Abstract
Objective
The objective of this study was to evaluate the effect of indocyanine green (ICG) concentration, distance, angle, and penetration depth on average fluorescence intensity (AFI) when using 2 imaging devices in a canine ex vivo model.
Methods
3 concentrations of ICG were evaluated (0.006, 0.011, and 0.022 mg/mL) for the effect of distance, penetration depth, angle, and ambient light on AFI. Penetration depth was assessed using serially stacked 1-mm sheets of a gelatin phantom. Each set of conditions was tested using a VITOM 3D and 30° 10-mm laparoscope. Polynomial regression was used to evaluate for significance.
Results
Average fluorescence intensity increased with increasing distance for the 30° 10-mm laparoscope. The VITOM 3D displayed higher AFIs at closer and farther distances. Average fluorescence intensity decreased as penetration depth increased for both devices. Ambient light had no effect on AFI, and AFI varied by angle, but statistical analysis for these variables was limited by insufficient data points.
Conclusions
Average fluorescence intensity was affected by distance, penetration depth, and angle for each ICG concentration and imaging device. There was no apparent effect on AFI with or without ambient light.
Clinical Relevance
The interest in near-infrared fluorescence with ICG is growing in veterinary medicine. Optimization of protocols using these techniques can provide veterinarians with guidelines for the implementation of near-infrared fluorescence with ICG. Further studies are needed to assess the relationship between distance, depth, angle, and ambient light with AFI in vivo.
Indocyanine green (ICG) is a nontoxic, fluorescent, tricarbocyanine iodide dye with many clinical applications when used in combination with near-infrared fluorescence (NIRF) imaging, including intraoperative tissue perfusion assessment, identification of neoplastic tissue (through preferential uptake and retention of ICG and via increased vascular density with abnormally structured vessels), tumor resection margin evaluation, and evaluation of organ function.1–8 The use of NIRF with ICG provides a method in which anatomical structures can be highlighted with fluorescence in real time without altering the operative field. The images provide objective data to aid in intraoperative decision-making. Near-infrared fluorescence imaging systems use light spectra with wavelengths between 700 and 900 nm, allowing for deeper tissue penetration than is possible with the human eye.9 Indocyanine green can be administered as an IV, SC, or intradermal injection to the patient, binds to plasma proteins such as albumin, and is exclusively metabolized by the liver.1,2 The use of NIRF with ICG is gaining interest within veterinary medicine and has been successfully used in hepatic and pulmonary tumor identification, cholangiography, sentinel lymph node evaluation, ophthalmic angiography, skin flap viability, and thoracic duct ligation.2–6,10 These studies show the diversity in the potential clinical applications of ICG. However, when reviewing the methods used for each study, the protocols and techniques used by various institutions vary widely.
Despite the increased interest in NIRF with ICG in veterinary medicine, few established guidelines are available to advise surgeons on its usage. Two systematic reviews of the use of NIRF with ICG for human skin flap reconstructive surgeries and laparoscopic cholecystectomies highlight the variation in the administered dose and concentration of ICG, positioning of the imaging devices, and interpretation of the signal.11,12 In these studies, ex vivo experiments were also performed to test the effect of distance to the target, angle, and penetration depth on average fluorescence intensity (AFI). Because ICG binds to plasma proteins in vivo, human albumin was used for these studies, and general considerations were made for the future use of NIRF with ICG for human procedures. The objective of this study was to test the effect of ICG concentration, distance, angle, and penetration depth on AFI in a canine ex vivo model.
Methods
Experiments were performed using 3 previously established doses of ICG (0.5, 1, and 2 mg/kg; Diagnostic Green LLC).2,3,5,10–12 The concentration of ICG was determined by first calculating the total milligram dose for each milligram per kilogram dose based on a 20-kg canine patient. Then, the estimated blood volume (90 mL/kg) of a 20-kg dog was calculated. The milligram dose was divided by the average canine blood volume (1,800 mL) to achieve the concentration to be used in the 24-well plates, which were 0.006, 0.011, and 0.022 mg/mL. Indocyanine green was first diluted using sterile water following manufacturer instructions. The ICG-sterile water solution was then further diluted with canine plasma to achieve the appropriate concentrations. This plasma was obtained from a blood donor dog with an albumin level of 3.8 g/dL (normal reference range: 3.3 to 4.2 g/dL).
