Heat production: Longitudinal versus torsionalphacoemulsification

Young Keun Han, MD, Kevin M. Miller, MD

PURPOSE: To compare the heat production of longitudinal versus torsional phacoemulsification un-der strict laboratory test conditions.

SETTING: Department of Ophthalmology, David Geffen School of Medicine at UCLA, and Jules SteinEye Institute, Los Angeles, California, USA.

METHODS: Two Infiniti phacoemulsification handpieces were inserted into silicone test chambersfilled with a balanced salt solution and imaged serially using a thermal camera. Incision compres-sion was simulated by suspending 25.3 g weights from the silicone chambers. To simulate occlu-sion of the phacoemulsification tip, the aspiration line was clamped. Peak temperatures weremeasured 0, 10, 30, 60, and 120 seconds after the commencement of continuous ultrasoundpower. The 2 handpieces, operating exclusively in longitudinal or torsional modes, were compared3 ways: (1) using the same power displayed on the instrument console, (2) using identical strokelengths, and (3) using the same applied energy, a product of stroke length and frequency.

RESULTS: For all 3 comparisons, torsional phacoemulsification resulted in lower temperatures ateach time point. At the same displayed power setting, the scenario most familiar to cataract sur-geons, longitudinal phacoemulsification elevated temperatures up to 41.5�C more than torsionalphacoemulsification.

CONCLUSIONS: Torsional phacoemulsification generated less heat than longitudinal phacoemulsifi-cation in all 3 comparison tests. Lower operating temperatures indicate lower heat generation withinthe same volume of fluid, and this may provide additional thermal protection during cataract surgery.

J Cataract Refract Surg 2009; 35:1799–1805 Q 2009 ASCRS and ESCRS

LABORATORY SCIENCE

During phacoemulsification, an ultrasonically vibrat-ing needle generates heat in a variety of ways. The pri-mary source of heat at the incision is friction betweenthe vibrating needle and the silicone sleeve covering it.

Submitted: March 12, 2009.Final revision submitted: April 19, 2009.Accepted: April 29, 2009.

From the Department of Ophthalmology, David Geffen School ofMedicine at UCLA, and the Jules Stein Eye Institute, Los Angeles,California, USA.

Neither author has a financial or proprietary interest in any materialor method mentioned.

Supported in part by a grant from the Carl and Roberta DeutschFoundation, Santa Monica, and the Paul J. Vicari endowed researchfund at UCLA, Los Angeles, California, USA.

Corresponding author: Kevin M. Miller, MD, Jules Stein Eye Insti-tute, 100 Stein Plaza, UCLA, Los Angeles, California 90095-7002,USA. E-mail: kmiller@ucla.edu.

Q 2009 ASCRS and ESCRS

Published by Elsevier Inc.

If sufficient heat is generated, the temperature of thephacoemulsification needle can rise high enough toburn tissue at the incision site. Burns can be severe,making it difficult to close an incision and inducingconsiderable astigmatism.1,2

The temperature required to produce a burn ina porcine eye can be as low as 44.2�C.3,4 In our anec-dotal experience, we found that a temperature of50�C or higher for several seconds is required to pro-duce a corneal burn.1

In longitudinal phacoemulsification, movement ofthe tip of the needle where lens disruption occurs,the so-called stroke, is equal to the movement of theshaft of the needle inside the incision. In torsional pha-coemulsification, small-angle rotational movement ofthe shaft is translated into horizontal stroke at the tipof the needle. The stroke amplification depends onthe angle of the bend near the distal end of the needleand the length of the shaft beyond the bend (Figure 1).Because of the geometric design of available bent tips,for equivalent longitudinal and horizontal strokes at

0886-3350/09/$dsee front matter 1799doi:10.1016/j.jcrs.2009.04.046

1800 LABORATORY SCIENCE: HEAT PRODUCTION DURING PHACOEMULSIFICATION

the tip, torsional phacoemulsification should generateless frictional heat at the incision because of reducedmovement at that location.

