September 1999
Volume 40, Issue 10
Free
Retina  |   September 1999
Thresholds for Visible Lesions in the Primate Eye Produced by Ultrashort Near-Infrared Laser Pulses
Author Affiliations
  • Clarence P. Cain
    From the TASC, San Antonio, Texas; the
  • Cynthia A. Toth
    Duke University Eye Center, Durham, North Carolina; and the
  • Gary D. Noojin
    From the TASC, San Antonio, Texas; the
  • Val Carothers
    From the TASC, San Antonio, Texas; the
  • David J. Stolarski
    From the TASC, San Antonio, Texas; the
  • Benjamin A. Rockwell
    United States Air Force Research Laboratory, Brooks Air Force Base, Texas.
Investigative Ophthalmology & Visual Science September 1999, Vol.40, 2343-2349. doi:
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Clarence P. Cain, Cynthia A. Toth, Gary D. Noojin, Val Carothers, David J. Stolarski, Benjamin A. Rockwell; Thresholds for Visible Lesions in the Primate Eye Produced by Ultrashort Near-Infrared Laser Pulses. Invest. Ophthalmol. Vis. Sci. 1999;40(10):2343-2349.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

purpose. To evaluate the effects of near-infrared (near-IR) ultrashort laser pulses on the retinas of rhesus monkey eyes and to perform threshold measurements for minimum visible lesions (MVLs) at pulse widths ranging from nanoseconds to femtoseconds.

methods. Near-infrared single laser pulses were placed within the macular area of live rhesus monkey eyes for five different pulse widths (7 nsec; 80, 20, and 1 psec; and 150 fsec). One visible wavelength of 530 nm at 100 fsec was also included in the study. Visible lesion thresholds (MVL-ED50) were determined 1 hour and 24 hours after exposure. Fluorescein angiography thresholds (FAVL-ED50) were also determined using a probit analysis of the dosage. Thresholds were calculated as that dosage causing a 50% probability for damage, and the fiducial limits were calculated at the 95% confidence level.

results. For all pulse widths, the 24-hour MVL-ED50 was lower than the 1-hour MVL-ED50, and they both decreased with decreasing pulse width. Thresholds at the 1-hour reading decreased from 28.7 μJ at 7 nsec to 1.8 μJ at 150 fsec, whereas thresholds at 24 hours decreased from 19.1 μJ at 7 nsec to 1.0 μJ at 150 fsec. The doubled 1060-nm wavelength of the 530-nm threshold decreased from 0.36 to 0.16 μJ after 24 hours. FAVL-ED50s were much higher than MVL-ED50s, showing that FA was not as sensitive in determining damage levels.

conclusions. Laser pulse widths less than 1 nsec in the near-IR are capable of producing visible lesions in rhesus monkey eyes with pulse energies between 5 and 1 μJ. Also, the near-IR thresholds for these pulse widths are much higher than for the visible wavelengths. As with visible wavelengths, FA is not as sensitive in determining threshold levels as is visually observing the retina through a fundus camera.

