for the ambient temperature.
Regarding the electrical boundary conditions, computer simulations were conducted using a constant electrical voltage of 50 V between the metallic part of the device and dispersive electrode, which was assumed to be on the bottom surface (see Figure 4). The electrical voltage on the dispersive electrode was fixed at zero volts. An electrical condition of null electrical current was used at the symmetry plane. The same condition was used on the surfaces at a distance from the device (except for the bottom surface), and on the tissue-ambient and device-ambient interfaces.
Physical dimensions of the applicator correspond with those of the actual device. However, the tissue dimensions (A and B in Figure 4) and handle length (H in Figure 4) were calculated by means of a sensitivity analysis in order to avoid boundary effects. A convergence test was performed to obtain the adequate spatial and temporal resolution . The value of the maximal temperature achieved in the hepatic tissue (Tmax) after 1 s of heating was used as a control parameter in these sensitivity and convergence tests. The spatial resolution was heterogeneous. We used the following criterion: the finest zone was always the blade-tissue interface because it is known that this has the largest voltage gradient and hence the maximum value of current density. In the tissue, grid size was increased gradually with distance from the blade-tissue interface. First, we considered a tentative spatial and temporal resolution. We then conducted a computer analysis to determine the appropriate values of A, B and H (see Figure 4). These simulations were made by increasing the value of the three parameters by equal amounts. When the difference between the Tmax and the Tmax in the previous simulation was less than 0.5°C, we considered the former values to be adequate. Finally, we performed convergence tests to determine adequate spatial and temporal resolution. The spatial resolution was achieved by refining the mesh near the blade so that Tmax was within 0.5% of the value obtained from the previous refinement step . With an adequate spatial resolution achieved, we decreased the time step until Tmax was within 0.5% of the value obtained from the last time step.
In this study, the simulations were stopped when maximal temperature in the tissue reached 100°C . This was because no experimental data have been reported dealing with the thermal and electrical characteristics of hepatic tissue above this temperature. We then analyzed the voltage and temperature distributions in the tissue using the 55°C isotherm as thermal lesion boundary. The model was employed to assess the electrical-thermal performance of the proposed device, i.e. to determine the location of the hottest points. It should be pointed out that at the hottest points, i.e. in the tissue zones where temperature reaches 100°C, dehydration and carbonization processes occur. These phenomena can drastically limit the progress of the lesion, and hence the expansion of the coagulated area. Furthermore, carbonization can cause carbonized tissue to adhere to the electrode, which could affect the mechanical characteristics of some of its critical parts, such as the blade. In fact, in view of the results of the first simulation, we modified the model to study the effect on the electrical characteristics of the blade. To be more precise, we performed a further computer simulation considering a blade made of the same non conducting material as the handle (see Table 1).
In vivo study on pig liver
Sixteen partial hepatectomies were performed on pig liver with the proposed device by both laparotomy (n = 8) and a laparoscopic hand-assisted approach (n = 8). Eight similar partial hepatectomies were also performed using a conventional saline-linked device through laparotomy (DS 3.0; Tissue-Link Medical, Dover, New Hampshire, USA) acting as a control group (used in accordance with the manufacturer's recommendations). A total of 12 female farm pigs were employed in the study. The animals were deeply anesthetized with Tiletamine-Zolazepam (7 mg/kg, im), medetomidine (0.03 mg/kg) and maintained with propofol (10 mg/kg) or sevoflurane. In all cases, the specific efficiency of each medical device (proposed and the Tissue-link® device) was evaluated in division and hemostasis without the aid of any other instruments unless a hemorrhage of more than 2 min was observed. After performing the experiment the animals were euthanized by exanguination. All the experimental procedures were conducted in a laboratory authorized for animal research and approved by the local institutional ethical committee.