The purpose of this study was to compare the heatproduction of longitudinal versus torsional phacoe-mulsification under strict laboratory test conditions.In our experiments, we used temperature measuredin degrees Celsius as a proxy for the heat generated.Heat is thermal energy that can be transferred from 1object or system to another as a result of a differencein temperatures.

MATERIALS AND METHODS

A straight 30-degree non aspiration bypass system (ABS) mi-crotip on an Infiniti ultrasound handpiece (Alcon, Inc.) anda 45-degree Kelman mini-flared ABS tip on an OZil torsionalhandpiece (Alcon, Inc.) were each capped with silicone testchambers filled with balanced salt solution (BSS) (volume0.85 cc, approximately 4 times greater than the anterior cham-ber of the human eye). The outside surfaces of the test cham-bers were imaged in the infrared spectrum usinga ThermaCAM (model P60, FLIR Systems, Inc.). All objectsemit infrared radiationasa functionof temperature.A thermalcamera captures the infrared radiation that is emitted from anobject’s surface and converts it to temperature readings. In theexperiments, the camera was set to display temperatures be-tween 20�C and 90�C (dynamic range of camera: �40�C to120�C). Camera calibration was verified by measuringwarm- and cold-water baths using the infrared camera andthen comparing the readingswith those from a thermocoupleat temperatures across the experimental range (Figure 2).

The Infiniti Vision System with Intrepid Fluidics Man-agement System (Alcon, Inc.) was used to operate bothhandpieces. Elastomeric bands were used to suspend 25.3g weights from the silicone test chambers to simulate tightcorneal incisions (Figure 3). The aspiration line wasclamped using needle-nosed pliers to simulate completeocclusion of the tip.

Figure 1. Relationship between movement of the tip and movementof the shaft of a phacoemulsification needle in longitudinal and tor-sionalmodes.Note the amplification of the side-to-sidemovement atthe tip in torsional mode compared with the movement of the shaft.

J CATARACT REFRACT SUR

The thermal camera captured still images of the probes be-fore power application (at 0 seconds) and then 10, 30, 60, and120 seconds after power application. The longitudinal hand-piece and torsional handpiece were tested separately. Experi-ments were allowed to run for 2 minutes because the largevolume of fluid in a test chamber takes significantly longertoheat than theaqueoushumor in thehumaneye.The thermalcamera has a real-time analysis tool withwhich a circle can bedrawn in the field of view so that one can determine the high-est surface temperature inside the circle. A circle was drawnaround the point of contact where the edge of the elastomericband pressed against the silicone test chamber. The imageswere recorded and saved to a compact flash card. Figure 4shows sequential thermal images of a silicone test chamber.

An infrared thermal camera is nonintrusive and useful inthese dynamic situations because it allows measurement ofthe temperature of a silicone test chamber in real time with-out direct contact. Unlike a thermocouple, a thermal camerais not adversely affected by vibration. Compared with previ-ous experiments1,5 in which the temperature was measuredacross the entire area of the test chamber, this experiment

Figure 2. Calibration of the infrared camera was verified againsta thermocouple by measuring water baths of different temperaturesacross the experimental range.

Figure 3. Wound compression was simulated by suspending 25.3 gweights from BSS-filled silicone test chambers using elastomericbands.

G - VOL 35, OCTOBER 2009

1801LABORATORY SCIENCE: HEAT PRODUCTION DURING PHACOEMULSIFICATION

Figure 4. Thermal images obtained with thehandpiece operating in longitudinal modeat 100% power. The first image was recordedbefore power application and subsequentimages were recorded 10, 30, 60, and 120 sec-onds after power application. The highesttemperature (�C) in each circle is displayedat the top right corner.

focused on the more clinically relevant area of interest, theincision area, where the corneal incision would apply a loadand interact with the vibrating needle. However, becausethe thermal camera can only analyze exposed surface areasand not any location beneath the surface, this method couldnot determine the actual temperature of contact area be-tween the irrigation sleeve and the test chamber. Eventhough access to this location is not possible with thismethod, the focus on the intersection between the loadedelastomeric band and the silicone chamber provides usefuland relevant comparative thermal data between ultrasonicmodalities.

Three types of experiments were performed to simulate 3types of equivalent power applications.

