Luís Oliveira a,b , Maria Inês Carvalho c , Elisabete Nogueira a , Valery V. Tuchin d,e,f

1. COMPARISON BETWEEN OPTICAL MEASUREMENTS MADE FROM NATURAL AND FROZEN SAMPLES AT OPTICAL CLEARING Luís Oliveira a,b, Maria Inês Carvalho c, Elisabete Nogueira a,Valery V. Tuchind,e,f a DFI – Polytechnic of Porto, School of Engineering, Rua Dr. António Bernardino de Almeida, 431, 4200-072 Porto, Portugal; b PhD student at FEUP – University of Porto, Rua Dr. Roberto Frias, 4200-465 Porto, Portugal c DEEC/FEUP and INESC TEC, University of Porto, Rua Dr. Roberto Frias, 4200-465 Porto, Portugal d Research-Educational Institute of Optics and Biophotonics, Saratov State University, 83 Astrakhanskaya str., Saratov 410012, Russia; e Laboratory of Laser Diagnostics of Technical and Living Systems, Institute of Precise Mechanics and Control RAS, Saratov 410028, Russia; f Optoelectronics and Measurement Techniques Laboratory, P. O. Box 4500, University of Oulu, FIN-90014, Oulu, Finland Saratov Fall Meeting – 2012 September 25 – 28, 2012 Saratov, Russia

2. 1. Introduction To evaluate the differences between the optical clearing effects created in fresh and frozen muscle samples, we have performed several measurements in two independent studies. As before (see presentation “Optical measurements of rat muscle samples under treatment with ethylene glycol and glucose” from Saratov Fall Meeting 2012), we have selected muscle samples from the abdominal wall of the Wistar Han rat. From a single animal sacrifice, we have collected several samples with 0.5 mm thickness [1]. From this collection of samples, we have considered two groups: fresh and frozen. Using the fresh samples, we immediately performed a set of measurements to obtain the response to light stimulation. The so called frozen samples were kept at -20 ºC for 72 hours before being submitted to a similar study. In both cases the measurements were performed while samples were immersed in ethylene glycol which is a well known optical clearing agent [1] [2] [3] [4].

3. The results obtained from the measurements with the two sets of samples were submitted to comparison so we could evaluate differences between the optical clearing effects created and this comparison is presented here. To compare between the two cases, we have used the measurements of total transmittance, collimated transmittance, specular reflectance and total reflectance measured from the two classes of muscle samples under treatment with ethylene glycol. Additionally we have also used the calculated absorbance and diffuse reflectance. After comparing results from the two studies we can be sure that the fact of freezing the muscle samples from a period of 72 hours produces some changes in the natural responses of the muscle to light. Such changes can be identified both in transmittance and in reflectance measurements. In general, we observed that the optical clearing effect is similar between the two states of samples (fresh or frozen), but some particular differences were observed in the time dependence of some measurements that deserve our attention and explanation.

4. 2. Experimental method Apart from using fresh and frozen samples, the experimental methodology was the same used in the studies of the other presentation(see presentation “Optical measurements of rat muscle samples under treatment with ethylene glycol and glucose” from Saratov Fall Meeting 2012). In the present case, we have started our first study with the fresh samples. The samples were submitted to measurements of total transmittance and total reflectance with the integrating sphere and collimated transmittance and specular reflectance with the assemblies described in the above indicated presentation. While the fresh samples were being studied, the samples for the second study were kept frozen for 72 hours at -20 ºC. After this time period, the samples had defrost naturally immediately before being used in the same type of measurements. In both cases ethylene glycol was used as optical clearing agent to perform the studies.

5. 3. Experimental results In this chapter, and since the two studies were performed with ethylene glycol, we will consider each of the particular measurements and present the results obtained from the two types of samples side by side. This representation allows to perform adequate comparison and immediate identification of the differences between the two cases. We will start in the following subsection with the results of total transmittance, followed by the results of collimated transmittance, total reflectance, specular reflectance, absorbance and diffuse reflectance. In each type of measurement, total transmittance, collimated transmittance, total reflectance and specular reflectance, we have used the same measuring and calculating methodology as in the other presentation (see presentation “Optical measurements of rat muscle samples under treatment with ethylene glycol and glucose” from Saratov Fall Meeting 2012). We will now describe the calculations to obtain results for each particular case.

