• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of annsurgLink to Publisher's site
Ann Surg. Feb 2003; 237(2): 186–191.
PMCID: PMC1522135

How Safe is High-Power Ultrasonic Dissection?

Tarek A. Emam, MB BCh, MCh and Alfred Cuschieri, MD, ChM, FRSE, FRCS



To evaluate the safety of ultrasonic dissection.

Summary Background Data

High-power ultrasonic dissection is in widespread use for both open and laparoscopic operations and is generally perceived to carry a low risk of collateral damage, but there is no published evidence for this.


Under controlled experimental conditions, ultrasonic dissections were performed in pigs using Ultracision (Ethicon) or Autosonix (Tyco/USSC) at the three power settings (3, 4, and 5) in random fashion to mobilize the cardia and fundus, bile duct, hepatic artery, portal vein, aorta from the inferior vena cava, renal vessels, colon, and ureters. The dissections (open and laparoscopic) were carried out on pigs at each power setting with each device. Thermal mapping of the tissues during dissection was performed with an infrared thermal camera and associated software. The animals were killed at the end of each experiment and specimens were harvested for quantitative histology.


Extreme and equivalent temperature gradients were generated by ultrasonic dissection with both systems. Heat production was directly proportional to the power setting and the activation time. The core body temperature of the animals after completion of the laparoscopic dissections rose by an average of 2.3°C. The zone around the jaws that exceeded 60°C with continuous ultrasonic dissection for 10 to 15 seconds at level 5 measured 25.3 and 25.7 mm for Ultracision and Autosonix, respectively. At this power setting and an activation time of 15 seconds, the temperature 1.0 cm away from the tips of the instrument exceeded 140°C. Although there was no discernible macroscopic damage, these thermal changes were accompanied by significant histologic injury that extended to the media of large vessels and caused partial- to full-thickness mural damage of the cardia, ureter, and bile duct. Collateral damage was absent or insignificant after dissections at power level 3 with both systems and an activation time not exceeding 5 seconds.


High-power ultrasonic dissections at level 5 and to a lesser extent level 4 result in considerable heat production that causes proximity collateral damage to adjacent tissues when the continuous activation time exceeds 10 seconds. Ultrasonic dissections near important structures should be conducted at level 3. At power levels of 4 and 5, the ultrasonic energy bursts to the tissue should not exceed 5 seconds at any one time.

High-power ultrasonic dissection systems that cut and coagulate tissues have been introduced in both open and endoscopic surgery. They carry undoubted advantages over high-frequency electrosurgery in that they do not generate smoke. It is generally assumed that ultrasonic dissection systems disperse less energy to surrounding tissue during activation and thus have a reduced propensity for collateral or proximity thermal damage. The two most widely used systems, Ultracision (Ethicon) and Autosonix (USSC, Tyco), incorporate piezoelectric transducers that induce a vibration frequency at the functional tip (through different transmission systems along the shaft) of 55 kHz over a 50- to 100-μm arc (movement of the tips). At this level, the ultrasonic energy can cut and coagulate, especially when the functional end consists of shears such that the tissue is compressed between the sharp and blunt blades. Thus, the benefit is that of a multifunctional instrument that can cut and achieve local hemostasis of vessels of up to 2 to 3 mm, ostensibly with minimal injury to surrounding tissues. 1–3 By reducing instrument traffic through the ports, these high-power ultrasonic dissection systems expedite considerably the conduct of complex laparoscopic operations, especially colorectal resections.

One reported in vivo study in pigs 4 demonstrated proximity injury detected by histologic examination to important structures such as the bile duct, aorta, and inferior vena cava during high-power ultrasonic dissection. In certain structures, the extent of the damage was marked: up to 80% of the thickness of the bile duct was coagulated in most areas, and there was 30% transmural necrosis/injury to the ureter, aorta, and inferior vena cava. Of even greater concern was the observation that this damage was not macroscopically apparent at the time of surgery but was apparent only on histologic examination of the structures harvested at the time of sacrifice of the animals. This study, however, did not give information on the power level (excursion of the vibrating tip) and the activation time used. These two variables determine the extent of frictional heat produced. The present in vivo study was designed to evaluate the extent of any collateral/proximity damage to important structures using both devices at power setting 3, 4, and 5 on a random basis. The design included real-time mapping of the heat production during use of the ultrasonic shears and the harvest of tissue specimens for quantitative histology for the detection of collateral/proximity damage to adjacent vital structures during ultrasonic dissection.