The main outcome measures of the study were: 1) Transection time: total time of transection including time for achieving complete hemostasis; 2) Blood loss: total amount of blood loss during transection (from the sucker and bloody gauzes minus saline dripped onto the tissue); 3) Transection area: obtained by delineation of the transection plane (digital photograph) using appropriate software (3D Doctor, Able Software Corp, Lexington, MA, USA); 4) Transection speed: ratio of the transection area to transection time, 5) Blood loss per transection area and 6) Tissue coagulation depth: mean tissue coagulation depth calculated by histological assessment from four points equidistant from the mid transverse line of the transection area in every case. The hemorrhagic rim surrounding ablation tissue was not included in coagulation depth measurements.
The statistical analyses were performed on mean values in the open approach (proposed device vs. Tissue-link® method). Mean values of numerical data were compared for both groups using the Student t test or the U-Mann-Whitney's test when appropriate. Categorical data were compared for both groups using Fisher's exact text. Differences in variables were considered to be significant at a threshold of P < 0.05. Statistical analyses were performed with statistical software (version 12.0; SPSS, Chicago, IL, USA).
Technical modifications for clinical trials
After the sensitivity analyses, we obtained the following dimensions: A = 30 mm, B = 45 mm and H = 10 mm. The convergence test provided a grid size of 0.1 mm in the finest zone (blade-tissue interface), and a step time of 0.05 s. The model had nearly 14,800 nodes and used over 66,000 tetrahedral elements (see C and D in Figure 5).
A comparison of the results provided by the two blades showed that the electrically insulated blade took longer to reach 100°C tissue temperature (10 s) than the electrically conducting blade (3.85 s). As a result, the lesion boundary (assessed as the 55°C isotherm) was more uniform and larger in the case of the electrically insulated blade (see D in Figure 7).
Performance of the proposed device on in vivo pig liver
Additional file 1: Transection of the liver with a hand-assisted laparoscopic approach with the proposed device. Note that no other device is used to achieve dissection, parenchyma division and hemostasis. (MOV 1 MB)
Results of the in vivo studies comparing the new device with Tissue-Link®.
New device under test
Transection time (min)
12 ± 3
13 ± 7
21 ± 7
Blood loss (mL)
70 ± 74
26 ± 34
527 ± 273
Transection area (cm2)
35 ± 7
34 ± 11
41 ± 8
Transection speed (cm2/min)
3 ± 0
3 ± 1
2 ± 1
Blood loss per transection area (mL/cm2)
2 ± 2
1 ± 1
13 ± 6
Tissue coagulation depth (mm)
6 ± 2
9 ± 2
3 ± 1
This paper describes the different steps followed by our group in the R&D of a new RF-assisted device for rapid, bloodless liver resection. We first employed the computer modeling technique to predict and assess the temperature profile of the new device. This approach has been widely employed in surgical research, not only in RF heating of biological tissue , but also to assist the design of endoscopic instruments , to assess the impact of surgical procedures  and to evaluate the performance of different surgical devices . The advantages of this approach are its low cost and the short time needed to determine basic performance and the effect of different parameters (e.g. dispersion in the geometry and physical characteristics of the tissues, materials and designs of surgical devices). The main disadvantage is obviously the lack of accurate characterization of the biological tissue (i.e. full data on thermal, electrical and mechanical properties). In our study, the results of the computer modeling showed a marked edge effect at the blade tip (see B in Figure 7), which has also been observed in long electrodes in surface application for cardiac ablation . This overheating at the tip could produce rapid tissue carbonization adhering to the blade, and hence could impair the cutting quality. For this reason, we conducted additional simulations with a non (electrically) conducting blade and the results showed a considerable reduction of the edge effect. Maximal temperature was not located at the blade tip, avoiding the risk of carbonization in this zone, and hence preserving cutting quality. In view of these results, future studies should be conducted using prototypes with electrically insulated blades. However, it may be difficult to find a high-performance non electrically conducting material appropriate for surgical blades.