Three 24-well plates were used in this experiment, each corresponding to a different concentration of ICG (0.006, 0.011, and 0.022 mg/mL; Figure 1). For each dose, the effect of distance, angle, and penetration depth on the fluorescence intensity was measured. Each set of conditions was tested using a VITOM 3D (0°; Karl Storz) and 30° 10-mm laparoscope (Karl Storz; Figure 1). Each experiment testing a specific parameter (ie, distance, angle, penetration depth) was performed with the room lights off and with windows covered to minimize the effects of ambient light.
The setup of ex vivo experiments. A—Example of image of the experimental setup with the VITOM 3D; the same setup was used for the 30° laparoscope. B—24-well plates with each concentration of indocyanine green diluted in canine albumin: 0.022 mg/mL (top), 0.011 mg/mL (middle), and 0.006 mg/mL (bottom). C—Image from a penetration depth measurement with a 2-mm gelatin phantom placed over the wells. D—Example of an image taken with the VITOM 3D at a 90° angle. Average fluorescence intensity was determined by measuring the central well.
Citation: American Journal of Veterinary Research 2025; 10.2460/ajvr.25.01.0001
Distance
The effect of distance on AFI was measured by incrementally increasing the distance between the tip of the telescope and the surface of the well plate. When using the VITOM 3D, distances ranging from 20 to 50 cm in 5-cm increments were used based on previously published data.11 When using the 30° laparoscope, distances ranging from 1 to 10 cm in 1-cm increments were used.12 Shorter distances were used in the laparoscope group as compared to the VITOM 3D group due to limited working distance within the abdominal cavity of veterinary patients as compared to an open abdominal approach when using the VITOM 3D and due to the magnified field of view with a VITOM 3D. All distances were measured at a 90° angle to the well plate.
Angle
Four angles, 90°, 75°, 60°, and 45°, were tested using the VITOM 3D and 30° laparoscope. The selected angles were based on previous studies.11,12 The angle was measured between the tip of the telescope and the center well using a goniometer. The selected angles were measured at a distance of 5 cm for the 30° laparoscope and at a distance of 35 cm for the VITOM 3D as these were both the median distance in the range of distances measured for each imaging modality.
Penetration depth
Penetration depth was measured using a gelatin tissue phantom.13 The gelatin model was prepared by adding 5 g of gelatin into 25 g of deionized water. The amount of gelatin was doubled as compared to the original protocol to create firmer sheets of the tissue phantom. Changing the percentage of gelatin did not affect the light-scattering properties of the gelatin phantom in the original study.13 This mixture was placed in the microwave for 30-second intervals until the gelatin was completely dissolved, making sure to mix the solution slowly between intervals to avoid excess bubbling. The mixture was cooled to 37 °C. Then, 0.1 mL of whole blood, which was collected in a heparinized blood collection tube, and 3 g of Intralipid 20% (Fresenius Kabi) were gently mixed together. The canine whole blood and plasma used in this experiment came from the same canine donor. The mixture was added to the gelatin slowly. The resulting gelatin mixture was drawn up into a 20-mL syringe to a predetermined amount (14.5 mL) that would create a thickness of 1 mm in a petri dish, and then it was added to the petri dish. The petri dishes were set in a 4 °C refrigerator until firm. The resulting gelatin phantom was trimmed to the desired size and serially stacked onto the 24-well plates to mimic normal tissue overlying the ICG signal until a maximum penetration depth was reached (Figure 1). Due to the staining of the gelatin phantom with ICG in a full well, 1 mL of the ICG diluted in plasma was removed from each well, leaving 2 mL total per well. This prevented the gelatin from contacting the phantom, which caused areas of hyperfluorescence in the resulting images. Penetration depth was measured at a distance of 5 cm for the 30° laparoscope and at a distance of 35 cm for the VITOM 3D, as these were both the median distance in the measured distances for each imaging modality. All penetration depths were measured at an angle of 90° to the well plate.