The high-energy ultrasonic dissection systems used were Ultracision (Ethicon-LCS-B5) with 5.0-mm shears and AutoSonics (USSC, Tyco) with 5.0-mm shears. Experiments were carried out in Large White postweaning young pigs (20–30 kg) randomly allocated to two groups. In group 1, laparoscopic dissections were carried out in four animals at each power setting (3, 4, and 5) for both devices (total of 24 animals). In group 2, open dissections were carried out with two animals at each power setting (3, 4, and 5) for both devices (total of 12 animals). These fixed power settings were necessary to obtain valid thermal isotherms during dissections with both types of ultrasonic dissectors, enabling valid comparisons between the two.


The animals were anesthetized using the following sedation, relaxation, and narcosis regimen: ketamine (Ketanest 10% with a dose of 20 mg/kg, intramuscular), xylazine (Rompun 2% in a dose of 2 mg/kg, intramuscular), atropine sulfate 1% in a dose of 3 mL/animal), and propofol (Disoprivan 1% in a dose of 2–5 mL/animal). Endotracheal anesthesia was with isoflurane (1–1.5 vol%), nitrous oxide (max. 75 vol%), and oxygen (25 vol%). The core body temperature was recorded using a gastric probe every 2 minutes during all laparoscopic and open dissections.


For the open cases, a midline laparotomy was made in the anesthetized pigs. Various structures were dissected using each of the three randomly selected power settings. Dissection was carried around the cardia with isolation of both vagi and division of the short gastric vessels as needed for a total fundoplication. The gastroepiploic vessels (3 mm in diameter) were also coagulated and cut with the ultrasonically activated shears using a single-grasp technique. The efficiency of coagulation following division was assessed by a grading system (grade 4, no bleeding; 3, slight ooze that stopped spontaneously; grade 2, ooze that needed reapplication of the shears; grade 1, active bleeding requiring clipping for control). Dissection was then carried around the bile duct to isolate it from the surrounding tissue and around the portal vein, which was separated from the pancreas. Next the cecum and colon were totally mobilized with division of the mesentery and the colon cut transversely as in colonic resection. Circumferential specimens were obtained to determine the extent of damage from the transection margin. The aorta was dissected from the inferior vena cava. The ureter was freed from the retroperitoneum by performing dissection parallel to the ureter along tissue planes in the upper half of its course.

For the laparoscopic cases, identical dissections were carried out laparoscopically using a positive-pressure pneumoperitoneum (10 mmHg). Core body temperature was recorded as in the open cases.

In all the experiments (open and laparoscopic), the activation time of the ultrasonic dissector was recorded by a stopwatch activated each time the surgeon pressed the foot pedal.

An infrared camera (AGEMA 900 Series, FLIR Systems AB, Danderyd, Sweden) was used to record precise heat tissue mapping during each of the open ultrasonic dissections; thermal mapping was impossible during the laparoscopic dissections.

Complete infrared thermal recordings were obtained of all the dissections for subsequent analysis by the infrared camera software. Real-time minimum, maximum, and different spot/zone temperatures were determined. Using the stored picture data, temperatures in each frame were measured and analyzed and saved. The time-discrete thermal changes and the heat diffusion range around the vibrating blade were calculated from the picture data for each open experiment.

Following completion of the dissections, the animals were killed with intravenous barbiturates. The following organs, all of which looked macroscopically normal, were harvested in formol saline for subsequent histology: esophagus, fundus of the stomach, bile duct and hepatic artery, portal vein, renal vessels, ureters, inferior vena cava and aorta, cecum and colon, splenic and hepatic biopsies. After processing, sections were cut from the harvested specimens and stained with hematoxylin and eosin and Elastic Van Gieson (for vessels). The histology studies were performed and reported by an independent histopathologist in a blind fashion. Each dissection was coded for pig number/tag, ultrasonic device, and power setting used.