6. For each particular measurement assembly, we measured the light reference spectrum and the spectra for natural tissue and while under optical clearing treatment using the Scan mode of the spectrometer. Considering that the light reference is 100% of light used in each experiment, we have calculated from the measurements the total transmittance, total reflectance, collimated transmittance and specular reflectance spectra by using equations 1, 2, 3 and 4 below. The results obtained with these equations are presented in percentage of the light reference used in each case. All these calculations are explained next: • For the case of total transmittance, we consider Stt(λ) as the spectrum of the light reference and Rtt(λ,t) as the total transmitted spectrum measured from the sample at a time t of the treatment. Using these symbols we can present equation 1 that is used to calculate the total transmittance spectrum at any time t: (1)

7. In the case of collimated transmittance, we consider Stc(λ) as the spectrum of the light reference and Rtc(λ,t) as the collimated transmitted spectrum measured from the sample at a time t of the treatment. Equation 2 allows the calculation of the collimated transmittance spectrum of the sample at time t: (2) • To calculate the total reflectance spectrum of the sample at a time t of the treatment, Srt(λ) is the measured spectrum from the light reference and Rrt(λ,t) is the total reflected spectrum measured from the sample at a time t of the treatment. Equation 3 shows this calculation: (3)

8. In a similar manner, equation 4 shows the calculation of the specular reflectance spectrum at a time t that uses the measurements of the source spectrum (Ssr(λ)) and the specular reflected spectrum (Rsr(λ,t) ) from the sample at time t: (4) • Using the experimental measurements we can also calculate the absorbance and diffuse reflectance of the tissue. • Considering the integrated measurements of total transmittance and total reflectance presented in equations 1 and 3 and assuming once again that the light spectrum that is used in the integrated measurements is 100%, the absorbance of the sample can be determined as the difference between 100% of light used and the integrated measurements [5]. • Equation 5 shows the calculation of sample’s absorption spectrum at a time t during the optical clearing treatment: (5)

9. On the other hand, we know that total reflectance contains both specular and diffuse terms. This way, using the measurements of total and specular reflectance (performed in the same conditions), we can use equation 6 to calculate diffuse reflection of the sample at time t: (6) • The results presented in all graphics below were calculated using all the equations above. • As mentioned above, we will present in the sub-sections below both the results for fresh and frozen samples. In each particular sub-section there are 4 figures for each type of sample (fresh or frozen). For the fresh samples, figures will always be presented on the left, while for frozen samples, figures will always be presented on the right. We will now make a brief explanation of each of these figures.

10. For each of the two cases (fresh or frozen), the 4 figures will be accordingly to the following description: • In the first figure we present the corresponding spectrum for the “natural” sample (before adding the optical clearing agent). • The second figure shows the time dependence lines for some particular wavelengths during the first minute of treatment. These lines have 1 second resolution. The chosen wavelengths were 400 nm, 500 nm, 600 nm, 700 nm, 800 nm, 900 nm and 1000 nm. • Similarly to the second figure, the third figure shows the time dependence curves for the same particular wavelengths (400 nm ….. 1000 nm), but now for 30 minutes of treatment and with one minute resolution. • The fourth figure shows the spectral variation during the first 15 minutes of treatment. The sequential spectra represented in this figure are separated by 1 minute. Natural spectrum is represented as a thicker line and viewing point may vary from case to case to optimize the identification of time variations.

11. This set of figures allows immediate identification of the “natural” characteristics of the muscle’s response to light stimulations. • On the other hand, the time dependence caused in the muscle’s response to light can also be identified easily. • By comparing between side-by-side figures, we can evaluate the differences between fresh and frozen samples. • By comparing between the two cases, we can also identify the quality of the optical clearing effect created by ethylene glycol and identify which case is better. • As we shall see by analyzing the figures below, both cases present advantages. Generally, the frozen samples show higher magnitude effect, but its time dependence is not so good as in the case of fresh samples. These facts are certainly related with the fact of samples being frozen for 72 hours and consequently the optical clearing effect is not so straight-forward as in the case of fresh samples.

12. 3.1 Total transmittance The total transmittance spectrum obtained from the sample before applying the immersion treatment in ethylene glycol can be seen in figure 1 (fresh sample) and in figure 2 (frozen sample): No absorption band seen in frozen sample at 520 nm. Different levels of total transmittance. Figure 1: Total transmittance spectrum from natural fresh muscle. Figure 2: Total transmittance spectrum from natural frozen muscle. We can see that in general the total transmittance spectrum is similar between both cases, but some differences can be identified.