In a previous study we reported 1 the retinal damage thresholds arising from single ultrashort laser pulses of visible wavelengths compared with other reported threshold measurements. In this study we determined retinal damage thresholds for single laser pulses of near-infrared (near-IR) wavelengths and compared our results with those previously reported for the near-IR and visible wavelengths. 1 2 3 4 5 6 7 8 9 10 11 12  
We have determined the threshold dosages for ophthalmoscopically minimum visible lesions (MVL-ED50) for pulse widths of 150 fsec; 1, 20, and 80 psec; and 7 nsec and the fluorescein angiography threshold dosages (FAVL-ED50) at 1 hour and 24 hours after exposure. Laser–ocular tissue interaction studies for pulse widths less than 1 nsec are critical to the development of safety standards and in identifying hazards to the human eye from those systems presently operating in the near-IR regime. Also, new national laser safety standards for laser systems operating in the near-IR with pulse widths less than 1 nsec are in development, and we are providing the urgently needed data in the primate fundus to assess potential human retinal hazards from these laser sources. 
The maximum permissible exposure for the retina has been established by the national laser safety standard, American National Standards Institute (ANSI) Z136.1-1993, 13 and the International Electrotechnical Commission standard, IEC-825:1-1993, 14 for visible and near-IR laser radiation at pulse widths as short as 1 nsec. There is no set standard for pulse widths less than 1 nsec, only a guideline that recommends keeping a constant irradiance that may be overly conservative. The standard was based on retinal injury studies that were conducted on primate eyes for continuous wave and pulsed laser systems with pulse widths greater than 2 nsec. We have previously reported retinal injury studies of visible wavelengths for pulse widths down to 90 fsec for pigmented rabbit eyes 15 16 and for rhesus monkey eyes. 1 17  
Methods
Experimental Systems
Two laser systems were required to produce the six pulse widths ranging from nanoseconds down to femtoseconds. Thus, the three longer pulse widths were from one system with a wavelength of 1064 nm, and the shortest pulse widths, 150 fsec and 1 psec, were from a second system operating at 1060 nm. We designated the two systems Laser System I and Laser System II. 
Laser System I consisted of a mode-locked Nd:YAG laser and an Nd:YAG regenerative amplifier. The long pulses (7 nsec) were obtained by operating the regenerative amplifier as a standard Q-switched Nd:YAG. The shorter pulses of 80 and 20 psec were generated by injection-seeding the regenerative amplifier with pulses from the mode-locked Nd:YAG. Energies up to several millijoules were available for all pulse widths, and the beam divergence was approximately 0.5 milliradians. The pulse widths were measured with a fast photodiode for 7-nsec pulses and with a slow-scan autocorrelator or streak camera for the 20-psec pulses. 
The incident laser beam from System I was apertured to 2.5 mm at a 1-m distance from the cornea to provide a uniform spatial profile for delivery to the corneal surface. This beam was delivered to the eye by deflecting it from a 1064-nm quartz beam splitter mounted on a fundus camera (Carl Zeiss, Thornwood, NY). The beam splitter was adjusted so that the deflected beam was collinear with the optical axis of the fundus camera. A low power HeNe laser was aligned collinear to the incident laser beam for location of retinal exposure sites. The 82-MHz, mode-locked Nd:YAG beam that was pulse compressed to 5 psec and frequency doubled was used to provide marker lesions within the fundus surrounding the macula. 
Laser System II operated with a mode-locked Ti:Sapphire oscillator operating at 76-MHz, 1060 nm, and a pulse width of 150 fsec. These pulses were amplified by a doubled Nd:YAG-pumped Ti:Sapphire regenerative amplifier that provided single pulses at 150 fsec with energies up to 3 mJ. Pulse widths were measured with a slow-scan autocorrelator after tuning the compressor for minimum pulse width. Marker lesions were accomplished with a shuttered continuous wave, krypton laser beam for 3 msec. Similar arrangements of the laser beam delivery were accomplished with System II and a Topcon (Paramus, NJ) fundus camera. 
For both systems, the cornea was positioned approximately 1 cm from the beam splitter so that the reflected portion of the beam entered the eye. The retina was at the focal plane of the fundus camera. The transmitted portion of the beam was directed to an energy meter so that the energy of each pulse could be recorded. The ratio of the reflected and transmitted portions of the beam was measured for each experiment and recorded. This was accomplished so that the actual energy delivered to the cornea could be calculated for each pulse. These energies and ratios were measured with a joulemeter-ratiometer (Model JD2000; Molectron, Sunnyvale, CA) with J4-09 or J3-09 detectors. Throughout this report, “laser energy delivered” is the energy delivered to the corneal surface without a contact lens or any other device to control the image size on the retina. The range of energies used for femtosecond pulses was from less than 0.1 to 14 μJ and for picosecond pulses, from 0.5 to 50 μJ. For nanosecond pulses, energies ranged from 8 μJ to more than 100 μJ. 
In Vivo Model
Mature Macaca mulatta primates from 2.2 to 6.9 kg were maintained under standard laboratory conditions (12 hours light, 12 hours dark). All primates were screened before exposure to ensure that no eye was more than 0.5 D from being emmetropic. All procedures were performed during the light cycle. Animals involved in the study were procured, maintained, and used in accordance with the Federal Animal Welfare Act, the Guide for the Care and Use of Laboratory Animals prepared by the Institute of Laboratory Animal Resources, National Research Council, and the ARVO Resolution for the Use of Animals in Ophthalmic and Vision Research. 
In Vivo Preparation
All animals were chemically restrained using 10 mg/kg ketamine HCl intramuscularly. Once restrained, 0.16 mg atropine sulfate was administered subcutaneously. Two drops each of proparacaine HCl 0.5% , phenylephrine HCl 2.5%, and tropicamide 1% were administered to both eyes. Under ketamine restraint, the primate had intravenous catheters placed for administration of warmed lactated Ringer’s solution (10-ml/kg per hour flow rate) and for administration of propofol. An initial induction dose of propofol (5 mg/kg) was administered to effect. The state of anesthesia was maintained in the subject by administering 0.2 to 0.5 mg/kg per minute propofol by syringe pump. The animal was intubated with a cuffed endotracheal tube. A peribulbar injection of 2% lidocaine was administered to reduce extraocular muscular movement. The subject was securely restrained prone on an adjustable stage for fundus photography, laser exposure, and FA. Immediately before FA, 0.6 ml Fluorescite 10% (Alcon, Fort Worth, TX) was administered as an intravenous bolus. The subject’s blood pressure, temperature, and pulse were continuously monitored throughout the experimental protocol. Normal body temperature was maintained by the use of circulating warm water blankets. 
The eyelids were held open with a wire lid speculum, and the cornea was moistened throughout the procedures with 0.9% saline solution. The retina was viewed with a modified fundus camera at all times, and all macular exposures (15–30) were delivered to the eye, without an external lens system, in a rectangular grid pattern in the macular region of the fundus. Immediately visible retinal marker lesions (created by shuttered exposures of the mode-locked, doubled, compressed Nd:YAG output at 82 MHz for the three longer pulse widths and a 3-msec shuttered exposure of krypton laser output for the 1-psec and 150-fsec pulses) were made in an L-shaped grid pattern of columns and rows to aid in localizing the exposure sites. Fundus photography (including FA) and observation of lesion formation by the researchers were performed by monocular viewing through a fundus camera’s optical system (Zeiss or Topcon). Photographs of the fundus were taken immediately before the dye injection, during fluorescein angiography, and at intervals of a few seconds until 5 minutes had elapsed, thus providing a sequence of photographs for the development of fluorescein leakage. After fluorescein injection and angiography, the lesions were also assessed for fluorescence by viewing through the camera system with excitation and a barrier filter in place. However, fluorescein leakage for the smaller lesions could not be identified by direct observation through the fundus camera, and no results using this technique are reported in this article. 
