Combined Video Analysis of ICG and 5-ALA Induced
Protoporphyrin IX and Hemoglobin Oxygen Saturation in near
Infrared
T. A. Savelieva
1,2
, D. M. Kustov
1
, P. V. Grachev
1,2
, V. I. Makarov
1
, E. E. Osipova
1
and
V. B. Loschenov
1,2
1
Prokhorov General Physics Institute of the Russian Academy of Sciences, Vavilov st., Moscow, Russia
2
National Research Nuclear University MEPhI, Kashirskoe sh., Moscow, Russia
Keywords: Indocyanine Green, Protoporphyrin IX, Haemoglobin Oxygenation, Fluorescence, Diffuse Reflectance.
Abstract: Due to the high recurrence rate after the glial brain tumor removal, methods of intraoperative navigation
have a high relevance, providing the most complete removal of tumor tissues with maximum preservation of
healthy ones. In this work a combined visualization method is proposed with an assessment of fluorescence
and diffuse reflectance images. Fluorescence intensity of 5-ALA-induced protoporphyrin IX allows
visualization of tumor cells, distribution of indocyanine green fluorescence helps to visualize the vascular
system of the tumor, and parallel mapping of the degree of oxygenation demonstrate the hypoxic regions.
The images were obtained in the near infrared range of the optical spectrum in order to maximize the optical
probing depth in the window of biological transparency.
1 INTRODUCTION
This paper presents a video analysis technique,
which allows parallel-sequential implementation of:
- simultaneous registration of fluorescent
intraoperative images of tumors and surrounding
tissues in two wavelengths, corresponding to the
fluorescence of protoporphyrin IX, which
characterizes the activity of metabolic processes in
cells, and indocyanine green fluorescence,
characterizing the parameters of blood supply to the
tissue with the display in the overlay mode of two
fluorescent patterns;
- registration of intraoperative images of tumors and
surrounding tissues in diffusively reflected light at
wavelengths characterizing the ratio of oxy- and
deoxyhemoglobin to assess the degree of hypoxia of
different sections of tumor tissue.
Currently there are no systems for simultaneous
visualization of fluorescence of indocyanine green
and protoporphyrin IX in clinical practice, however,
these modes are available sequentially in the Zeiss
Opmi Pentero microscope, but there are no oxygen
saturation mapping regime. The possibility of a
multi-factorial intraoperative assessment of the
tumor state can improve the accuracy of resection.
There are a number of works on simultaneous
registration of the photosensitizer fluorescence
distribution and the image of the same area in
diffusively reflected broadband light to observe the
fluorescent tissues on the full color background
image and thereby improve navigation (Sato et al.,
2015; Loshchenov et al., 2018). In other side the
hemoglobin oxygenation visualization of brain
tissues by optical methods are performed in two
ways as a rule. Oximetric video systems in the
visible or near infrared spectral range are based on
mathematical processing of full-color images
(Mustari et al., 2018) or hyperspectral imaging
(Giannoni et al., 2018). But using the near infrared
region for this purpose is limited as a rule by the
mapping of spectroscopic signals obtained from
specific points of the cortex projection from a set of
fiber optic probes, as in Hitachi, fNIR Devices LLC,
Biopac, Artinis systems for functional near infrared
spectroscopy. These systems are designed for
functional transcranial spectroscopy in the near
infrared, while proposed system is intended for
intraoperative imaging of the tumor bed. And it has
such an important advantage over visualization
systems in the visible range as work in the biological
transparency window. As it is well known, in near
320
Savelieva, T., Kustov, D., Grachev, P., Makarov, V., Osipova, E. and Loschenov, V.
Combined Video Analysis of ICG and 5-ALA Induced Protopor phyrin IX and Hemoglobin Oxygen Saturation in near Infrared.
DOI: 10.5220/0007693703200324
In Proceedings of the 7th International Conference on Photonics, Optics and Laser Technology (PHOTOPTICS 2019), pages 320-324
ISBN: 978-989-758-364-3
Copyright
c
2019 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
infrared range of optical spectrum the absorption of
main tissue chromophores is minimal so the light
can penetrate deeper in the tissue. But in this optical
range we should consider the prevalence of
scattering under the absorption in tissue.