13. Figure 3: Total transmittance evolution in the first minute of treatment (fresh sample). Figure 4: Total transmittance evolution in the first minute of treatment (frozen sample). From figures 5 and 6, we see that the frozen sample shows a little decreasing behavior in saturation regime. Also, we can see different levels of total transmittance between cases. Comparing figures 3 and 4, we see that the frozen sample presents a peak in total transmittance at 7 seconds and it shows a smaller rise than the fresh sample during the 1st minute of treatment with ethylene glycol. Figure 5: Total transmittance evolution during the treatment (fresh sample). Figure 6: Total transmittance evolution during the treatment (frozen sample).

14. Figure 7: Total transmittance spectral evolution in the first 15 minutes (fresh sample). Figure 8: Total transmittance spectral evolution in the first 15 minutes (frozen sample). By analyzing the spectral evolution during the first 15 minutes of treatment (figures 7 and 8), we see that in the case of the fresh sample, the total transmittance spectrum tends to grow during this entire period, while in the case of frozen sample only an initial increase is verified in the first minute of treatment. After the first minute, the spectrum of the frozen sample tends to remain constant.

15. In figure 9 (fresh) and figure 10 (frozen) we can see the collimated transmittance spectrum for the sample before applying the immersion treatment with ethylene glycol. 3.2 Collimated transmittance We observe similarity in the form of the spectra between the two cases. The case of frozen sample presents higher collimated transmittance. Figure 9: Collimated transmittance spectrum from natural muscle (fresh sample). The absorption peak at lower wavelengths shows different magnitudes in both cases. Figure 10: Collimated transmittance spectrum from natural muscle (frozen sample).

16. Figure 11: Collimated transmittance evolution in the first minute of treatment (fresh sample). Figure 12: Collimated transmittance evolution in the first minute of treatment (frozen sample). From figures 13 and 14, we see different behavior between cases along the treatment period. Frozen sample shows a more rapid increase in Tc in the first 4 or 5 minutes. Then it lowers a little before entering an almost constant saturation regime. From figures 11 and 12, we see the similarity in the behavior of the optical clearing effect during the first minute. Frozen sample has higher magnitude of collimated transmittance. Figure 13: Collimated transmittance evolution during the treatment (fresh sample). Figure 14: Collimated transmittance evolution during the treatment (frozen sample).

17. Figure 15 shows the time variation of the collimated transmittance spectrum of the fresh sample during the first 15 minutes of treatment with ethylene glycol. In this case we see that the spectrum always rises during this period (especially for longer wavelengths). Figure 15: Collimated transmittance spectral evolution in the first 15 minutes (fresh sample). In the case of the frozen sample (figure 16), we verify that the Tc spectrum only rises in the first 3 or 4 minutes. After that it shows a little decreasing tendency. Figure 16: Collimated transmittance spectral evolution in the first 15 minutes (frozen sample).

18. 3.3 Total reflectance Figure 17 and 18 show the total reflectance (Rt) spectrum of the natural samples (fresh and frozen, respectively). In this case, we see that the fresh sample has a higher Rt spectrum than the frozen sample. Figure 17: Total reflectance spectrum from natural muscle (fresh sample). The magnitude of the absorption peak at lower wavelengths is different between cases. Figure 18: Total reflectance spectrum from natural muscle (frozen sample).

19. Figure 19: Total reflectance evolution in the first minute of treatment (fresh sample). Figure 20: Total reflectance evolution in the first minute of treatment (frozen sample). Apart from the different levels of Rt verified between cases, the similarity of behavior is evident. Even regarding the oscillations, they are observed with great similarity between the two studies, during the treatment with ethylene glycol. Figure 22: Total reflectance evolution during the treatment (frozen sample). Figure 21: Total reflectance evolution during the treatment (fresh sample).

20. Figure 23 shows the time variation of the total reflectance spectrum of the fresh sample during the first 15 minutes of treatment. Figure 24 shows the same variation for the case of the frozen sample. Figure 23: Total reflectance spectral evolution in the first 15 minutes (fresh sample). In both cases the Rt spectrum lowers significantly only in the first minute. After that it tends to stay at the same level. In the case of the frozen sample, spectra appear to be more constant along the wavelength scale. This is only due to the point of view of the graph, which is different from the case of figure 23. Figure 24: Total reflectance spectral evolution in the first 15 minutes (frozen sample).