Ambient light
The effect of ambient light was also investigated by turning the room lights on. Experiments were performed in a 251–sq ft room with 4 overhead fluorescent lighting fixtures. Two windows within the doors to the room were covered with impermeable draping material. The AFI with lights on was measured at a distance of 5 cm for the 30° laparoscope and at a distance of 35 cm for the VITOM 3D, with all imaging devices and ICG concentrations at an angle of 90° to the well plate.
Fluorescence intensity
In all experiments, AFI was measured on a grayscale from 0 (black) to 255 (white) using ImageJ software (ImageJ, version 1.51; NIH). A higher AFI corresponds to a brighter and whiter image. To calculate AFI, the free-hand tool in ImageJ was used to trace the center well of the image, and the grayscale value was recorded (Figure 1). Tracings were performed by one researcher (AKS). To account for variability in tracing the center well, the final AFI for each image was measured three times and the values were averaged.
Statistical analysis
The association between distance, angle, and penetration depth with AFI was defined using scatter plots with lines fitted using polynomial regression due to an observed curvilinear association. Polynomial regression (quadratic, cubic, or quartic) was used to test for significance with higher R2 indicating reduced variance in the model. P < .05 was considered acceptable. All analyses were performed using JMP Pro, version 17.0.0 (JMP Statistical Discovery LLC).
Results
Average fluorescence intensity was positively correlated with distance for both imaging devices (Figure 2; Table 1). Average fluorescence intensity generally increased as distance increased when using the 30° laparoscope. The lowest AFIs were seen at a distance of 1 cm and AFIs stabilized between 4 and 10 cm. The VITOM 3D displayed higher AFIs at closer and farther distances, while the lowest AFI was seen at 35 cm. These trends remained true across each concentration of ICG for each imaging device.
Effect of distance (A and B) and penetration depth (C and D) on average fluorescence intensity (FI) using a 30° 10-mm laparoscope (A and C) and VITOM 3D (B and D) for different concentrations: 0.006 mg/mL (blue), 0.011 mg/mL (red), and 0.022 mg/mL (green).
Citation: American Journal of Veterinary Research 2025; 10.2460/ajvr.25.01.0001
Results of polynomial regression for distance measurements with the 30° laparoscope and the VITOM 3D for each tested concentration.
| 0.006 mg/mL | 0.011 mg/mL | 0.022 mg/mL | |
|---|---|---|---|
| 30° laparoscope | P = .001 (0.961) | P < .001 (0.968) | P = .001 (0.960) |
| VITOM 3D | P < .001 (0.990) | P = .026 (0.987) | P = .002 (0.961) |
Values in parentheses represent the R2 value. Distance for the 30° laparoscope was evaluated with quartic polynomial regression, and distance for the VITOM 3D was evaluated with quadratic polynomial regression for the 0.006- and 0.022-mg/mL concentrations and quartic polynomial regression for the 0.011-mg/mL concentration.
Average fluorescence intensity decreased as penetration depth increased for each ICG concentration and imaging device (Figure 2; Table 2). The maximum penetration depth for each imaging device ranged from 5 to 6 mm. The maximum penetration depth for the 30° laparoscope was 6 mm for the 0.006- and 0.022-mg/mL concentrations and 5 mm for the 0.011-mg/mL concentration. The VITOM 3D had the lowest penetration depths; thus, there were not enough data points available for statistical analysis. The maximum penetration depth when using the VITOM 3D was 5 mm, which was only achieved with the 0.011-mg/mL concentration.