2 MATERIALS AND METHODS
2.1 Device
The four laser light sources were used. The source
with 635 nm wavelength was used for PpIX
fluorescence excitation. The source with 785 nm
wavelength was used for ICG fluorescence
excitation. The 687 nm laser light source was used
for getting diffuse reflectance image in the position
where deoxygenated hemoglobin absorption
significantly higher than oxygenated hemoglobin
absorption. The source with 805 nm wavelength was
used to obtain diffuse reflectance image in the
position close to isosbestic point of deoxygenated
and oxygenated hemoglobin absorption spectra. Two
video cameras were used for parallel-serial data
reception in two modes: fluorescence and diffuse
reflectance image registration.
In fluorescence mode in front of the cameras were
installed such optical band-pass filters as 710/40
Semrock for Pp IX fluorescence registration and
832/37 Semrock for ICG fluorescence registration.
In diffuse reflectance mode in front of the cameras
were installed 685/40 Semrock and 810/10 Semrock
band-pass filters to eliminate cross light from these
spectrum ranges.
The control of oxygenation level was performed
with fiber-optic spectroscopy technique with LESA-
01-BIOSPEC using algorithm described earlier
(Stratonnikov et al., 2006). The control of
fluorescence intensity was carried out with the same
spectroscopic device in fluorescence mode.
2.2 Diffuse Reflectance Image
Processing Algorithm
Schmitt and Kumar, (Schmitt and Kumar, 1996;
Kumar and Schmitt, 1997), demonstrated the
dependence of the diffuse reflectance spectrum in
the near infrared region for tissue with different
optical properties on the distance between the
illuminating and receiving fibers, using the diffusion
approximation to approximate the obtained
dependences. For the fluorescence spectra, it is usual
to assume that measuring the diffuse reflectance
spectrum of tissues in the same range as
fluorescence one is sufficient to take into account
the effect of the sample geometry. However, as it
was shown in Gebhart et al., 2007, the spectra of
diffuse reflectance and fluorescence show significant
differences when registering by the method of point
to point spectroscopy in contact with the sample and
at a certain distance, as well as by the method of
spectrally resolved visualization. An increase in the
contrast between the red and blue-green ranges of
the diffuse reflectance spectrum was found for the
illumination-reception geometry implemented in the
visualization system and point spectroscopy at a
certain distance from the sample. This effect is most
likely due to the low contribution of multiply
scattered photons in case of proximity of light
source and receiver. Differences in the diffuse
reflectance spectrum from samples containing
hemoglobin in various concentrations were greater
for the imaging system compared to the point
spectroscopy system, but they had similar trends. At
the same time, the dependence of the diffuse
reflectance spectra of samples with different
contents of scatterers differed for the considered
geometries both in nature and in magnitude. In this
connection, in the present work, in order to consider
the influence of the measurement geometry, a
mathematical simulation was carried out for the
cases of the diffusely reflected and fluorescent
radiation registration at different distances between
the illumination and receiving fibers for different
numerical apertures.
2.3 Objects for Validation
Validation of proposed method of combined video
analysis and image processing algorithms was
carried out on an array of optical phantoms that
simulate the optical properties of nerve tissues
containing the studied photosensitizers with a
concentration 0.625 mg/l, 1.25 mg/l, 2.5 mg/l and 5
mg/l of indocyanine green and 1.25 mg/l , 2.5 mg/l,
5 mg/l and 10 mg/l of Pp IX. Photosensitizers were
diluted with the 1% fat emulsion solution as a
scattering agent to simulate multiple scattering
occurring in investigated tissues.
Phantoms were performed in the cuvettes. To
analyze their images in environments similar to the
natural brain, they were immersed in a scattering
gelatin matrix (5% gelatin with a 1% intralipid
solution).
Optical phantoms simulating biological tissue with
blood were also performed. Whole blood was
diluted 16 times so that the hemoglobin
concentration was approximately equal to the
Combined Video Analysis of ICG and 5-ALA Induced Protoporphyrin IX and Hemoglobin Oxygen Saturation in near Infrared
321
average concentration in the white matter of the
brain. Fat emulsion solution was used as a scattering
medium. The samples were deoxygenated through
nitrogen ventilation.