21. 3.4 Specular reflectance Figure 25 shows the “natural” specular reflectance spectrum of the fresh sample. In figure 26 we see the natural specular reflectance for the frozen sample. Figure 25: Specular reflectance spectrum from natural muscle (fresh sample). The frozen sample shows smaller levels of specular reflectance for all wavelengths. The form of the specular reflectance spectrum is almost equal between the two cases. The only difference is seen for intermediary wavelengths. Figure 26: Specular reflectance spectrum from natural muscle (frozen sample).

22. Figure 27: Specular reflectance evolution in the first minute of treatment (fresh sample). Figure 28: Specular reflectance evolution in the first minute of treatment (frozen sample). From figures 27 and 28, we see that the frozen sample produces some oscillations in the first seconds of treatment. Also the reduction in specular reflectance is smaller in the frozen sample. Considering figures 29 and 30, we see some similarity in behavior, but the frozen sample presents oscillations along the treatment with ethylene glycol. Figure 29: Specular reflectance evolution during the treatment (fresh sample). Figure 30: Specular reflectance evolution during the treatment (frozen sample).

23. Now considering the time dependence of the specular reflectance we have in figure 31 the case of the fresh sample and the case of the frozen sample in figure 32. Figure 31: Specular reflectance spectral evolution in the first 15 minutes (fresh sample). Once again we see variations with different levels of magnitude for the optical clearing effects created in the two samples. Apart from the different magnitudes, the time dependence of the specular reflectance seems very similar between the two cases. Figure 32: Specular reflectance spectral evolution in the first 15 minutes (frozen sample).

24. 3.5 Absorbance Figure 33 shows the “natural” absorbance spectrum of the fresh sample. In figure 26 we see the “natural” absorbance spectrum of the frozen sample. Figure 33: Absorbance spectrum from natural muscle (fresh sample). The absorbance levels are different , as we can verify by comparing figures 33 and 34. The spectral form is similar, but the frozen sample does not show absorbance peak around 520 nm. Figure 34: Absorbance spectrum from natural muscle (frozen sample).

25. Figure 35: Absorbance evolution in the first minute of treatment (fresh sample). Figure 36: Absorbance evolution in the first minute of treatment (frozen sample). Comparing between figures 37 and 38, we observe that the absorbance shows different saturation regimes when treated with ethylene glycol: in the case of the fresh sample it is almost constant, but in the case of the frozen sample it is increasing with time. From figures 35 and 36, we see that only the frozen sample shows an oscillation around 7 seconds. Additionally, the absorbance levels are smaller for the frozen sample and the decreasing behavior is more intense in the case of the fresh sample. Figure 38: Absorbance evolution during the treatment (frozen sample). Figure 37: Absorbance evolution during the treatment (fresh sample).

26. Figure 39 shows that the spectral evolution for the fresh sample is always decreasing in the first 15 minutes. For the case of the frozen sample, figure 40 indicates that the major decrease in absorbance is verified in the first minutes of treatment. After approximately 7 minutes, we see a little increasing behavior in the spectral evolution. Figure 39: Absorbance spectral evolution in the first 15 minutes (fresh samples). Figure 40: Absorbance spectral evolution in the first 15 minutes (frozen samples).

27. To finalize this comparative study, we present the calculated diffuse reflectance. Figure 41 shows the natural diffuse reflectance spectrum for the fresh sample and figure 42 shows the natural diffuse reflectance spectrum for the frozen sample. 3.6 Diffuse reflectance We can see a similar spectral form between the two cases, but there are some particular differences. The diffuse reflectance levels are different. Figure 41: Diffuse reflectance spectrum from natural muscle (fresh sample). Different magnitude peaks at lower wavelengths. Some spectral oscillations at middle wavelengths for the frozen sample. Figure 42: Diffuse reflectance spectrum from natural muscle (frozen sample).

28. Figure 43: Diffuse reflectance evolution in the first minute of treatment (fresh sample). Figure 44: Diffuse reflectance evolution in the first minute of treatment (frozen sample). In the first minute of treatment with ethylene glycol (figures 43 and 44), we see that the fresh sample shows an almost constant behavior after the 4th minute (well above natural values), which is not verified for the frozen sample. From figures 45 and 46 we can compare between the two cases for the long time period. We see that both cases show decreasing behavior, but the frozen sample presents oscillations, which is not the case for the fresh sample. Figure 46: Diffuse reflectance evolution during the treatment (frozen sample). Figure 45: Diffuse reflectance evolution during the treatment (fresh sample).