Experiment 1

This experiment compared heat production in longitudi-nal and torsional phacoemulsification modes under thesame power settings displayed by the instrument console.Powerswere set to 40%, 60%, 70%, and 100% for longitudinalphacoemulsification and then separately for torsional pha-coemulsification at the same amplitude values.

Experiment 2

In this experiment, heat production under conditions ofidentical stroke was compared. To measure the strokelength, sleeveless phaco tips were placed in a fluid tankwith a viewing window and imaged with a video micro-scope (Meiji Techno). When ultrasound power was applied,the tip movement was measured using a VIA-150 video im-age marker-measurement system (Boeckeler Instruments)(Figure 5). Stroke lengths for longitudinal mode at 40%,60%, 70%, and 100% power were 1.55, 2.40, 2.80, and 3.45mils (thousandths of an inch), respectively, whichwere in ac-cordance with stroke lengths in the torsional mode at 25%,

J CATARACT REFRACT SURG

45%, 60%, and 80% power. The temperatures produced byoperating the handpieces at these identical longitudinaland torsional stroke lengthsweremeasured out to 2minutes.

Experiment 3

In this experiment, heat production under the same theo-retical applied energy (kHz*mils) was compared. Appliedenergy for an oscillating object is the product of frequencyand stroke length (ie, energy Z frequency � stroke length).Because the 2 phacoemulsification modalities have differentfrequencies (37 kHz for longitudinal mode and 32 kHz fortorsional mode), appropriate phacoemulsification powersfor each modality were selected to achieve stroke lengthsthat would result in the comparable applied energies re-quired for this experiment. Therefore, the temperature dur-ing torsional mode operation at 30%, 60%, 75%, and 95%power was measured, which resulted in comparable appliedenergy during longitudinal mode operation at 40%, 60%,70%, and 100% power, respectively. Table 1 compares thecalculated energy for each mode and power. From this anal-ysis, control-panel power settings were chosen to producecomparable applied energies.

RESULTS

Figures 6 to 8 show the results of the 3 experiments.Each data point represents the highest temperature re-corded at the point of greatest test chamber compres-sion. This point would correspond to the incision sitein a clinical scenario.

Experiment 1

Longitudinal-mode operation produced highertemperatures than torsional-mode operation at 60%,

- VOL 35, OCTOBER 2009

1802 LABORATORY SCIENCE: HEAT PRODUCTION DURING PHACOEMULSIFICATION

70%, and 100% displayed console power settings at 30seconds and beyond. The temperature difference in-creased over time, although not necessarily propor-tional to the power setting.

At 40% displayed power, both modes showed noappreciable temperature differences in the first 30 sec-onds. Temperature measured using the longitudinal

Figure 5. Stroke length was measured using a video image marker-measurement system.

mode was 4.1�C higher than using the torsionalmode at 60 seconds and 14.3�C higher at 120 seconds.

At 60%, 70%, and 100%, the temperature differenceswere less than 1.5�C in the first 10 seconds. The tem-perature increases for longitudinal mode averaged3�C at 30 seconds and 13.9�C at 60 seconds. At 120 sec-onds, the increases ranged between 22.4�C and 36.8�C(Figure 6).

Experiment 2

Figure 9 compares the stroke length (mils) at the endof the selected tips for longitudinal and torsionalmodes in 10% power increments. At every power set-ting, the longitudinal mode showed a shorter strokelength than the torsional mode.

Figure 7 shows the maximum temperatures gener-ated by the longitudinal and torsional modes whenthe stroke lengths were equivalent. At 1.55 mils, bothmodes showed no appreciable temperature differ-ences in the first 30 seconds. Temperature measuredusing the longitudinal mode was 4.8�C higher thantorsional mode at 60 seconds and 14.6�C higher at120 seconds.

At 2.40, 2.80, and 3.45 mils, longitudinal mode tem-peratures were higher than torsional mode tempera-tures by a mean of 1.6�C at 10 seconds, 4.5�C at 30seconds, and 16.8�C at 60 seconds, respectively. At120 seconds, the increases ranged between 29.2�Cand 41.5�C.