Testing of the developed method and system on
laboratory animals was also carried out in the
fluorescence mode (with simultaneous
administration of indocyanine green and
protoporphyrin IX) and in the diffuse reflectance
mode (to control the degree of oxygenation) before,
during and after deoxygenation of the animal’s
blood by nitrogen ventilation.
3 RESULTS AND DISCUSSION
3.1 Analysis of Co-distribution of ICG
and PP IX on Optical Phantoms
The results of the video system validation on optical
phantoms are shown in Figure 1 in the fluorescence
mode for fluorescence phantoms. In the left column
of images in black-white channel the laser light
source with 785nm wavelength for fluorescence
excitation and optical filter 832/37 nm for
fluorescence registration were used. In the right
column of images in color channel the laser light
source with 633 nm wavelength for fluorescence
excitation and optical filter 710/40 nm for
fluorescence registration were used. These two
pictures can be combined in one in special mode to
navigate during surgery.
Validation showed that the sensitivity of the
system in the fluorescence mode is no worse than
1.25 mg/l for indocyanine green detection and 0.125
mg/l for Pp IX detection without hardware signal
amplification in real time for fluorescent inclusions
in the scattering medium, whose transverse
dimensions do not exceed 1 mm.
3.2 Algorithm of Diffuse Reflectance
Image Processing with Separation
of Absorption and Scattering
Component
The diffuse reflectance spectra depends on scattering
as well as absorption of light by the tissue. If the
purpose of registering a diffuse reflectance signal is
to analyze the content of chromophores in a tissue
by their absorption, these two processes should be
separated. At the same time, information about
tissue scattering can be an independent diagnostic
criterion.
In a highly scattering medium, the radiation
corresponding to one pixel of the image is
characterized by a wide distribution of photon path
lengths in the tissue. Thus, the greater the scattering
compared to the absorption, the greater the influence
of the neighboring pixels on each other. In general
case, the solution for one pixel of the image is a
linear combination of solutions for a point source
and a receiver for the sum of possible photon paths
in the tissue.
As a result of numerical simulation of the
propagation of light in a high-scattering medium
with optical properties, which varied in a series of
physiologically relevant values, the following
approximating dependence of the diffuse reflection
signal (Rd) on absorption coefficient
a
) and
scattering coefficient (µ’
s
) was proposed:





(1)
Here, A
1
and B
i
are adjustable parameters which
depend on the illumination and signal detection
geometry, μ'S, reduced scattering coefficient defined
as μ'
s
= μ
s
·(1 g), g, scattering anisotropy factor,
and μа, absorption coefficient. Detailed description
of this approximation is given in earlier publication
(Savelieva et al., 2015).
R
d
is the signal that we receive from each camera
(on two wavelengths) in the diffuse reflection mode
of registration, taking into account the normalization
to the image of standard white illuminated with the
same broadband light source. Solving equation (1)
we get the values of the absorption coefficients on
chosen wavelengths.
The matrix for converting the concentrations of
tissue chromophores studied in this work to
absorption coefficients is as follows:













(2)
where µ
a
λi
absorption coefficient on
wavelength λ
i
, ε
j
λi
molar extinction coefficient of j
chromophore on wavelength λ
i
, C
j
the
concentration of j chromophore.
Solving system of equations (2) we get the
values of deoxygenated hemoglobin concentration
(

) and oxygenated haemoglobin concentration
(

). From these parameters we can calculate the
deoxygenation parameter as:





(3)
This parameter was introduced in this work for
more convenient observation and it is additional to
PHOTOPTICS 2019 - 7th International Conference on Photonics, Optics and Laser Technology
322
the level of hemoglobin oxygen saturation; in total,
they give 1.
On Figure 2 we can see two pictures in diffuse
reflectance mode as an example of registered
images.
Figure 1: Imaging of distribution of ICG and Pp IX in
optical phantoms in fluorescence mode.
3.3 Analysis of Hemoglobin Oxygen
Saturation with Diffuse Reflectance
Visualization in NIR Region
The time dependences of the deoxygenation
parameter (the ratio of the deoxygenated
hemoglobin concentration to the total concentration
of hemoglobin) were obtained during nitrogen
ventilation of an optical phantom containing a
solution of whole blood in the scattering medium,
which served as calibration curves for analyzing
similar dependencies recorded in experiments with
laboratory animals. On Figure 2 we can see the
example of pictures in two wavelengths obtained
during in vivo experiment. Laboratory animals were
ventilated with nitrogen to achieve significant
changes in the deoxygenation parameter without
harming the animal's health.