Results of polynomial regression for penetration depth measurements with the 30° laparoscope for each tested concentration.
| 0.006 mg/mL | 0.011 mg/mL | 0.022 mg/mL | |
|---|---|---|---|
| 30° laparoscope | P = .007 (0.964) | P = .029 (0.971) | P = .010 (0.955) |
Values in parentheses represent the R2 value. Penetration depth for the 30° laparoscope was evaluated with quadratic polynomial regression.
The effect of angle on AFI was measured using the VITOM 3D and 30° laparoscope (Figure 3). Average fluorescence intensity varied based on angle; however, statistical analysis could not be performed due to too few data points. When using the 30° laparoscope, the maximum AFI for each concentration was highest at a 45° angle. The maximum AFI for the VITOM 3D was at 90°.
Effect of angle on average fluorescence intensity (FI) using a 30° 10-mm laparoscope (A) and VITOM 3D (B) for different concentrations: 0.006 mg/mL (blue), 0.011 mg/mL (red), and 0.022 mg/mL (green).
Citation: American Journal of Veterinary Research 2025; 10.2460/ajvr.25.01.0001
Average fluorescence intensity did not appear to vary between groups when testing the effect of ambient light (Table 3); however, statistical analysis could not be performed due to too few data points.
Results of the presence of ambient light on average fluorescence intensity for each imaging device and concentration of indocyanine green.
| 0.006 mg/mL | 0.011 mg/mL | 0.022 mg/mL | ||||
|---|---|---|---|---|---|---|
| Lights on | Lights off | Lights on | Lights off | Lights on | Lights off | |
| 30° laparoscope | 115.729 | 115.484 | 112.6 | 113.026 | 111.766 | 112.55 |
| VITOM 3D | 99.165 | 98.107 | 99.509 | 98.866 | 100.662 | 99.432 |
The laparoscope was held at 5 cm, while the VITOM 3D was held at 35 cm.
Discussion
The aim of this study was to test the effect of various factors on AFI using 2 imaging devices in a canine ex vivo model. Distance from the imaging device and AFI was similar for each concentration of ICG; however, the trend varied between the imaging devices. With the 30° laparoscope, the lowest AFIs were seen at 1 cm, whereas higher AFIs were seen at 4- and 10-cm distances. These findings differ from an ex vivo study11 that showed a lower AFI was seen at greater distances when using a 30° 10-mm laparoscope. Similarly, a study14 comparing 5 laparoscopic systems in their use for fluorescence cholangiography found that AFI decreased by 50% when the distance between the bile duct and laparoscope increased from 5 to 15 cm. A possible explanation for this discrepancy seen in the present study is that dark shadows created by the indentations at the bottom of the 24-well plates were better seen at shorter distances. When measuring the AFI, these shadows could not be excluded from the fluorescence measurement, therefore artificially decreasing the AFI (Supplementary Figure S1). With the VITOM 3D, AFI was higher at shorter and longer distances and was lowest at a midrange distance. Similar shadowing at the bottom of the wells was seen with the VITOM 3D as the laparoscopes; however, it was less obvious as compared to the laparoscopic images.
Penetration depth was evaluated using a phantom consisting of canine whole blood, lipid, and gelatin. The protocol for the gelatin phantom was previously developed with the intention to be used as a low-cost optical tissue-mimicking phantom for biomedical optics research.13 In the present study, the formulation was altered to include double the amount of gelatin to provide a more rigid structure to allow ease of manipulation of the gelatin sheets. The original study13 found that altering the gelatin-to-water ratio did not alter the light-scattering properties of the tissue phantom. Previous studies11,12 have used sheets of beeswax to simulate tissue in the same manner. The gelatin phantom was chosen in the present study with the intention of more closely simulating real tissue. However, this is the first use of this specific tissue phantom to evaluate NIRF with ICG; thus, the true similarity to animal tissue is unknown.