All the experiments were carried out after approval was obtained from the animal investigational review board of the Institute in Hamburg, and a veterinary surgeon was present throughout all of the experiments.


Thermal Mapping

The temperature varied with both the activation time and power setting. For this reason, the data for the thermal tissue profiles were grouped into three continuous activation time zones: ultrasonic dissection up to 5 seconds, up to 10 seconds, and up to 15 seconds uninterrupted activation. The peak temperature was recorded at each of these time zones.

Table 1 shows that the mean peak temperature for activation of up to 5 seconds at level 3 was more than 60°C with both devices at the jaws but less than 60° at 1.0 cm from the vibrating end piece. However, at power settings 4 and 5, the mean peak temperature exceeded 60°C at the vibrating tips and at 1.0 cm from the shears. As shown in Tables 2 and 3, the mean peak temperatures were directly proportional to the activation time and power setting and exceeded 100°C at the tips of the shears at level 4 and 200°C when the activation time exceeded 10 seconds (Fig. 1).

figure 7FF1
Figure 1. Top: Thermal mapping during dissection with Ultracision at power setting 5 for 15 seconds: jaw temperature (A) = 224.3°C, tissue temperature at 10 mm from jaws (B) = 102°C, tissue temperature at 20 mm from jaws ...
Table thumbnail

The highest temperature reached at level 5 was 294°C for Ultracision and 297°C for Autosonix at activation times of 13 and 13.1 seconds, respectively. The instrument shaft distal to the functional tip heated up during activation to a similar extent with both systems. Thus, 3 cm from the jaws the shaft temperature was 63.55% (± 12.39) of the tip for Autosonix and 62.55% (± 13.88) for Ultracision.

The zone of hyperthermia above 60°C is a reflection of the heat production by the instrument and the diffusion/dissipation of this energy by the blood flow (heat sink effect). This zone or map defines the territory in which proximity or collateral thermal damage may occur (see Fig. 1). The width of this zone on activation for more than 10 seconds with both instruments averaged 25.28 mm (± 2.28) for Ultracision and 25.7 mm (± 3.08) for Autosonix.

At the end of the procedure, the core body core temperature of the pigs had risen by 2.32°C (± 0.53) for Autosonix and 2.25°C (± 0.41) for Ultracision during the laparoscopic dissections. The core body temperature rise during open dissections was 1.18°C (± 0.42) for Autosonix and 1.1°C (± 0.48) for Ultracision.

Histologic Studies

The wall thickness of the esophageal specimens averaged 4.3 mm. Similar mild adventitial damage only was observed with both dissection systems when the power output was set at level 3. Esophageal damage extending to the longitudinal muscle layer and involving 15% to 33% mural thickness of the organ was present in all dissections with both systems set at level 4 power output with activation time exceeding 10 seconds. The mural damage extended to the circular muscle layer/submucosa with dissections at power setting 5 and exceeding 10 seconds by both systems in all cases with 30% to 53% mural thickness involvement. There was no instance of mucosal damage.

The average wall thickness of the gastric specimens was 3.7 mm. In the stomach serosal coagulation damage was encountered following dissections with both systems at power setting 3, with the depth of coagulation averaging 1.4% of the wall thickness of the specimen. The damage caused by dissections at power setting 4 and activation time of more than 5 seconds with both systems extended to the muscle layer (muscularis propria) in all cases with 12% to 20% intramural involvement. Deeper damage reaching and involving the submucosal layer and vessels with 30% to 40% involvement of the wall thickness was encountered after dissections with the power set at 5 for longer than 10 seconds. As in the esophagus, there was no mucosal damage.

The colonic specimen thickness ranged from 1.7 to 2.2 mm. In the cecum/colon, power level 3 dissections resulted in equivalent serosal damage with both systems and a coagulation depth ranging from 3% to 8% of specimen wall thickness. At power level 4 dissections, the muscle layer was affected and the depth of the mural damage ranged from 12% to 54% of the wall thickness. Involvement of the submucosa with extent of damage ranging from 60% to 68% of the wall thickness was observed with power level 5 dissections of more than 10 seconds. Mucosal damage (transmural) was encountered in one case (Ultracision group). The extent of damage to the colonic wall from the line of transection ranged from 4.2 to 6 mm.

