Inquiry on the extraction of optical properties for different port sizes using the integration sphere combined with Inverse Adding-Doubling

[xiaoxuinchina]

Dear Professor Scott Prahl,

I am Chengli Xu, a graduate student from the School of Medical Technology at the Beijing Institute of Technology in China. I am writing to seek your expertise in a matter related to our ongoing research.

As we embark on our project to extract tissue optical properties from small-sized samples, we have encountered several challenges and questions during our experimentation. We have been utilizing various ports of an integrating sphere with diameters of 25.4 mm, 12.7 mm, and 10 mm to test a 10% fat emulsion. Our primary goal is to measure the total transmittance and reflectance of the same sample through these different ports and subsequently apply the Inverse Adding-Doubling (IAD) methodology to determine the optical properties of the tissue.

However, our results indicate significant variations in the optical properties of the same sample tissue when measured through the different ports, particularly in the reduced scattering coefficient (as evident in Figure 1, where we compare the reduced scattering coefficient of 10% Intralipid measured through the various ports using IAD-3.12 and 3.16 versions).

Additionally, we have observed discrepancies in the tissue optical properties calculated by reflectance and transmittance using the same set of parameters between iad-win-3-12-0 and iad-win-3-16-2. The overlapping sections of the curves in Figures 2 and 3, for instance, exhibit inconsistent code display statuses. IAD-3.12 indicates "*" for normal status, whereas IAD-3.16 displays "R," suggesting excessively high reflectance. We are curious to understand why the same data set exhibits different statuses across different codes.

Furthermore, in the calculation results for 20% Intralipid, we find that the IAD results seem to deviate significantly from Mie theory predictions, particularly in the IAD-3.12 version, where the deviation is more pronounced. The results are not nearly double those obtained for 10% Intralipid, which is unexpected.

Given your expertise in this field, I would greatly appreciate your insights on how you addressed this issue in your subsequent research work. Any advice or suggestions would be invaluable to us in advancing our project.

Thank you for considering my request and for your time. I am eager to learn from your experiences and expertise.

All the best,

Chengli Xu

Appendix: Equipment used in this study: Double integrating sphere (RT-060-SF, Labsphere) system, reflectance calibration plate (SRS-99-010, Labsphere), sphere wall reflectance, reflectance and transmittance measurement procedures were carried out according to the "Everything I think you should know about Inverse Adding-Doubling" manual. Test sample used: 20% Intralipid(Fresenius Kabi). Fig. 1. The reduced scattering coefficient results of 10% Intralipid calculated at different ports and different IAD versions. Fig. 2. The reduced scattering coefficient results of 10% Intralipid calculated at different IAD versions. Fig. 3. The reduced scattering coefficient results of 20% Intralipid calculated at different IAD versions.

[scottprahl]

1) Measuring Intralipid is difficult because the absorption is so small (relative to scattering). Hopefully we can work together to sort out the inconsistencies that you have found.

2) The differences between v3.12 and v3.16 are the result of improved MC estimates for lost light. You will want to use v3.16.2 or later for the most correct implementation. (You can see this improvement because v3.16 gives slightly closer scattering values for the 12.7 and 25.4 port sizes.)

3) I would not worry about matching Mie estimates because there is a wide range of fat droplet sizes. One has to use an (expensive) monodisperse solution of latex or polystyrene microspheres as a "gold" standard.

4) As you mentioned, the 20% solution should be 2X the 10% solution and it is not. This is concerning.

5) To comment more, I would need to see your input (.rxt) files for the 10% measurements with 12.7 and 25.4mm ports. You can email them to me or add them here.

[xiaoxuinchina]

Thank you for your response. Since GitHub does not support uploading .rat format files, it has been converted to .txt for upload. I have uploaded here the input (.txt) files of my 10% Intralipid measurements using the 12.7 and 25.4 mm ports. 2024_05_28_10%_25.4.txt 2024_05_28_10%_12.7.txt Perhaps you would like me to also provide measurements for the 20% Intralipid with the 12.7 and 25.4mm ports?

[scottprahl]

Your 25.4 measurements are curious because MR+MT > 1 for all wavelengths longer than about 480nm. That is why the program kept emitting 'R' for the measurements.

Another detail is that the index of refraction of Intralipid should be essentially that of water. Making this change does not make a significant change.

I assume you made the same measurements as shown in figure 4 in the manual?

[xiaoxuinchina]

Thank you very much for your reply, it makes a lot of sense! I will pay attention to the issue of refractive index later. MR+MT > 1 is too bad. I wonder if it is caused by the port and thickness ratio, as shown in Figure 1; or if I put a cuvette filled with deionized water between the two balls during the measurement of T2(0,0,1,1), and measured T2(0,0,1,1), which led to a high transmittance and MR+MT > 1. I am eager for your valuable feedback! Fig. 1. The total collected light versus the optical thickness of the sample(From figure 7 in the manual).

[shanghongY]

Your 25.4 measurements are curious because MR+MT > 1 for all wavelengths longer than about 480nm. That is why the program kept emitting 'R' for the measurements.

Another detail is that the index of refraction of Intralipid should be essentially that of water. Making this change does not make a significant change.

I assume you made the same measurements as shown in figure 4 in the manual?

Dear professor, I have checked the measured MR and MT, but i could not find out why the program kept emitting '+' (did not converge )for the measurements using the latest version of the IAD program (iad 3-16-3).