Figure 2: Imaging of optical scattering phantoms with
haemoglobin in diffuse reflectance mode.
The curve of change in the deoxygenation
parameter of the animal brain depending on the
ventilation time is shown in Figure 3.
Figure 3: Analysis of deoxygenation in NIR in vivo.
Combined Video Analysis of ICG and 5-ALA Induced Protoporphyrin IX and Hemoglobin Oxygen Saturation in near Infrared
323
4 CONCLUSIONS
The video analysis technique for intraoperative
navigation during neurosurgery is proposed that
allows simultaneous registration of fluorescent
intraoperative images of tumors and surrounding
tissues at two wavelengths corresponding to the
fluorescence of protoporphyrin IX, which
characterizes the activity of metabolic processes in
cells, and fluorescence of indocyanine green, which
characterizes the parameters of tissue blood supply
with display in mode overlaying two fluorescent
pictures, as well as recording of intraoperative
images of tumors in diffusevely reflected broadband
light to measure the degree of tissue hypoxia.
ACKNOWLEDGEMENTS
The reported study was funded by RFBR according
to the research project 17-00-00162 (K) (17-00-
00159) and partially supported by the
Competitiveness Program of MEPhI.
REFERENCES
Sato, T., Suzuki, K., Sakuma, J., Takatsu, N., Kojima, Y.,
Sugano, T., Saito, K., 2015, Development of a new
high-resolution intraoperative imaging system (dual-
image videoangiography, DIVA) to simultaneously
visualize light and near-infrared fluorescence images
of indocyanine green angiography.
ActaNeurochirurgica, Volume 157, Issue 8, pp 1295
1301.
Loshchenov, M., Blondel, W., Amouroux, M., Steiner, R.,
Potapov, A., Golbin, D., Loran, O., Babaev, A.,
Borodkin, A., Daul, C., Trinh, D.-H., Loschenov, V.,
2018, Multimodal fluorescence imaging navigation for
surgical guidance of malignant tumors in
photosensitized tissues of neural system and other
organs, Proc. SPIE 10677, Unconventional Optical
Imaging, 106773N.
Mustari, A., Nakamura, N., Kawauchi, S., Sato, S., Sato,
M., Nishidate, I., 2018, RGB camera-based imaging of
cerebral tissue oxygen saturation, hemoglobin
concentration, and hemodynamic spontaneous low-
frequency oscillations in rat brain following induction
of cortical spreading depression, Biomed. Opt.
Express 9, 933-951.
Giannoni, L., Lange, F., Tachtsidis I., 2018, Hyperspectral
imaging solutions for brain tissue metabolic and
hemodynamic monitoring: past, current and future
developments. J. Opt. 20 044009
Stratonnikov, A.A., Meerovich, G.A., Ryabova, A.V.,
Savel'eva, T.A., Loshchenov, V.B., , 2006,
Application of backward diffuse reflection
spectroscopy for monitoring the state of tissues in
photodynamic therapy, Quantum Electronics, vol. 36,
no. 12, pp. 11031110.
Schmitt, J. M., and Kumar, G., 1996, Spectral distortions
in near-infrared spectroscopy of turbid materials.
//Appl. Spectrosc. 50, 10661073.
Kumar, G., Schmitt, J. M., 1997, Optimal probe geometry
for near-infrared spectroscopy of biological tissue.
//Appl. Opt., 36, 22862293.
Gebhart, S. C., Majumder, S. K., Mahadevan-Jansen, A.,
2007, Comparison of spectral variation from
spectroscopy to spectral imaging. //Applied Optics,
46(8).
Savelieva, T. A., Loshchenov, V. B., Goryainov, S. A.,
Shishkina, L. V., and Potapov. A. A., 2015, A
spectroscopic method for simultaneous determination
of protoporphyrin IX and hemoglobin in the nerve
tissues at intraoperative diagnosis, Russian Journal of
General Chemistry, vol. 85, no. 6, pp. 15491557.
PHOTOPTICS 2019 - 7th International Conference on Photonics, Optics and Laser Technology
324