A minimum of two examiners evaluated all eyes at 1 hour and 24 hours after exposure. Visible lesions at a given exposure site were reported to be present only if two or more\E examiners identified a lesion. Color fundus photographs were taken at 1 hour and 24 hours after exposure, along with black-and-white photographs of the FA. 
Statistical Analysis
The Probit Procedure 18 was used to estimate the ED50 for creating an MVL in the retina for all pulse widths and to estimate the 95% confidence intervals for the ED50s. Enough data were recorded to ensure that the fiducial limits were reasonable and within the following limits at the 24-hour postexposure reading for visible lesions only: The upper fiducial limit could be no more than 50% greater than the ED50, and the lower fiducial limit could be no less than 50% of the ED50. 19 The above procedures were used for ophthalmoscopically visible lesions and fluorescein angiogram data at 1 hour and 24 hours after exposure. 
Results
Visible Lesion Thresholds
Visible lesion threshold data for six pulse widths are listed in Table 1 , for 1 hour and 24 hours after exposure, along with the slopes of the probit curves for the 24-hour readings. These slopes were calculated using the a probit computer program (SAS, Cary, NC). The threshold dosages at 24 hours after exposure were lower than for the 1-hour reading at all pulse widths. Also, Table 1 lists the number of subjects, eyes used, and exposures counted in the probit analysis. The right or left eye was selected randomly and depending on availability. 
Under direct ophthalmoscopic observations, the retinal response to minimal exposures was visible as a pale gray-to-white lesion increasing in whiteness and in size as energy increased in all exposures, as shown in Figure 1 . This photograph of the fundus shows the marker grid pattern for 16 exposures and visible lesions within the macular area. Lesions are clearly shown to increase in size with increasing energy of the laser pulse. 
For the 7-nsec pulses the number of lesions observable increased with time, and there was a 30% increase in the number of lesions between the 1-hour and 24-hour readings. These additional visible lesions decreased the calculated ED50 by one third, from 28.7 μJ at 1 hour to 19.1 μJ at 24 hours. The slope of the probit curve was 3.3 for the 24-hour reading, and both fiducial limits fell within the range of 0.5 to 1.5 times the ED50 value. 
Similar results were recorded for picosecond pulse widths. For 80 psec, the number of visible lesions increased by one third between the 1-hour and 24-hour readings. These additional visible lesions decreased the calculated ED50 threshold to one half the value at 1 hour. The slope of the probit curve was greater than 2, and the fiducial limits were reasonable, as described. 
The number of visible lesions for 20-psec pulse widths did not increase by as large a factor between the 1-hour and the 24-hour readings as did the other pulse widths and therefore, there was not as large a difference between the two readings. Also the slope of the probit curve at 24 hours was very large (6.7) compared with all our previous measurements. 
The three pulse widths were all generated by the Laser System I, and they were obtained from a Q-switched or mode-locked and pulse-compressed Nd:YAG laser. The other two pulse widths, 1 psec and 150 fsec, were produced by System II, which was a seeded Ti:Sapphire regenerative amplifier operating at 1060 nm and 530 nm and producing single pulses. 
The number of lesions developing within 24 hours versus 1 hour was almost double for the 1-psec pulse width (42 versus 23). Thus, the ED50 at the 24-hour reading (2.0 μJ) was slightly more than one half the ED50 at the 1-hour reading (3.8 μJ). The slope of the probit at 24 hours was 3.2, well above the desired 2 value, and the fiducial limits for both readings were within the allowed values. 
When the pulse width was reduced to 150 fsec at 1060 nm, the energy required to produce a visible lesion was also reduced. At the 1-hour reading, the ED50 for the threshold was 1.8 μJ, and that value decreased to 1.0 μJ at the 24-hour reading. The actual number of visible lesions at 24 hours increased by one third from the 1-hour reading (52 versus 38), and the slope of the probit was large (4.4). Four eyes were used for this pulse width, to ensure reasonable fiducial limits at the 24-hour reading (both were within ±20% of ED50). Measurements at 530 nm (doubled, 1060 nm) were recorded as listed in Table 1 and at the 1-hour reading. The ED50 was smaller by a factor of five below the 1060-nm threshold. After 24 hours, the ED50 for 530 nm was reduced by six times and was calculated to be the smallest threshold reported so far (0.16 μJ) for any pulse width or wavelength. At the 24-hour reading, there were almost 1.5 times the number of visible lesions at the 1-hour reading, which reduced the threshold by a factor of more than 2. 
Fluorescein Thresholds
Fluorescein angiographic threshold data for five pulse widths are listed in Table 2 for 1 hour and 24 hours after exposure, along with the slopes of the probit curves for the 24-hour readings. Across all pulse widths, the threshold for FA visibility was much higher than the threshold for MVLs. The thresholds for FAVLs at 1 hour and 24 hours decreased with pulse width until 1 psec and then increased at 150 fsec. Side-by-side comparison of FAs and fundus photographs demonstrate sites of visible retinal laser lesions without FA evidence of damage (Figs. 2 A, 2B). 
In Figure 2 note the granular patchy pattern of choroidal fluorescence in the normal primate macula (arrowheads). This pattern limits our ability to resolve a lesion on FA, particularly when it does not exhibit late leakage (increased intensity of fluorescein). The lesions in row 2, (lesions 1 and 3 from the left) do not show fluorescein leakage. 
For 7-nsec and 80-psec pulse widths, the threshold laser energy for FAVLs did not change between the 1-hour and 24-hour examinations (Figs. 3 A, 3B). With the 20-psec pulse width; however, 10 fewer lesions were visible at 24 hours, resulting in a much higher threshold for FAVLs at 24 hours. The 20-psec data at 24 hours were insufficient for fiducial limits, and the data point was therefore not statistically significant. For 1-psec and 150-fsec pulse widths, the number of angiographically visible lesions increased only slightly between 1 hour and 24 hours, thus lowering the threshold by only a small amount for FAVLs at 24 hours. Only three more lesions were visible for 1 psec at 24 hours, and the ED50 value for FAVLs was at 90% probability instead of the 95% confidence level. For the 150-fsec laser pulses, only five lesions were visible by FA for both the 1-hour and 24-hour readings. However, only three of the five lesions were the same ones for both readings. Because the FA thresholds were so much larger than the MVL thresholds, there were not enough data points recorded to determine the true FAVL-ED50 fiducial limits for both 20 psec and 150 fsec. 
The FA pattern of the test lesions was a fine, pale, hyperfluorescent spot that appeared within the first 30 seconds of the angiogram and was most prominent in the mid-arteriovenous phase or later venous phase of the angiogram. Although many of the lesions, particularly those of higher energy, had persisting or increasing hyperfluorescence in later phases of the angiogram, numerous lesions did not have any persisting hyperfluorescence or leakage beyond the margins of the lesion in late phases of the angiogram. Because of the small size of the laser lesions and the absence of persisting hyperfluorescence later in the angiogram, lower-energy laser lesions were difficult to differentiate from the normal pattern of choroidal fluorescence. Although the fundus photograph with visible lesions was used as a guide in searching for angiographic lesions, no change in the fluorescein angiogram could be found at the site of many ultrashort pulse laser retinal lesions that were visible using the fundus camera. The higher energy mode-locked marker lesions demonstrated central hypofluorescence with a ring of hyperfluorescence in early phases of the angiogram—most prominently at 24 hours. There was enlargement and blurring of the margins of all marker lesions in late phases of the angiogram, causing the lesions to stand out in contrast to the fading fluorescence of the choroidal pattern. 
Discussion
Retinal thresholds reported for the near-IR wavelengths vary over a broad range, depending on the pulse widths and experimental conditions. Most thresholds reported have been in the nanosecond regime, and the threshold for a 20-nsec pulse width 6 at 1064 nm has been reported as high as 99 μJ. In the picosecond regime, energies reported for ED50s vary from 2.2μ J at 6 psec to 13 μJ at 30 psec for 1064 nm. 5 8 The lowest retinal threshold reported for near-IR has also been for the shortest pulse width reported (6 psec). 8 Our energies for the 3 pulse widths described above (20 and 80 psec and 7 nsec) fall between these limits. 