Regarding the research on a surgical device for bloodless hepatic resection, it is very important to point out that ex vivo setups do not allow modeling of blood perfusion, and it will therefore be necessary to plan further in vivo experiments. We did, in fact, conduct in vivo experiments on the pig model to assess the true potential of the new device [19, 20], as other authors had done previously [22, 30]. In addition, our in vivo studies were always planned as comparative assessments. We paid particular attention to comparing our device with the Tissue-Link® device (DS 3.0; Tissue Link Medical, Dover, New Hampshire, USA). This is a conventional saline-linked system which acted as the control group in our experiments. It employs similar technology to the proposed device and shows promise but is still under evaluation [11, 31]. The performance of our device was demonstrated in the open comparative approach, especially in the 30% increase in liver transection speed (over the Tissue-Link® device) (see Table 2). Even more important, our device demonstrated a seven-fold reduction in blood loss during liver transection (over the Tissue-Link® device) (see Table 2). With the proposed method a greater mean coagulation depth was obtained compared to the saline-linked device (6 ± 2 mm and 3 ± 1 mm, for the new device and Tissue-link device, respectively). This was the only macroscopic or microscopic difference found between transection surfaces in both groups.
The laparoscopic study with our device on the other hand, demonstrated similar or even better results than in the open approach with our device . Hence, our device may be used with a minimally invasive approach in a performance fashion.
Even though the device gave good results in both open and laparoscopic approaches in the tests on animals, further clinical research, especially through a Randomized Controlled Trial -RCT- will be necessary before the system can be introduced into clinical practice . At the moment we are in the process of designing a two-phase clinical testing program. Phase I study will examine the safety of the procedure with a small sample of patients suffering from liver metastases. When and if the safety of the procedure is guaranteed, Phase II study will evaluate the new device in comparison to a conventional method of liver transection with randomly selected patients.
Limitations of the study
The method studied has been compared with the saline-linked dissecting sealer, which is a promising technology but is still under evaluation. It will therefore be necessary for future research by other groups to compare the studied method with other available technologies, such as CUSA or Kelly crashing.
The method sacrifices parenchymal tissue that is usually spared in other resectional techniques. This could be especially relevant in cirrhotic patients with limited remnant reserve.
In the study of a pig liver model, we tested the specific efficiency of each procedure in division and hemostasis without the aid of other instruments. It is conceivable that in an actual clinical setting (combining these methods with other systems of division and hemostasis when necessary) differences in transection time and transection speed between both methods may be smaller.
In this paper we have presented our research and development plan for a new RF-assisted device for rapid, bloodless liver resection, covering a variety of different techniques ranging from computer modeling to clinical trials. The interaction of these multidisciplinary methodologies is expected to provide further data on the new device.
This work was partially supported by a medical research grant from the Spanish Government (PETRI 2005/0353), from the "Programa de Promoción de la Investigación Biomédica y en Ciencias de la Salud del Ministerio de Sanidad y Consumo" of Spain (PI052498), and by the "Spanish Plan Nacional de I+D+I del Ministerio de Ciencia e Innovación" (TEC2008-01369/TEC). Finally, the publication cost were provided by the "Programa de Apoyo a la Investigación y el Desarrollo" of the Universidad Politécnica de Valencia. We would like to thank the R+D+i Linguistic Assistance Office at the Universidad Politécnica of Valencia for their help in revising this paper.