Based on studies testing NIRF with ICG in human patients, similar maximum penetration depths were found using this phantom. In a study8 evaluating the use of NIRF with ICG for laparoscopic hepatectomy in human patients, only hepatic tumors that were 5 mm or closer to the liver surface could be identified. Previous studies11,12 using stacked sheets of beeswax found a maximum penetration depth of 3.6 mm when using a 30° laparoscope and 4 mm when using a handheld NIRF camera (Fluobeam; Fluoptics). Based on the similarity of these results to the findings of the present study, the gelatin phantom may be an acceptable experimental replacement for tissue. However, beeswax sheets were not compared to the gelatin phantom in this study, therefore a direct comparison cannot be made. Knowledge of maximal penetration depths can assist clinicians in understanding the limitations of NIRF with ICG when evaluating tumors deeper within the parenchyma of an organ or tissue. Interestingly, similar experiments utilizing beeswax sheets found that lower fluorescence intensities were seen with higher doses of ICG.11 The authors hypothesized this phenomenon occurred because a 40 mg/mL albumin and 0.9% NaCl solution was used to dilute the ICG. Therefore, lower concentrations of ICG received a higher absolute quantity of albumin due to receiving a larger amount of the dilutant. In the present study, canine plasma was used to dilute the ICG and was measured to have an albumin concentration of 38 mg/mL. Similarly, with the 30° laparoscope, the lowest concentration of ICG had a higher AFI at deeper depths than the other concentrations. Having knowledge of the patient’s albumin level may be helpful to determine appropriate dosing of ICG, as hypoalbuminemic patients may have a lower AFI due to decreased albumin available for ICG to bind to compared to patients with normal albumin levels.
In the present study, the effect of the angle at which the VITOM 3D or 30° laparoscope was held in relation to the ICG plates was evaluated. For both imaging devices, the AFI varied based on the angle. A previous study11 found that holding the 30° laparoscope at angles ranging from 60° to 90° did not influence AFI; however, an angle of 45° resulted in a decreased AFI for a range of concentrations of ICG. In the present study, a maximum AFI was achieved when the 30° laparoscope was held at 45° to the well plate, which made the lens surface of the laparoscope more perpendicular to the well plate, almost eliminating the 30° angle of the laparoscope itself. The VITOM 3D had a maximum AFI when held at 90°. A previous study11 did not find a substantial difference between AFI when using a handheld 0° NIRF camera (Fluobeam) for any angle except at 45°, which showed a lower AFI. Overall, the angle may affect the AFI, but the AFI at each angle was adequate to visually detect ICG. It may be ideal to hold the lens surface of the imaging device as close to perpendicular to the tissue of interest in vivo to maximize AFI, but a perfectly perpendicular angle is not required.
No difference in AFI was seen in the presence or absence of ambient light in the present study, although statistical analysis could not be performed. In a similar study11 using a handheld NIRF camera (Fluobeam), when the ambient light was on, the AFI was slightly higher as compared to the same conditions with the lights off. These differences in AFI may only be detectable with calculated measurements of AFI rather than with the naked eye. Additional studies could be performed in different levels of light and with more data points to allow for statistical analysis of this variable.
A major limitation of the present study was the ex vivo experimental design. The results of this experiment cannot be directly translated to the use of NIRF with ICG in veterinary patients. More specifically, an ex vivo model does not consider the interaction of ICG with live tissue or account for differences in tissue behavior among specific organs (eg, skin vs liver). However, the findings can facilitate future investigations into variables affecting ICG usage and assist surgeons with general guidelines to consider when working with NIRF with ICG.
In conclusion, this study aimed to test the effect of ICG concentration, distance, penetration depth, angle, and ambient light on AFI when using various imaging devices in a canine ex vivo model. Further studies are needed to objectively assess the relationship between distance, penetration depth, angle, and ambient light with AFI in vivo. The use of NIRF with ICG in veterinary medicine remains in its early stages, and further research is essential to fully explore its potential.
Supplementary Materials
Supplementary materials are posted online at the journal website: avmajournals.avma.org.
Acknowledgments
None reported.
Disclosures
The authors have nothing to disclose. No AI-assisted technologies were used in the composition of this manuscript.
Funding
The authors have nothing to disclose.