At 7 nsec, our 1064-nm threshold, 19 μJ, was 21 times the threshold we measured at 4 nsec at 532 nm. This value is reasonable, because we expected an order of magnitude difference between the two thresholds. However, it is considerably lower than the 99 μJ reported by Lund and Beatrice 6 at a pulse width of 20 nsec or the 69 μJ reported by Ham et al. 7 for 30 nsec at 1064 nm. A much larger threshold of 158 μJ at the 1-hour postexposure reading was reported by Allen et al. 10 for 4-nsec pulse widths at 1064 nm. They found no difference between the 1-hour reading and the 24-hour reading, whereas our threshold decreased by one third between the 1-hour and 24-hour readings (28.7 μJ versus 19.1 μJ). 
Our measured thresholds at 20 psec and 80 psec of 4.6 and 4.2 μJ, respectively, are almost exactly one order of magnitude larger than the ED50 of 0.43 μJ for 532-nm exposures at 60 psec. However, both are approximately double the 2.2 μJ at 24 hours reported by Taboada and Gibbons 8 for a pulse width of 6 psec, but they are smaller than the 8.7 μJ reported by Goldman et al. 2 for a 30-psec pulse width. The thresholds for both pulse widths decreased with time, and the fiducial limits were reduced as well. A greater change was recorded for the 80-psec pulse width, because the number of visible lesions almost doubled between 1 and 24 hours, whereas there was only a 20% increase in the number of visible lesions after 24 hours for the 20-psec pulses. It is not known at this time why there was such a large difference between the thresholds for visible lesions at 1 hour between the 20-psec and 80-psec pulses, whereas this difference disappeared at the 24-hour reading. Our 1-psec ED50 of 2.0 μJ at 24 hours compares favorably with 2.2 μJ for the 6 psec pulse width. As with the 80-psec pulse width, the MVL threshold at 1 psec decreased by one half between the 1-hour and 24-hour readings. What causes these large decreases in the thresholds cannot be explained at this time; better understanding will have to wait for histology results. 
At 150 fsec and 1060 nm, our ED50 of 1.0 μJ was recorded at the shortest ever pulse width for near-IR and was the lowest threshold of all our near-IR measurements. However this 1.0 μJ was still six times the threshold measured at 530 nm. The difference between the 1060-nm and the 530-nm thresholds for 7 nsec was 21 times (19.1 μJ versus 0.90 μJ), and for 150 fsec the difference was 6 times (1.0 μJ versus 0.16 μJ). This 1.0 μJ was only slightly more than double the 0.43 μJ recorded for the visible wavelength of 580 nm at 90 fsec. The number of lesions visible after 24 hours was almost double the number after 1 hour, and the ED50 thresholds therefore decreased by almost one half at the 24-hour reading. The difference between the 1-hour and 24-hour thresholds for the 100-fsec and 530-nm pulses was more than a factor of 2, because the threshold decreased from 0.36 to 0.16 μJ. It is obvious from these data that as the pulse width became shorter, the wavelength dependence of the thresholds became less. However, the time required for the lesions to develop and become visible increased as the pulse width decreased and did not depend on the wavelength to a great extent. 
No hemorrhagic lesions were created with any of the pulse widths during the MVL threshold measurements. This was not surprising, because Allen et al. 10 reported the retinal hemorrhagic threshold for 4-nsec, 1064-nm pulses to be 340 μJ, and our pulse energies for the 7-nsec pulse width varied from 8.8 to 188 μJ within the macular area. Thus, we never came close to the ED50 threshold for hemorrhagic lesions. At 80 psec, the pulse energies varied between 1 and 54 μJ and again, no hemorrhagic lesions were created. Pulse energies varied between 0.5 and 44 μJ for the 20-psec pulse widths and from 0.1 to 7 μJ for the pulse width of 1 psec. For the 150-fsec pulse width, we varied the pulse energies from less than 0.1 to 14 μJ without producing any hemorrhagic lesions. The absence of hemorrhages from 1064-nm ultrashort pulses is in contrast to our previous reports of 532-nm or 580-nm ultrashort pulse laser delivery, when intraretinal hemorrhages were produced by energies as low as two times the MVL-ED50 threshold of 0.43 μJ. 1 We will have to wait for the results of histology before determining why no hemorrhagic lesions were produced. 
The FAVL findings suggest the retinal laser lesions from pulse energies below the FAVL threshold do not allow the leakage of fluorescein in a pattern greater than that seen in the normal choroid and retina. In related studies of 532-nm or 580-nm, ultrashort pulse laser retinal lesions, Toth et al. 20 found that visible retinal laser lesions with focal RPE damage but without extensive vacuolization did not leak fluorescein dye, and Chiu et al. 21 demonstrated intact zonula occludens between laser-damaged and adjacent RPE cells. The lower energy 1064-nm ultrashort laser pulses may produce similar RPE injury that does not break down the barrier function of the RPE. 
In Toth et al. 20 and in this study, the MVL lesion size was between 15 and 50 μm. These small lesions, if they do not allow leakage of fluorescein dye, are difficult to differentiate from the background punctate fluorescent pattern of the primate choroid. Borland et al., 9 using a higher concentration of fluorescein dye (20%), also noted this problem of “confusion between threshold lesions of small image size and the background of the choroidal flush.” 
Our FAVL-ED50 of 58 μJ for a 1064-nm, 7-nsec pulse is comparable to the 47-μJ FAVL-ED50 for the 1059-nm, 15-nsec pulse reported by Borland et al. 9 In contrast, however, we observed visible retinal lesions at much lower energies than those reported by Borland et al. (MVL-ED50 of 28.7 μJ versus 135 μJ). The difference in observation of fundus lesions may be in the method of retinal examination. Borland et al. used a direct ophthalmoscope to examine lesion development at 1 hour after exposure, which was subsequently confirmed by studying photographic enlargements of the retinal exposures taken with a fundus camera. In the current study, the fundus was observed using a fundus camera (Zeiss or Topcon) with the observers looking at magnified images 1 and 24 hours after exposure. 
Conclusions
Our data using near-IR laser pulses for the rhesus monkey can be compared with other published data included in the database used to establish the ANSI Z136.1-1993 standard. There are several reported data points for rhesus monkeys for pulse widths less than 1 nsec at near-IR wavelengths, as shown in Figure 4 , including Goldman et al., 2 Ham et al., 7 Taboada and Gibbons, 8 and Lund and Beatrice. 6 Our data points are shown below all other data points with the exception of the Taboada and Gibbons data for 6 psec. Our data show a definite downward trend from 7 nsec to 150 fsec for the 24-hour reading. For a decrease in pulse width of five orders of magnitude, our MVL-ED50 thresholds decreased from 19 to 1 μJ, or a factor of 20 times. The upper line shown in Figure 4 represents the ANSI retinal maximum permissible exposure for wavelengths between 1.05 nm and 1.15 nm for pulse widths down to 1 nsec, which is 5μ J/cm2. Thus, 2 μJ at the cornea for a pulse width of 1 nsec is considered safe; however, this safe level (upper line for 1060 nm) cannot be extrapolated to pulse widths less than 1 nsec. This is because our data include 26 lesions that resulted from pulse energies of less than 2 μJ from the total of 94 lesions created for combined pulse widths of 1 psec and 150 fsec. Thus, the ANSI standard for retinal maximum permissible exposure cannot be extended less than 1 nsec unless the MPE limits are reduced. However, it is not reasonable to continue using the ANSI recommendation of a constant irradiance for decreasing pulse width less than 1 nsec, because the safe limit for a pulse width of 100 fsec would be only 0.2 nJ or more than three orders of magnitude below our MVL-ED50
It is important to recognize that although a retinal lesion from low-energy ultrashort pulse laser injury may be seen on fundus examination, it will often not be visible on FA. This could make the assessment of a possible mild acute injury more difficult, because a small lesion may be difficult to differentiate from a drusen, and the clinician would typically expect the acute laser lesions to fluoresce. We do not yet know the impact on vision from the presumed minimal focal laser injury to RPE and possibly photoreceptors. With higher energy lesions (above the FAVL threshold) this would not be a problem. Unfortunately, at energies close to MVL-ED50, and with small retinal lesions, fluorescein leakage may not be evident. 
 