Experiment 3

Figure 8 shows the maximum temperatures gener-ated by the longitudinal and torsional modes whenthe theoretical applied energies were equivalent. Thelongitudinal mode resulted in higher temperaturesthan torsional mode at all equivalent applied energiesat each time point.

At 57 kHz*mils, both modes showed no appreciabletemperature differences in the first 30 seconds. Themeasured temperature in longitudinal mode was

Table 1. Calculated applied energies.

Approximate AppliedEnergy (kHz*mils)

Longitudinal Mode (37 kHz) Torsional Mode (32 kHz)

Power (%)Stroke

Length (mils)Calculated AppliedEnergy* (kHz*mils) Amplitude (%)

StrokeLength (mils)

Calculated AppliedEnergy (kHz*mils)

57 40 1.55 57.4 30 1.25 56.089 60 2.40 88.8 60 2.80 89.6104 70 2.80 104 75 3.25 104129 100 3.45 128 95 4.05 130

kHz*mils Z theoretical applied energy; mils Z thousandths of an inch*Ultrasound frequency � stroke length

J CATARACT REFRACT SURG - VOL 35, OCTOBER 2009

1803LABORATORY SCIENCE: HEAT PRODUCTION DURING PHACOEMULSIFICATION

Figure 6. Graphs from experiment1 showing the maximum tempera-tures generated by longitudinaland torsional phacoemulsificationmodes when the power displayedon the instrument console was40%, 60%, 70%, and 100%.

4.4�C higher than torsional mode at 60 seconds and14.3�C higher at 120 seconds.

For the other applied energies, the longitudinalmode temperatures were higher than the torsionalmode temperatures by a mean of 1.4�C at 10 seconds,3.2�C at 30 seconds, and 13.3�C at 60 seconds. At 120seconds, the increases ranged from 19.9�C to 36.8�C.

DISCUSSION

Torsional phacoemulsification has several advantagesover longitudinal phacoemulsification, including in-creased followability, decreased energy use, shortersurgical time, and less endothelial damage in cataractsurgery.6,7 Mechanically, many of the differences canbe explained by the side-to-sidemotion of the torsional

Figure 7. Graphs from experiment2 showing the maximum tempera-tures generated by longitudinaland torsional phacoemulsificationwhen the stroke lengths wereequivalent at 1.55, 2.40, 2.80, and3.45 mils (thousandths of an inch).

J CATARACT REFRACT SURG - VOL 35, OCTOBER 2009

1804 LABORATORY SCIENCE: HEAT PRODUCTION DURING PHACOEMULSIFICATION

Figure 8. Graphs from experiment3 showing the maximum tempera-tures generated by longitudinaland torsional phacoemulsificationwhen the theoretical applied ener-gies were equivalent at 57, 89, 104,and 129 kHz*mils (theoretical ap-plied energy).

phacoemulsification tip with reduced repulsion of lensfragments.6,7

In an effort to repeat the success of a previous thermalstudy inwhichwe compareddifferent longitudinal plat-forms,5wemeasured thermaldifferences between longi-tudinal and torsional modalities. In experiment 1, thetorsional mode always generated less heat than the lon-gitudinal mode at the same displayed power setting.However, because of the marked difference in how thetorsional tip moves, the actual effective stroke, not thepower settingdisplayedby the instrument,maybeabet-ter way to make comparisons. Thus, we performed ex-periment 2, in which equal stroke lengths at the cuttingtipwere equatedandthe thermal resultsweremeasured.

Experiment 2 also resulted in lower temperaturesfor the torsional mode. Again, we recognized an op-portunity to improve the comparison and level theplaying field. The theoretical delivered ultrasonic en-ergy is proportional to the product of the ultrasonicfrequency and the stroke length.8 In practice, lowerfrequencies appear to be less efficient for emulsifyinga lens, whereas higher frequencies may have the po-tential to generate more heat. The frequency of theOZil torsional technology is 32 kHz, which is lowerthan the 37 kHz of the manufacturer’s longitudinal ul-trasound handpiece, and therefore results in an energysavings of 13.5% over any given period of time.