ORCID
Janet A. Grimes https://orcid.org/0000-0003-4554-8013
Mandy L. Wallace https://orcid.org/0000-0002-5286-4287
References
- 1.↑
Reinhart MB, Huntington CR, Blair LJ, Heniford BT, Augenstein VA. Indocyanine green: historical context, current applications, and future considerations. Surg Innov. 2016;23(2):166–175. doi:10.1177/1553350615604053
- 2.↑
Sakurai N, Ishigaki K, Terai K, et al. Impact of near-infrared fluorescence imaging with indocyanine green on the surgical treatment of pulmonary masses in dogs. Front Vet Sci. 2023;10:1018263. doi:10.3389/fvets.2023.1018263
- 3.↑
Sakurai N, Ishigaki K, Terai K, et al. Clinical impact of near-infrared fluorescence imaging with indocyanine green on surgical treatment for hepatic masses in dogs. BMC Vet Res. 2022;18(1):374. doi:10.1186/s12917-022-03467-2
- 4.
Eiger SN, Bertran J, Reynolds PS, et al. Use of near-infrared fluorescence angiography with indocyanine green to evaluate direct cutaneous arteries used for canine axial pattern flaps. Vet Surg. 2024;53(6):1073–1082. doi:10.1111/vsu.14121
- 5.↑
Larose PC, Brisson BA, Sanchez A, Monteith G, Singh A, Zhang M. Near-infrared fluorescence cholangiography in dogs: a pilot study. Vet Surg. 2024;53(4):659–670. doi:10.1111/vsu.14007
- 6.↑
Michalik D, Nolff MC. Case report: indocyanine green-based angiography for real-time assessment of superficial brachialis axial pattern flap vascularization in two dogs. Front Vet Sci. 2022;9:859875. doi:10.3389/fvets.2022.859875
- 7.
Madajewski B, Judy BF, Mouchli A, et al. Intraoperative near-infrared imaging of surgical wounds after tumor resections can detect residual disease. Clin Cancer Res. 2012;18(20):5741–5751. doi:10.1158/1078-0432.CCR-12-1188
- 8.↑
Kudo H, Ishizawa T, Tani K, et al. Visualization of subcapsular hepatic malignancy by indocyanine-green fluorescence imaging during laparoscopic hepatectomy. Surg Endosc. 2014;28(8):2504–2508. doi:10.1007/s00464-014-3468-z
- 9.↑
Gioux S, Choi HS, Frangioni JV. Image-guided surgery using invisible near-infrared light: fundamentals of clinical translation. Mol Imaging. 2010;9(5):237–255. doi:10.2310/7290.2010.00034
- 10.↑
Iida G, Asano K, Seki M, et al. Intraoperative identification of canine hepatocellular carcinoma with indocyanine green fluorescent imaging. J Small Anim Pract. 2013;54(11):594–600. doi:10.1111/jsap.12148
- 11.↑
Pruimboom T, van Kuijk SMJ, Qiu SS, et al. Optimizing indocyanine green fluorescence angiography in reconstructive flap surgery: a systematic review and ex vivo experiments. Surg Innov. 2020;27(1):103–119. doi:10.1177/1553350619862097
- 12.↑
van den Bos J, Wieringa FP, Bouvy ND, Stassen LPS. Optimizing the image of fluorescence cholangiography using ICG: a systematic review and ex vivo experiments. Surg Endosc. 2018;32(12):4820–4832. doi:10.1007/s00464-018-6233-x
- 13.↑
Jonasson H, Anderson C, Saager R. Water and hemoglobin modulated gelatin-based phantoms to spectrally mimic inflamed tissue in the validation of biomedical techniques and the modeling of microdialysis data. J Biomed Opt. 2022;27(7):074712. doi:10.1117/1.JBO.27.7.074712
- 14.↑
Kono Y, Ishizawa T, Tani K, et al. Techniques of fluorescence cholangiography during laparoscopic cholecystectomy for better delineation of the bile duct anatomy. Medicine (Baltimore). 2015;94(25):e1005. doi:10.1097/MD.0000000000001005