Table 1.
 
MVL Threshold for 1064-nm and 532-nm Wavelength ED50 with Fiducial Limits at the 95% Confidence Level
Table 1.
 
MVL Threshold for 1064-nm and 532-nm Wavelength ED50 with Fiducial Limits at the 95% Confidence Level
MVL Pulse Width 1-Hour Reading ED50 * 24-Hour Reading ED50 * Slope of Probit Curve
7 nsec 1064 nm (3 Subjects, 3 Eyes, 69 Exp) 28.7 (22.3–39.3) 19.1 (13.6–24.4) 3.3
80 psec 1064 nm (5 Subjects, 5 Eyes, 100 Exp) 8.1 (5.1–16.0) 4.2 (3.0–5.8) 2.2
20 psec 1064 nm (3 Subjects, 3 Eyes, 72 Exp) 5.6 (4.6–6.9) 4.6 (3.8–5.5) 6.7
1 psec 1060 nm (2 Subjects, 3 Eyes, 72 Exp) 3.8 (3.0–5.6) 2.0 (1.4–2.5) 3.2
150 fsec 1060 nm (4 Subjects, 4 Eyes, 81 Exp) 1.8 (1.2–2.7) 1.0 (0.8–1.2) 4.4
100 fsec 530 nm (2 Subjects, 4 Eyes, 63 Exp) 0.36 (0.22–0.63) 0.16 (0.11–0.23) 3.0
Figure 1.
 