- Poon RT: Current techniques of liver transection. HBP (Oxford) 2007,9(3):166–173.View ArticleGoogle Scholar
- Sitzmann JV, Greene PS: Perioperative predictors of morbidity following hepatic resection for neoplasm. A multivariate analysis of a single surgeon with 105 patients. Ann Surg 1994, 219: 13–17.View ArticleGoogle Scholar
- Stephenson K, Steinberg S, Hughes KS, Vetto JT, Sugarbaker PH, Chang AE: Perioperative blood transfusions are associated with decreased time to recurrence and decreased survival after resection of colorectal liver metastases. Ann Surg 1988, 208: 679–687. 10.1097/00000658-198812000-00002View ArticleGoogle Scholar
- Kooby DA, Stockman J, Ben-Porat L, Gonen M, Jarnagin WR, Dematteo RP, Tuorto S, Wuest D, Blumgart LH, Fong Y: Influence of transfusions on perioperative and long-term outcome in patients following hepatic resection for colorectal metastases. Ann Surg 2003, 237: 860–869. 10.1097/00000658-200306000-00015Google Scholar
- Yamamoto J, Kosuge T, Takayama T, Shimada K, Yamasaki S, Ozaki H, Yamaguchi N, Mizuno S, Makuuchi M: Perioperative blood transfusion promotes recurrence of hepatocellular carcinoma after hepatectomy. Surgery 1994, 115: 303–309.Google Scholar
- Takayama T, Makuuchi M, Kubota K, Harihara Y, Hui AM, Sano K, Ijichi M, Hasegawa K: Randomized comparison of ultrasonic vs clamp transection of the liver. Arch Surg 2001, 136: 922–928. 10.1001/archsurg.136.8.922View ArticleGoogle Scholar
- Baer HU, Stain SC, Guastella T, Maddern GJ, Blumgart LH: Hepatic resection using water jet dissector. HBP Surg 1993, 6: 189–196.Google Scholar
- Rau HG, Buttler ER, Baretton G, Schardey HM, Schildberg FW: Jet-cutting supported by high frequency current: New technique for hepatic surgery. World J Surg 1997, 21: 254–259. 10.1007/s002689900225View ArticleGoogle Scholar
- Postema RR, Plaisier PW, Ten Kate FJ, Terpstra OT: Haemostasis after partial hepatectomy using argon beam coagulation. Br J Surg 1993, 80: 1563–1565. 10.1002/bjs.1800801224View ArticleGoogle Scholar
- Poon RT, Fan ST, Wong J: Liver resection using a saline-linked radiofrequency dissecting sealer for transection of the liver. J Am Coll Surg 2005, 200: 308–313. 10.1016/j.jamcollsurg.2004.10.008View ArticleGoogle Scholar
- Arita J, Hasegawa K, Kokudo N, Sano K, Sugawara Y, Makuuchi M: Randomized clinical trial of the effect of a saline-linked radiofrequency coagulator on blood loss during hepatic resection. Br J Surg 2005, 92: 954–959. 10.1002/bjs.5108View ArticleGoogle Scholar
- Weber JC, Navarra G, Jiao LR, Nicholls JP, Jensen SL, Habib NA: New technique for liver resection using heat coagulative necrosis. Ann Surg 2002, 236: 560–563. 10.1097/00000658-200211000-00004View ArticleGoogle Scholar
- Hering J, Garrean S, Saied A, Helton WS, Espat NJ: Use of radiofrequency hepatic parenchymal transection device in hepatic hemangioma resection: early experience and lessons learned. HBP (Oxford) 2007,9(4):319–323.View ArticleGoogle Scholar
- Haghighi KS, Wang F, King J, Daniel S, Morris DL: In-line radiofrequency ablation to minimize blood loss in hepatic parenchymal transection. Am J Surg 2005, 190: 43–47. 10.1016/j.amjsurg.2005.04.011View ArticleGoogle Scholar
- Clancy TE, Swanson RS: Laparoscopic radiofrequency-assisted liver resection (LRR): a report of two cases. Dig Dis Sci 2005, 50: 2259–2262. 10.1007/s10620-005-3045-3View ArticleGoogle Scholar
- Haghighi KS, Steinke K, Hazratwala K, Kam PC, Daniel S, Morris DL: Controlled study of inline radiofrequency ablation (ILRFA) assisted transection of ovine liver. J Surg Res 2005, 123: 139–143. 10.1016/j.jss.2004.07.021View ArticleGoogle Scholar
- Yao P, Morris DL: Radiofrequency ablation-assisted liver resection: review of the literature and our experience. HBP (Oxford) 2006,8(4):248–254.View ArticleGoogle Scholar