29. From figure 47, we see that the spectral evolution for the fresh sample shows only a strong increase during the first minute and then it tends to remain constant. For the case of the frozen sample, figure 48 shows also that the major increase is verified during the first minute of treatment. After that it tends to remain also constant, but some oscillations are seen during the first 15 minutes. Figure 47: Diffuse reflectance spectral evolution in the first 15 minutes (fresh sample). Figure 48: Diffuse reflectance spectral evolution in the first 15 minutes (frozen sample).

30. 4. Conclusions We have used fresh and frozen samples from the abdominal wall muscle of rat to study and compare the optical responses of those samples to light stimulation during treatment with ethylene glycol. We have verified that the frozen sample presents higher levels of total and collimated transmittance and also smaller levels of reflectance (total, specular and diffuse) and absorbance than the fresh sample. On the other hand, if we consider the natural spectra for total transmittance, collimated transmittance, total reflectance and absorbance, we see that the fact of freezing the sample for 72 hours removes some spectral signatures (absorption band around 520 nm). In the case of the natural specular and diffuse reflectance spectra, we see the appearance of some new spectral signatures in the case of the frozen sample that were not visible for the fresh sample. Considering the optical clearing effects created by ethylene glycol presented for both cases, some significant differences could be identified. In opposition to the case of the fresh sample, we have registered the appearance of oscillations in the first seconds of treatment for the frozen sample in the cases of total transmittance, collimated transmittance, specular reflectance, absorbance and diffuse reflectance. In the case of total reflectance, oscillations appear in both cases studied (fresh and frozen). If we consider the total time of treatment, we observe great similarity between the two cases.

31. One of the differences observed is that due to higher levels of transmittance (total and collimated) and lower levels of reflectance (total, specular and diffuse) for the frozen sample, the optical clearing effect created by immersing the tissue in ethylene glycol shows higher levels of transmittance and lower levels of reflectance along the time for the frozen sample. In the case of absorbance, we have observed that the fresh sample presents a decreasing absorbance during all treatment time, while the frozen sample presents an initial decrease and an increasing saturation regime after approximately 6 or 7 minutes. By analyzing the comparisons made in the different measurements between fresh and frozen samples, we can recognize that the freezing process is responsible for significant changes in the interaction of the muscle with light. One of the changes is that for some of the measurements presented we see differences between the “natural” spectra of fresh and frozen samples. On the other hand, in some of the cases presented, we could see the appearance of oscillations in the case of the frozen sample during the optical clearing treatment with ethylene glycol.

32. Acknowledgements The authors would like to thank the following institutions for all the help in preparation of the samples and resources made available to perform the measurements: • CIETI – Centro de Inovação em Engenharia e Tecnologia Industrial, ISEP – Instituto Superior de Engenharia do Porto, Portugal. • LAIMM – Laboratório de Apoio à Investigação em Medicina Molecular, Departamento de Biologia Experimental, Faculdade de Medicina da Universidade do Porto, Portugal. • This work was also supported in part by grants: • 224014 PHOTONICS4LIFE of FP7-ICT-2007-2, 1.4.09 of RF Ministry of Education and Science; • RF Governmental contracts 02.740.11.0770, 02.740.11.0879, and 11.519.11.2035; • FiDiPro, TEKES Program (40111/11), Finland; • SCOPES EC, Uzb/Switz/RF, Swiss NSF, IZ74ZO_137423/1; • RF President’s grant “Scientific Schools”, 1177.2012.2.

33. References [1] Oliveira, L., [Study of the spectral transmission response of biological tissues under the influence of different osmoticagents], MsCthesis, FEUP – Faculdade de Engenharia da Universidade do Porto, Porto, Portugal (2007). [2] Oliveira, L. M., Lage, A. S., PaisClemente M. and Tuchin, V. V., “Concentration dependence of the optical clearing effect created in muscle immersed in glycerol and ethylene glycol” SPIE Proceedings of Saratov Fall Meeting – 2006, Vol 6535, V. V. Tuchin Ed. (2007). [3] Tuchin, V. V., [Optical Clearing of Tissues and Blood], SPIE Press, Bellingham, WA (2006). [4] Hirshburg, J., Choi, B., Nelson, J. S., Yeh, A. T., “Collagen solubility correlates with skin optical clearing”, Journal of Biomedical Optics 11(4), (2006). [5] Bass, M., Mahajan, V. N., Van Stryland, E., [Handbook of Optics – Vol. II], McGraw Hill, (2010).

34. Thank you for your attention.