Visible lesions in the fundus at 24 hours after exposure for 80-psec, 1064-nm laser pulses. Marker lesions are the column of five lesions at the left of the grid and the row of five lesions at the lower margin of the grid.
Figure 1.
 
Visible lesions in the fundus at 24 hours after exposure for 80-psec, 1064-nm laser pulses. Marker lesions are the column of five lesions at the left of the grid and the row of five lesions at the lower margin of the grid.
Table 2.
 
FAVL Threshold for 1064-nm Wavelength ED50 with Fiducial Limits at the 95% Confidence Level
Table 2.
 
FAVL Threshold for 1064-nm Wavelength ED50 with Fiducial Limits at the 95% Confidence Level
Fluorescein Angiography Pulse Width 1-HR Reading ED50* 24 HR Reading ED50* Slope of Probit Curve
7 nsec 1064 nm 56.0 (41.5–97.4) 57.6 (46.7–97.3) 7.7
80 psec 1064 nm 14.3 (10.8–22.7) 14.7 (12.0–21.0) 6.0
20 psec 1064 nm 11.8 (8.5–25.2) 22.4 (no limits) 1.1
1 psec 1060 nm 6.8 (5.2–29.4) 5.1 (4.5–6.5) 10
150 fsec 1060 nm 15.3 (9.2–262) 12.2 (8.8–none) 5.6
Figure 2.
 
(A) Fundus photograph showing the macular grid 24 hours after exposure. Marker lesions extend vertically along the left margin and horizontally across the base of the grid. At the 16 laser sites in the grid, 16 lesions were visible at 24 hours. Three additional unscored test lesions are visible outside the grid. (B) Fluorescein angiogram (FA) image at 24 hours. The marker lesions are visible, as are 13 lesions within the grid. At three sites, the laser lesion seen on the fundus photograph was not visible on FA (arrowheads).
Figure 2.
 
(A) Fundus photograph showing the macular grid 24 hours after exposure. Marker lesions extend vertically along the left margin and horizontally across the base of the grid. At the 16 laser sites in the grid, 16 lesions were visible at 24 hours. Three additional unscored test lesions are visible outside the grid. (B) Fluorescein angiogram (FA) image at 24 hours. The marker lesions are visible, as are 13 lesions within the grid. At three sites, the laser lesion seen on the fundus photograph was not visible on FA (arrowheads).
Figure 3.
 