- Yao P, Chu F, Daniel S, Gunasegaram A, Yan T, Lindemann W, Pistorius G, Schilling M, Machi J, Zuckerman R, Morris DL: A multicentre controlled study of the inline radiofrequency ablation device for liver transection. HBP (Oxford) 2007,9(4):267–271.View ArticleGoogle Scholar
- Burdío F, Navarro A, Berjano E, Sousa R, Burdío JM, Güemes A, Subiró J, Gonzalez A, Cruz I, Castiella T, Tejero E, Lozano R, Grande L, de Gregorio MA: A radiofrequency-assisted device for bloodless rapid transection of the liver: A comparative study in a pig liver model. Eur J Surg Oncol 2007, 34: 599–605.View ArticleGoogle Scholar
- Navarro A, Burdio F, Berjano EJ, Güemes A, Sousa R, Rufas M, Subirá J, Gonzalez A, Burdío JM, Castiella T, Tejero E, De Gregorio MA, Grande L, Lozano R: Laparoscopic blood-saving liver resection using a new radiofrequency-assisted device: preliminary report of an in vivo study with pig liver. Surg Endosc 2008, 22: 1384–1391. 10.1007/s00464-008-9793-3View ArticleGoogle Scholar
- Berjano EJ: Theoretical modeling for radiofrequency ablation: state-of-the-art and challenges for the future. Biomed Eng Online 2006, 5: 24. 10.1186/1475-925X-5-24View ArticleGoogle Scholar
- Haemmerich D, Schutt DJ, Will JA, Striegel RM, Webster JG, Mahvi DM: A device for radiofrequency assisted hepatic resection. In Proceedings of the 26th Annual International Conference of the Engineering in Medicine and Biology Society – IEEE: 1–5 September 2004; San Francisco. Volume 4. IEEE-Press; 2004:2503–2506.View ArticleGoogle Scholar
- Holman JP: Heat transfer. 9th edition. New York: McGraw-Hill; 2001.Google Scholar
- Berjano EJ, Hornero F: Thermal-electrical modeling for epicardial atrial radiofrequency ablation. IEEE Trans Biomed Eng 2004, 51: 1348–1357. 10.1109/TBME.2004.827545View ArticleGoogle Scholar
- Berjano EJ, Alió JL, Saiz J: Modeling for radio-frequency conductive keratoplasty: implications for the maximum temperature reached in the cornea. Physiol Meas 2005, 26: 157–172. 10.1088/0967-3334/26/3/002View ArticleGoogle Scholar
- Jernigan SR, Buckner GD, Eischen JW, Cormier DR: Finite element modeling of the left atrium to facilitate the design of an endoscopic atrial retractor. J Biomech Eng 2007, 129: 825–837. 10.1115/1.2801650View ArticleGoogle Scholar
- Hato T, Kawahara N, Tomita K, Murakami H, Akamaru T, Tawara D, Sakamoto J, Oda J, Tanaka S: Finite-element analysis on closing-opening correction osteotomy for angular kyphosis of osteoporotic vertebral fractures. J Orthop Sci 2007, 12: 354–360. 10.1007/s00776-007-1144-zView ArticleGoogle Scholar
- Korkmaz HH: Evaluation of different miniplates in fixation of fractured human mandible with the finite element method. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2007, 103: e1–13. 10.1016/j.tripleo.2006.12.016View ArticleGoogle Scholar
- Berjano EJ, Hornero F, Atienza F, Montero A: Long electrodes for radio frequency ablation: comparative study of surface versus intramural application. Med Eng Phys 2003, 25: 869–877. 10.1016/S1350-4533(03)00125-5View ArticleGoogle Scholar
- Rossi P, De Majo A, Mauti A, Mauti P, Quattrini V, Mattei M, Tognoni V, Cenci L, Manzelli A, Lorenzo ND, Gaspari AL: Bloodless hepatic resection with automatic bipolar radiofrequency generator and multielectrode device. Minim Invasive Ther Allied Technol 2007, 16: 66–72. 10.1080/13645700701191511View ArticleGoogle Scholar
- Lesurtel M, Selzner M, Petrowsky H, McCormack L, Clavien PA: How Should transection of the liver be performed? A prospective randomized study in 100 consecutive patients: comparing four different transection strategies. Ann Surg 2005, 242: 814–823. 10.1097/01.sla.0000189121.35617.d7View ArticleGoogle Scholar
- Jadad A: Randomized controlled trials. London: British Medical Journal Books; 1998.Google Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.