Noninvasive Determination of Depth in Transmission Raman

Subscriber access provided by UNIVERSITY OF ADELAIDE LIBRARIES

Article

Non-invasive Determination of Depth in Transmission Raman Spectroscopy in Turbid Media based on Sample Differential Transmittance Benjamin Gardner, Nicholas Stone, and Pavel Matousek Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b01469 • Publication Date (Web): 21 Aug 2017 Downloaded from http://pubs.acs.org on August 23, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 11

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Non-invasive Determination of Depth in Transmission Raman Spectroscopy in Turbid Media based on Sample Differential Transmittance Benjamin Gardner1, Nicholas Stone1*, Pavel Matousek2*

1) School of Physics and Astronomy, University of Exeter, Exeter EX4 4QL, UK 2) Central Laser Facility, Research Complex at Harwell, STFC Rutherford Appleton Laboratory, Harwell Oxford, OX11 0QX, UK

*Joint corresponding authors: [email protected] [email protected]

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract Here we propose a simple non-invasive approach to determine the depth of a buried object using transmission Raman spectroscopy. In accordance with theory, the photons arising from spectral peaks that are suitably separated will be subjected to different optical properties in the media through which they travel. These differences can impact the relative intensities of Raman peaks as a function of the transmission path length, thereby the depth of signal generation is inherently encoded in the spectra. In a proof-of-concept study, through only external calibrations, it was possible to accurately predict the depth of Polytetrafluoroethylene (PTFE) layer purely on the basis of relative intensity of two peaks in a predominantly absorbing solution Indian ink (0.1 µl /ml) (RMSE 0.42 mm) and a scattering solution (RMSE 0.50 mm). This simple approach offers the possibility to non-invasively identify the depth of a buried object, such as breast calcifications, using simple transmission measurement geometries for the first time.


Page 2 of 11

Page 3 of 11

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Introduction The recent development of deep Raman spectroscopy, an array of methods for subsurface analysis of turbid media, has opened a number of analytical fields including pharmaceutical analysis in quality control and disease diagnosis 1. The bedrocks of these methods are Spatially Offset Raman Spectroscopy (SORS)

2

and Transmission Raman Spectroscopy

(TRS) 3,4. The SORS approach relies on spatially separating the collection Raman zone from the laser illumination zone on sample surface. The spatially offset Raman spectra contain varying relative contributions from different sample depths. As such the method enables the pure chemical makeup of a specific layer as well its depth to be determined. In contrast TRS, which is performed by illuminating one side of object and collecting Raman signal from the opposing side, renders information on the chemical composition of the entire probed volume. This leads to both a lack of depth information as well as preventing recovery of any pure specific chemical constituents. This is due to the fact that only a single measurement is performed and the Raman signals from individual depths or layers are mixed up (averaged) in the collected Raman spectrum. TRS, with its deeper probing capability compared to SORS (approximately double), is being explored for a number of important applications including breast cancer diagnosis5,6 where, for example the determination of depth / location of detected calcifications would be a very useful additional diagnostic feature if viable. In recent years, a more complex TRS concept based around tomographic imaging has been proposed and demonstrated that can render the depth information of a target object within a turbid matrix 7. However this concept, which involves complex reconstruction algorithms and data derived from measurements performed in multiple sample orientations, is time consuming and not readily applied. Other recent innovations to recover depth information



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

using TRS involved the addition of optical elements (band pass filter) at the sample surface acting as a “photon dioide”8. Here we proposed a much simpler and more broadly applicable method enabling the depth of a single, chemically distinct object buried within a turbid matrix to be determined through monitoring the differential transmittance of two or more discrete Raman bands. It is well established that photons of different wavelengths are influenced by absorption (µa) (along with scattering (µs) and anisotropy (g)) to potentially differing degrees. Where such ‘differential transmittance’ is present the propagation distance of Raman photons in an absorbing or scattering matrix can be monitored through differences in the relative intensities in the Raman peaks of the target material. In TRS measurements laser photons are first delivered to a chemically distinct inclusion (e.g. calcification in breast tissue) buried in a matrix (surrounding breast soft tissue). Raman photons (consisting of multiple Raman bands at different wavelengths) are then subject to mutually differential transmittance, where present, when they propagate to the other side of sample where they can be detected. As the distance between the Raman creation point and the detection point increases, the more apparent differential transmittance becomes with respect to a non-buried standard measurement. Naturally, the Raman signal propagation distance is larger if the inclusion is closer to the illumination side and smaller if it is closer to the Raman collection side of sample. As such the depth of the inclusion is encoded in the Raman signal through matrix differential transmittance. This ‘depth information’ can be then ‘read’ non-invasively from the detected TRS signal.