Photographs of FA for a laser exposure of 20 psec, 1064 nm at (A) 1 hour and (B) 24 hours after exposure. Photographs taken at same time after fluorescein injection.
Figure 3.
 
Photographs of FA for a laser exposure of 20 psec, 1064 nm at (A) 1 hour and (B) 24 hours after exposure. Photographs taken at same time after fluorescein injection.
Figure 4.
 
Retinal maximum permissible exposure from ANSI Z136.1-1993. The solid lines indicate the current national standards below which radiant exposure levels are considered safe. The dots represent the database on which the safety standard is determined (triangles for laser pulses at visible wavelengths and circles for near-IR wavelengths), and the circled near-IR dots are from the present study.
Figure 4.
 
Retinal maximum permissible exposure from ANSI Z136.1-1993. The solid lines indicate the current national standards below which radiant exposure levels are considered safe. The dots represent the database on which the safety standard is determined (triangles for laser pulses at visible wavelengths and circles for near-IR wavelengths), and the circled near-IR dots are from the present study.
Cain CP, Toth CA, DiCarlo CD, et al. Visible retinal lesions from ultrashort laser pulses in the primate eye. Invest Ophthalmol Vis Sci. 1995;36:879–888. [PubMed]
Goldman AJ, Ham WT, Mueller HA. Ocular damage thresholds and mechanisms for ultrashort pulses of both visible and infrared laser radiation in the rhesus monkey. Exp Eye Res. 1977;24:45–56. [CrossRef] [PubMed]
Schuschereba ST, Lund DJ, Stuck BE, Picosecond Nd. YAG (1064 & 532) Laser effects on monkey retinal morphology. Lasers in Surgery & Medicine, Vol. 7, Supp., (15th annual meeting) April, 1995.
Bruckner AP, Schurr JM, Chang EL. Biological damage threshold induced by ultrashort 2nd and 4th harmonic light pulses from a mode-locked Nd:glass laser. Springfield, VA: National Technical Information Service, 1980, USAF SAM-TR-80–47.
Goldman AJ, Ham WT, Mueller HA. Mechanisms of retinal damage resulting from the exposure of rhesus monkeys to ultrashort laser pulses. Exp Eye Res. 1975;21:457–469. [CrossRef] [PubMed]
Lund DJ, Beatrice ES. Near infrared laser ocular bioeffects. Health Phys. 1989;56:631–636. [CrossRef] [PubMed]
Ham WT, Mueller HA, Goldman AJ, Newman BE, Holland LM, Kuwabara T. Ocular hazard from picosecond pulses of Nd:YAG laser radiation. Science. 1974;185:362–363. [CrossRef] [PubMed]
Taboada J, Gibbons WD. Retinal tissue damage induced by single ultrashort 1060 nm laser light pulses. Appl Opt. 1978;17:2871–2873. [CrossRef] [PubMed]
Borland RG, Brennan DH, Marshall J, Viveash JP. The role of fluorescein angiography in the detection of laser-induced damage to the retina: a threshold study for Q-switched, neodymium and ruby lasers. Exp Eye Res. 1978;27:471–493. [CrossRef] [PubMed]
Allen RG, Thomas SJ, Harrison RF, Zuclich JA, Blankenstein MF. Ocular effects of pulsed Nd laser radiation: variation of threshold with pulse width. Health Phys. 1985;49:685–690. [CrossRef] [PubMed]
Zuclich JA, Elliott WR, Coffey DJ. Suprathreshold retinal lesions induced by laser radiation. Lasers Light Ophthalmol. 1992;5:51–59.
Zuclich JA, Schuschereba S, Zwick H, Cheney F, Stuck BE. Comparing laser-induced retinal damage from IR wavelengths to that from visible wavelengths. Society for Photo-Optical Instrumentation Engineers. 1996;2674:66–79.
American national standard for the safe use of lasers. New York: American National Standards Institute, 1993. ANSI Standard Z136.1.
Radiation safety of laser products, equipment, classification requirements, and users guide. International Electrotechnical Commission 1993. IEC-825:1 1993.
Toth CA, Cain CP, Stein CD, et al. Retinal effects of ultrashort laser pulses in the rabbit eye. Invest Ophthalmol Vis Sci. 1995;36:1910. [PubMed]
Zuclich JA, Elliott WR, Cain CP, et al. Ocular damage induced by ultrashort laser pulses. Springfield, VA: National Technical Information Service, 1993, USAF AL/OE-TR-93-0099.
Cain CP, Noojin GD, Stolarski DJ, et al. Ultrashort Pulse Laser Effects in the Primate Eye. Springfield, VA: National Technical Information Service, 1994. USAF AL/OE-TR-1994–0141.
SAS/STAT Probit Procedure. Cary, NC: SAS Institute, 1994.
Cain CP, Noojin GD, Manning LW. A Comparison of Various Probit Methods for Analyzing Yes/No Data on a Log Scale. Springfield, VA: National Technical Information Service, 1996. USAF AL/OE-TR-1996-0102.
Toth CA, Narayan DG, Cain CP, et al. Pathology of macular lesions from subnanosecond pulses of visible laser energy. Invest Ophthalmol Vis Sci. 1997;38:2204–2213. [PubMed]
Chiu EK, Jumper JM, Rockwell BA, Worniallo E, Toth CA. The response of the retinal pigment epithelium to pico- and femtosecond laser pulses as viewed by electron microscopy [ARVO Abstract]. Invest Ophthalmol Vis Sci. 1998;39(4):S482.
Figure 1.
 