In longitudinal mode, movement of the tip’s cuttingedge is exactly the same as movement at the incisionsite along the shaft. In contrast, in torsional mode, mo-tion of the needle inside the incision site is less than thestroke of the tip’s cutting edge because the bent tip

J CATARACT REFRACT SU

amplifies the torsional movement and translates itinto a nearly horizontal motion.

It makes some intuitive sense that equal appliedultrasonic energy should generate equal heat and tem-perature rises during a procedure. In experiment 3,with equal applied energy, the torsionalmode resultedin lower peak temperatures at the point of siliconesleeve compression at every time point. This is the ef-fect of stroke amplification. Thus, we determined thatthe reduction in temperature with torsional operationis due the reduced stroke length at the incision coupledwith a lower operating frequency, the primary

Figure 9. Longitudinal and torsional stroke under the same dis-played console power.

RG - VOL 35, OCTOBER 2009

1805LABORATORY SCIENCE: HEAT PRODUCTION DURING PHACOEMULSIFICATION

elements in the applied energy equation. The results inexperiment 3 show that the applied energy required tobreak up lens material may not be the same energy de-livered to the incision site when comparing torsionaland longitudinal modalities.

We were interested in comparing the longitudinaland torsional modes of operation under conditionsthat would maximize the chance of finding differ-ences. Inflow was clamped to eliminate the cooling ef-fect of irrigation flow. Temperatures were measuredfor longer than would be necessary in clinical testingbecause it takes longer for frictional heat to warm thelarge volume of fluid inside a silicone sleeve than towarm the smaller volume of balanced salt solution in-side the anterior chamber of the eye. Temperature risesin the experimental setup can barely be measured forpower applications of fewer than 10 seconds and lessthan 100% in amplitude. We did not test pulse modesor combined torsional and longitudinal modes of op-eration, again because we wanted to maximize thechances of finding differences.

In summary, our study found that torsional phaco-emulsification produced less heat than longitudinalphacoemulsification in all 3 comparisons.

REFERENCES1. Olson MD, Miller KM. In-air thermal imaging comparison of Leg-

acy AdvanTec, Millennium, and Sovereign WhiteStar phacoemul-

sification systems. J Cataract Refract Surg 2005; 31:1640–1647

J CATARACT REFRACT SURG

2. Sugar A, Schertzer RM. Clinical course of phacoemulsification

wound burns. J Cataract Refract Surg 1999; 25:688–692

3. Bissen-Miyajima J, Shimmura S, Tsubota K. Thermal effect on

corneal incisions with different phacoemulsification ultrasound

tips. J Cataract Refract Surg 1999; 25:60–64

4. Ernest P, Rhem M, McDermott M, Lavery K, Sensoli A. Phacoe-

mulsification conditions resulting in thermal wound injury. J Cata-

ract Refract Surg 2001; 27:1829–1839

5. Miller KM, Olson MD. In water thermal imaging comparison of the Al-

con Legacy and AMO Sovereign phacoemulsification systems.

Open Ophthalmol J 2008; 2:20–26. Available at: http://www.pub

medcentral.nih.gov/picrender.fcgi?artidZ2687103&blobtypeZpdf.

Accessed June 1, 2009

6. Liu Y, Zeng M, Liu X, Luo L, Yuan Z, Xia Y, Zeng Y. Torsional mode

versus conventional ultrasound mode phacoemulsification; ran-

domized comparative clinical study. J Cataract Refract Surg

2007; 33:287–292

7. Vargas LG, Holzer MP, Solomon KD, Sandoval HP, Auffarth GU,

Apple DJ. Endothelial cell integrity after phacoemulsification

with 2 different handpieces. J Cataract Refract Surg 2004; 30:

478–482

8. Packer M, Fishkind WJ, Fine IH, Seibel BS, Hoffman RS.

The physics of phaco: a review. J Cataract Refract Surg 2005; 31:

424–431

First author:Young Keun Han, MD

Department of Ophthalmology, DavidGeffen School of Medicine at UCLA,and Jules Stein Eye Institute,Los Angeles, California, USA

- VOL 35, OCTOBER 2009