Page 4 of 11

Page 5 of 11

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Experimental Two different media were used to observe differential transmittance: a predominantly absorbing, non-scattering medium containing Indian ink (Histology Stain, American MasterTech) at 0.1 µl ink / ml (~µa 0.37 cm-1) in water and a predominantly scattering Intralipid solution (Sigma Aldrich) at 0.5% (~µs' 4.6 cm-1). Experiments looking at differential transmittance of the Raman signal of Polytetrafluoroethylene (PTFE, 50×23×3 mm) through water and Indian ink, were carried out in a 45×45×50 mm quartz cell (external dimensions), with the total internal path length [46 mm]. The PTFE sample was aligned centrally to the X-axis, and moved along the Z-axis in [1 mm] steps [~40 steps] by a motorized translation stage (Standa Ltd). Experiments started with the sample in position (z=0) which corresponds to the sample being closest to the Raman collection optics. As the z distance increases so does the path length of the collected Raman photons. A quartz cell with a smaller path length was used when the scattering agent was used (45×45×30 mm), with the total internal path length (~24 mm), the PTFE was moved in 0.5 mm Steps (~56). Experiments were carried out on a previously described home built Raman system operating in transmission mode9 (Figure 1). The excitation wavelength was 830 nm (Innovative Photonic Solutions: I0830MM0350MF-EM) delivering ~400 mW at the sample surface (~2 mm). Raman spectra were collected from the other side of sample from an area of approximately 2 mm diameter and relied via optical system to a Kaiser spectrometer (Holospec 1.8i) with a deep depletion CCD camera (Andor iDus-420). Raman spectra were measured for 5s × 60 accumulations in all experiments at each depth.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 1| Experimental setup of the near IR transmission Raman system. Results and Discussion As is shown (Figure 2A), for two Raman peaks that are well separated spectrally, e.g. PTFE Raman bands at 290 cm-1 and 1480 cm-1, there is a large corresponding difference in the wavelength of the Raman photons that are generated ~90 nm. When these photons travel through a medium that is minimally absorbing or scattering such as air, there should be no correlation between Raman creation point and collection, as is seen (Figure 2B). However, when the Raman photons travel through a medium where there are changes in absorption or scattering with wavelength, relative differences in intensity can be observed over significant distances (Figure 2C). As is shown in Figure 2C and D, when z=0 mm i.e. the sample is at the back of the cell wall (Raman collection side) there is a minimal travel distance for the Raman photons and the relative intensities of the spectrum are not distorted. However, as the migration distance of the Raman photons increases, not only is there a decrease in absolute Raman intensity, there are also changes in the relative intensity of the two chosen Raman bands (290 cm-1 and 1380 cm-1). This difference in relative intensities can be seen as an approximately linear problem when the natural logarithm of the ratio of the intensities of the two bands (1380 / 290 cm-1) is plotted against the z distance (Figure 2D).


Page 6 of 11

Page 7 of 11

2500

2000

Z = 0 mm Z = 42 mm

*

~87 nm

1500

*

1000

B|

-0.662 -0.664

Log Peak Ratio

Raman Intensity [a.u.]

A|

-0.666 -0.668 -0.67 -0.672

500 200

400

600

800

1000

1200

-0.674 0

1400

Wavenumber [cm-1]

10

20

30

40

50

Z [mm]

C| 5000

-1

1380 cm -1 290 cm

D| ln Peak Ratio

4000

Raman Intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


3000

2000

1000

0 0

5

10

15

20

25

30

35

40

Z [mm]

Figure 2| A, Raman Spectra of PTFE measured at two separate depths in 0.1 µl / ml Indian ink. B, Natural logarithm of the Raman PTFE peak ratio (1380 / 290 cm-1) measured through air (no solution in the cell). C, Raman intensities of PTFE peaks (1380 cm-1 blue, 290 cm-1 red) at depth in Indian ink solution (0.1 µl / ml water). D, Natural logarithm of the Raman PTFE peak ratio (1380 / 290 cm-1) measured through Indian ink solution. Through a leave one out (1 series of four) cross validation linear model, it was possible to accurately predict the depth of PTFE with a root mean square error (RMSE) of 0.21 mm (Figure 3A). However, in real applications this would have a more limited use, as it has been trained with prior information on PTFE locations to build the model. A more desirable solution would be to be able to predict depth with no true prior information regarding the depth of the target object i.e. PTFE. Figure 3B shows the log peak ratio of PTFE from the internal measurements (blue) and two external measurements of PTFE (red). The first external PTFE measurement is a standard non-attenuated measurement with sample located



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

externally at the Raman collection side of the cell, while the second is PTFE measured in transmission externally to the quartz cell containing Indian ink with sample located externally at the laser illumination side of the cell, with the Raman photons traveling the entire path length through the cell filled with solution. An alternative method for depth calibration could be provided by comparing the spectrum of an atomic emission lamp source (sharp spectral features across the range of interest) with the resulting spectrum after transmission through the sample. From the two external measurements of PTFE it was possible to construct a linear prediction model of the measurements with a RMSE of 0.44 mm (Figure 3C). Furthermore, if we consider the overall sample thickness of 46 mm, the equivalent percentage RMSE is 0.9%.