The wall thickness of the bile duct specimens averaged 0.5 mm. Partial-thickness damage of a depth ranging from 25% to 50% of the duct wall thickness was found after dissection at power level 3 exceeding 10 seconds. The mural damage was more extensive (depth of 50–75%) after power level 4 dissections and was full thickness after dissections at power level 5 of more than 10 seconds.

The thickness of the ureteric specimens examined histologically ranged from 0.6 to 0.8 mm. Partial-thickness damage of a depth of 25% to 50% was observed after power level 3 dissections exceeding 10 seconds, and damage extending to 50% to 55% of the wall thickness with gross muscle damage occurred after dissection at power level 4 for more than 10 seconds. At power level 5 dissections of this duration, two cases (one with each system) showed mucosal damage, with full-thickness muscle coat damage in the rest of the specimens.

The wall thickness of the aorta specimens ranged from 2.4 to 2.8 mm. Adventitial damage only was encountered at power level 3. Damage extending to the media and ranging between 18% to 27% of this coat was observed with power level 4 dissections and more extensive media damage (50–69%) with level 5 dissections of more than 10 seconds. There was no instance of intimal damage.

The wall thickness of the vena cava specimens ranged from 0.4 to 0.8 mm. The damage sustained was as follows: adventitia only with power setting 3, 50% to 75% of the wall thickness with power level 4 dissections, and full thickness (transmural) with level 5 dissections of more than 10 seconds.

The wall thickness of the portal vein specimens ranged from 0.2 to 0.5 mm. Adventitial damage only was seen after power level 3 dissections. This was deeper (up to 90% of the media) with level 4 dissections and 90% to total (transmural) after level 5 dissections for 10 to 15 seconds.

Efficiency of Coagulation During Division

The use of both instruments to coagulate and cut the gastroepiploic vessels resulted in good hemostasis after division with both instruments at power level 3 (grade 4). Ultracision coagulation/cutting at level 4 resulted in oozing that stopped spontaneously (grade 3) and at power level 5, persistent ooze that required reapplication (grade 2). Autosonix coagulation/division at power level 4 and 5 resulted in grade 2 hemostasis, except for three cases where persistent bleeding required clipping of the vessels (grade 1).


The development of dissecting systems based on high-power ultrasonic energy, especially for laparoscopic surgery, was instigated by the need for an alternative to high-frequency monopolar electrosurgery and to enable the use of multifunctional instruments (e.g., shears that can be used for mechanical dissection as well as energized cutting and coagulation). The perception has been that these high-power ultrasonic systems would improve the efficiency of laparoscopic dissections and at the same time reduce the morbidity from collateral/proximity iatrogenic injuries, well documented with high-frequency electrosurgery. 5–7 In addition, ultrasonic systems abolish smoke production, which aside from obscuring the view contains highly toxic and mutagenic polycyclic hydrocarbons produced by the high-temperature pyrolysis of fat and protein. 8

The energy produced by the high-power ultrasonic dissection systems (55 kHz) is frictional energy, proportional to both the vibration frequency and the displacement of the instrument tips. 1 In essence, therefore, these devices function because they generate heat from the mechanical friction between the tissue and the vibrating blade. 1 Proteins begin to denature when the temperature reaches 60°C. When heat is delivered relatively gradually, and provided the temperature remains below 100°C, proteins denature from the colloidal state into an insoluble gel, which is necessary for vessel coagulation. 7 In addition, the ultrasound energy induces a cavitational effect in water-containing tissues, and this facilitates tissue separation. The only physical dis-advantage is the “storm” effect (dust cloud caused by nonviable tissue particles sprayed by the electronic shears), which becomes problematic only with close-up work.