Visible lesions in the fundus at 24 hours after exposure for 80-psec, 1064-nm laser pulses. Marker lesions are the column of five lesions at the left of the grid and the row of five lesions at the lower margin of the grid.
Figure 1.
 
Visible lesions in the fundus at 24 hours after exposure for 80-psec, 1064-nm laser pulses. Marker lesions are the column of five lesions at the left of the grid and the row of five lesions at the lower margin of the grid.
Figure 2.
 
(A) Fundus photograph showing the macular grid 24 hours after exposure. Marker lesions extend vertically along the left margin and horizontally across the base of the grid. At the 16 laser sites in the grid, 16 lesions were visible at 24 hours. Three additional unscored test lesions are visible outside the grid. (B) Fluorescein angiogram (FA) image at 24 hours. The marker lesions are visible, as are 13 lesions within the grid. At three sites, the laser lesion seen on the fundus photograph was not visible on FA (arrowheads).
Figure 2.
 
(A) Fundus photograph showing the macular grid 24 hours after exposure. Marker lesions extend vertically along the left margin and horizontally across the base of the grid. At the 16 laser sites in the grid, 16 lesions were visible at 24 hours. Three additional unscored test lesions are visible outside the grid. (B) Fluorescein angiogram (FA) image at 24 hours. The marker lesions are visible, as are 13 lesions within the grid. At three sites, the laser lesion seen on the fundus photograph was not visible on FA (arrowheads).
Figure 3.
 
Photographs of FA for a laser exposure of 20 psec, 1064 nm at (A) 1 hour and (B) 24 hours after exposure. Photographs taken at same time after fluorescein injection.
Figure 3.
 
Photographs of FA for a laser exposure of 20 psec, 1064 nm at (A) 1 hour and (B) 24 hours after exposure. Photographs taken at same time after fluorescein injection.
Figure 4.
 
Retinal maximum permissible exposure from ANSI Z136.1-1993. The solid lines indicate the current national standards below which radiant exposure levels are considered safe. The dots represent the database on which the safety standard is determined (triangles for laser pulses at visible wavelengths and circles for near-IR wavelengths), and the circled near-IR dots are from the present study.
Figure 4.
 
Retinal maximum permissible exposure from ANSI Z136.1-1993. The solid lines indicate the current national standards below which radiant exposure levels are considered safe. The dots represent the database on which the safety standard is determined (triangles for laser pulses at visible wavelengths and circles for near-IR wavelengths), and the circled near-IR dots are from the present study.
Table 1.
 
MVL Threshold for 1064-nm and 532-nm Wavelength ED50 with Fiducial Limits at the 95% Confidence Level
Table 1.
 
MVL Threshold for 1064-nm and 532-nm Wavelength ED50 with Fiducial Limits at the 95% Confidence Level
MVL Pulse Width 1-Hour Reading ED50 * 24-Hour Reading ED50 * Slope of Probit Curve
7 nsec 1064 nm (3 Subjects, 3 Eyes, 69 Exp) 28.7 (22.3–39.3) 19.1 (13.6–24.4) 3.3
80 psec 1064 nm (5 Subjects, 5 Eyes, 100 Exp) 8.1 (5.1–16.0) 4.2 (3.0–5.8) 2.2
20 psec 1064 nm (3 Subjects, 3 Eyes, 72 Exp) 5.6 (4.6–6.9) 4.6 (3.8–5.5) 6.7
1 psec 1060 nm (2 Subjects, 3 Eyes, 72 Exp) 3.8 (3.0–5.6) 2.0 (1.4–2.5) 3.2
150 fsec 1060 nm (4 Subjects, 4 Eyes, 81 Exp) 1.8 (1.2–2.7) 1.0 (0.8–1.2) 4.4
100 fsec 530 nm (2 Subjects, 4 Eyes, 63 Exp) 0.36 (0.22–0.63) 0.16 (0.11–0.23) 3.0
Table 2.
 
FAVL Threshold for 1064-nm Wavelength ED50 with Fiducial Limits at the 95% Confidence Level
Table 2.
 
FAVL Threshold for 1064-nm Wavelength ED50 with Fiducial Limits at the 95% Confidence Level
Fluorescein Angiography Pulse Width 1-HR Reading ED50* 24 HR Reading ED50* Slope of Probit Curve
7 nsec 1064 nm 56.0 (41.5–97.4) 57.6 (46.7–97.3) 7.7
80 psec 1064 nm 14.3 (10.8–22.7) 14.7 (12.0–21.0) 6.0
20 psec 1064 nm 11.8 (8.5–25.2) 22.4 (no limits) 1.1
1 psec 1060 nm 6.8 (5.2–29.4) 5.1 (4.5–6.5) 10
150 fsec 1060 nm 15.3 (9.2–262) 12.2 (8.8–none) 5.6
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×