Figure 3| A, Leave one out cross validation of 4 repeat experiments predicting the depth of PTFE. B, Log Peak Ratio of PTFE in 0.1 µl / ml Indian ink (Blue) and PTFE measured alone and externally in transmission through the entire quartz cell (Red). C, prediction of PTFE position based on the two external measurements.


Page 8 of 11

Page 9 of 11

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


A|

Map 1 Map 2 Map 3

-0.6

B|

-0.8

-1

-1.2

-1.4

-1.6 0

5

10

Z [mm]

15

20

C|

Figure 4| A, The natural logarithm of the PTFE Raman peak ratio (1380 / 290 cm-1) at depth in 0.5% Intralipid solution. B, Leave one out cross validation of the three repeat experiments predicting the depth of PTFE, RMSE = 0.52 mm. C, Prediction of PTFE depth based on two external PTFE measurements, RMSE = 0.50 mm.

To further explore this approach, of determining inclusion depth using transmission Raman spectroscopy, a predominantly scattering media was used (0.5 % intralipid). Figure 4A shows a mostly linear relationship is still observed with the log of the ratio of peak intensities versus depth. The leave one out analysis achieved a RMSE of 0.52 mm (Figure 4B). While using PTFE measured externally as described previously, a RMSE of 0.50 mm was obtained (Figure 4C) a slight improvement on the leave one out model. This equates to a relative depth resolution of 2% of the overall sample thickness.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

In the presence of turbid media, the dependence of the logarithm of the peak ratio from the inclusion of interest becomes a non-linear function of depth, z. This implies that the existing (linear) regression models could potentially further be improved by using non-linear multivariate methods. Furthermore, any sufficiently strong source of photons, with wavelengths similar to the Raman peaks of interest of the inclusion, could be used to provide the external depth calibration. This is in effect measuring the relative attenuation profile for the peaks of interest.

Conclusions In conclusion, the presented concept provides a simple means of deriving depth of a single buried object in a diffusely scattering matrix in TRS measurements. Examples of applications might include the non-invasive determination of the depth of calcifications in breast tissue along with their chemical makeup; depth of active pharmaceutical ingredients in tablets or the multiplexed depth localisation of labelled surface enhanced nanoparticles as an extension of SESORS 10,11.

Acknowledgments EPSRC grant (EP/K020374/1) funded the work presented here. References (1)

Matousek, P.; Stone, N. Chem. Soc. Rev. 2016, 45 (7), 1794–1802.

(2)

Matousek, P.; Clark, I. P.; Draper, E. R. C.; Morris, M. D.; Goodship, A. E.; Everall,


Page 10 of 11

Page 11 of 11

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


N.; Towrie, M.; Finney, W. F.; Parker, A. W. Appl. Spectrosc. 2005, 59 (4), 393–400. (3)

Schrader, B.; Bergmann, G. Fresenius’ Zeitschrift für Anal. Chemie 1967, 225 (2), 230–247.

(4)

Matousek, P.; Morris, M. D. Emerging Raman Applications and Techniques in

Biomedical and Pharmaceutical Fields; Matousek, P., Morris, M. D., Eds.; Biological and Medical Physics, Biomedical Engineering; Springer Berlin Heidelberg: Berlin, Heidelberg, 2010. (5)

Stone, N.; Matousek, P. Cancer Res. 2008, 68 (11), 4424 LP-4430.

(6)

Matousek, P.; Stone, N. J. Biomed. Opt. 2007, 12 (2), 24008.

(7)

Srinivasan, S.; Schulmerich, M.; Cole, J. H.; Dooley, K. A.; Kreider, J. M.; Pogue, B. W.; Morris, M. D.; Goldstein, S. A. Opt. Express 2008, 16 (16), 12190–12200.

(8)

Vardaki, M. Z.; Sheridan, H.; Stone, N.; Matousek, P. Appl. Spectrosc. 2017, 3702817691540.

(9)

Vardaki, M. Z.; Gardner, B.; Stone, N.; Matousek, P. Analyst 2015, 140 (15), 5112– 5119.

(10)

Sharma, B.; Ma, K.; Glucksberg, M. R.; Van Duyne, R. P. J. Am. Chem. Soc. 2013,

135 (46), 17290–17293. (11)

Stone, N.; Kerssens, M.; Lloyd, G. R.; Faulds, K.; Graham, D.; Matousek, P. Chem.

Sci. 2011, 2 (4), 776–780.


Noninvasive Determination of Depth in Transmission Raman

Recommend Documents