The results of the present study have demonstrated that the frictional heat produced during activation of both ultrasonic systems depends on the power setting and the activation time. The magnitude of this heat energy and its dissipation within the tissues is indicated by the increase in the core temperature of the animals at the end of the laparoscopic dissections by 2.3°C. At the maximal power setting of 5 (equivalent to a vibration arc of 100 μm), the heat generation is substantial not only at the vibrating tips but also, as the heat diffuses, within a considerable zone of tissue around the vibrating tips: temperatures exceeded 60° over a distance of 25 mm from the instrument. The resulting high temperatures with dissections at power setting 4 and especially 5 cause significant collateral and proximity injury to important structures, at least within 1.0 cm of the dissection plane. Contrary to high-frequency electrosurgery, ultrasonically induced collateral/proximity damage is not macroscopically visible. Thus, the tissues and structures appear normal despite extensive histologic injury, which in the case of thin-walled structures (e.g., the bile duct and ureter) may be transmural.

With both systems at power level 3, no significant hyperthermia (temperature exceeding 60°C) in the surrounding tissues was observed, and only minor inconsequential (serosal or adventitial) coagulation damage was detected. Thus, at this setting, ultrasonic dissection is entirely safe; however, the dissection process is slow, and this may not be acceptable to some surgeons. Slow dissection is, however, preferable to faster but potentially dangerous ultrasonic dissections carried out at power level 5. The other important factor in proximity hyperthermic damage by ultrasonic dissection is the duration of activation. The risk of damage is reduced if this is kept to bursts of energy not exceeding 5 seconds at any one time. Thus, based on our results, he surgeon should use power level 3 near important structures and bursts (not exceeding 5 seconds at any one time) of level 4 energy elsewhere.

Technologic improvements that would improve the safety of high-power ultrasonic dissection include cooling systems and automatic cutoffs such that the device incorporates sensor feedback to the generator, shutting off power until the temperature of the jaws of the instrument has cooled to safe levels. Studies on the relation between ultrasonic vibration and amplitude in respect to heat production may improve the functionality and reduce proximity thermal damage caused by ultrasonic dissectors/coagulators.


The authors thank all the members of the European Surgical Institute, Treudelberg, Hamburg, Germany, for their cooperation and support during this study.


Funding for experiments provided by Ethicon Endosurgery.

Correspondence: Alfred Cuschieri, MD, ChM, FRSE, FRCS, Department of Surgery, Ninewells Hospital & Medical School, Dundee DD1 9SY, Scotland.

E-mail: a.cuschieri@dundee.ac.uk

Accepted for publication May 29, 2002.


1. Amaral JF. Laparoscopic application of an ultrasonically activated scalpel. Gastrointest Clin North Am 1993; 3: 381–391.
2. Humbly R, Hebda PA, Abell E, et al. Wound healing of skin incisions produced by the ultrasonically vibrating knife, scalpel, electrosurgery and carbon dioxide laser. J Dermatol Surg Oncol 1988; 14: 11. [PubMed]
3. Kinoshita T, Kanehira E, Omura K, et al. Experimental study on heat production by a 23. 5-kHz ultrasonically activated device for endoscopic surgery. Surg Endosc 1999; 13: 621–625. [PubMed]
4. Kadesky KM, Schopf B, Blair GK. Proximity injury by the ultrasonically activated scalpel during dissection. J Paediatr Surg 1997; 32: 878–879. [PubMed]
5. Haag R, Cuschieri A. Recent advances in high-frequency electrosurgery: development of automated systems. J R Coll Surg Edinb 1993; 38: 354–364. [PubMed]
6. Mueller W, Fritzsch G. Medico-technical basis of surgery using invasive ultrasonic energy. Endosc Surg 1994; 2: 205–210. [PubMed]
7. Reidenbach HD, Buess G. Ancillary technology: electrocautery thermocoagulation and laser. In: Cuschieri A, Buess G, Perrisat L, eds. Operative manual of endoscopic surgery. Berlin/Heidelberg/New York: Springer; 1994: 46–60.
8. Hensman C, Baty D, Willis RG, et al. Chemical composition of smoke produced by high-frequency electrosurgery in a closed gaseous environment: an in-vitro study. Surg Endosc 1998; 12: 1017–1019. [PubMed]

Articles from Annals of Surgery are provided here courtesy of Lippincott, Williams, and Wilkins
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • MedGen
    Related information in MedGen
  • PubMed
    PubMed citations for these articles

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...