TEXTURE ANALYSIS OF MILK PROTEIN GELS
USING DIGITAL IMAGE ANALYSIS
Juan Pablo Costa
1,2
, Horacio Castellini
3
, Patricia Risso
1
and Bibiana Riquelme
1,2
1
Dpto. de Química-Física, FCByF, Universidad Nacional de Rosario, Rosario, Argentina
2
Óptica Aplicada a la Biología, Instituto de Física Rosario (CONICET-UNR), Rosario, Argentina
3
Dpto. de Física, FCEIA, Universidad Nacional de Rosario, Rosario, Argentina
Keywords: Milk protein, Acid gel, Gel structure, Digital images, Glucono-delta-lactone, Bovine caseins.
Abstract: Sodium caseinate (NaCAS) is a very useful ingredient in food industry because of its nutritional and
functional properties. Acidification produces a gel structure as a result of the dissociation and aggregation
of caseinic fractions. Formation of these protein gels can be made by the slow reduction of pH through the
addition of glucono-delta-lactone (GDL). Depending on its concentration and temperature, hydrolysis speed
of GDL can affect the grade of hardness and elasticity of the formed gel. This study evaluated the effect on
the formation and structure of protein gels induced by different relations of GDL through analysis of digital
images obtained in an inverted conventional microscope and a confocal microscope. The entropy,
smoothness and variance decrease with the added GDL quantity, but the uniformity increases. Results
confirm that the texture depends on gelification speed, which is directly related to the amount of added
GDL. This digital image analysis technique using conventional or confocal microscopy is, therefore,
suitable and very useful for the texture analysis of acid gels formed by different GDL/NaCAS rates.
1 INTRODUCTION
The texture is a very important characteristic and its
analysis is a very useful tool to quantify and classify
objects or interest region in an image. Image texture
is a quantification of the space variation of
intensities that is impossible to define by its
sensorial character. There are several textural
parameters and algorithms proposed for the
quantification of an image texture such as the co-
occurrence matrix, statistical studies, the wavelet,
etc. (Jensen, 1996) All these techniques can be
useful to characterize a great variety of textures, but
they can be unsuccessful when the textures do not
show a periodic structure. A general assumption is
that the relevant information is in the space relation
inside the grayscale images.
Caseins (CN) represent the major protein
component of bovine milk. The CN precipitate at pH
4.6 and may be resolubilized by increasing the pH. If
the increase in the pH is carried out by the addition
of NaOH it is possible to end up obtaining sodium
caseinate (NaCAS). CN and NaCAS are extensively
used in food industry because of their
physicochemical, nutritional and functional
properties that make them valuable ingredients in
complex food preparations. Casein gels are
responsible for most of the rheological/textural
properties (i.e. stretch, fracture) of cheese and other
dairy products (Walstra, 1984; Mulvihill and Fox,
1989).
Dissociation and a further aggregation step of
CN fractions due to NaCAS acidification results in
the formation of a gel structure. A possible
explanation to this observation is that as pH is
adjusted towards the isoelectric point it causes a
decrease of the repulsive interactions, resulting in a
destabilization of the colloidal aggregates as pH
drops slightly below 5 at a given temperature (Braga
et al., 2006; Ruis et al., 2007). Nowadays, a process
that has gained the attention of food industry is
direct acidification by the addition of a lactone, such
as glucono-δ-lactone (GDL) which allows to
overcome some of the difficulties associated with
the traditional process of using bacteria. In fact, the
final pH of the system is a function of the amount of
GDL added whereas starter bacteria produce acid
until they inhibit their own growth as pH becomes
lower (Ruis et al., 2007; de Kruif, 1997).
322
Pablo Costa J., Castellini H., Risso P. and Riquelme B..
TEXTURE ANALYSIS OF MILK PROTEIN GELS USING DIGITAL IMAGE ANALYSIS .
DOI: 10.5220/0003289303220325
In Proceedings of the International Conference on Bioinformatics Models, Methods and Algorithms (BIOINFORMATICS-2011), pages 322-325
ISBN: 978-989-8425-36-2
Copyright
c
2011 SCITEPRESS (Science and Technology Publications, Lda.)
Depending on GDL concentration and
temperature, hydrolysis speed of GDL can affect the
grade of hardness and elasticity of the formed gel.
The compactness and elasticity of gels formed at the
end of the acidification process of NaCAS depend
on the kinetic of the aggregation phenomena. As the
aggregation process becomes slower the more easily
a polypeptide chain could acquire different
orientations leading to the formation of more
compact gels with more elasticity and hardness
(Nespolo et al., 2010).
The interaction mechanisms and rheological
properties of obtained gels can be analyzed and
characterized by optical techniques (Yan et al.,
2008; Lucey, 2002).
In the present work, the gel structure formation
kinetics and different compaction degree were
evaluated by means of textural parameters from
digital image analysis. In order to do this, images of
bottom gel surface were obtained by conventional
inverted microscopy and images of gel internal
structure were obtained by confocal microscopy.
2 MATERIALS AND METHODS
2.1 Sample Preparation
In this work, the formation and structure of NaCAS
gels (3%) induced by different GDL/NaCAS ratio
(0.35; 0.5; 0.7 y 1.0) was evaluated by means of
analysis of digital images obtained in an inverted
optical and a confocal microscope.
An aqueous solution of bovine caseinate
(NaCAS) 3% w/w was prepared from the
commercial drug (Sigma Co.). The gelification
process was started adding GDL on 5 g of NaCAS
solution, at 35 ºC. The different amounts of GDL
corresponded to different relations GDL/NaCAS
(0.35, 0.5, 0.7 and 1) according to:
CN
GDL
R
%
%
=
(1)
To obtain the microscopic images, 80 μL of each
sample at different R were placed in compartments
of the LAB-TEK II cells. The samples were made in
duplicate maintaining the incubation temperature at
35ºC.
2.2 Stained Protocol for Confocal
Microscopy
The gels were stained with Rhodamine B. In order to
do, a dilution 1/50 v/v of 1 mL of Rhodamine B at
0.01% in 49 mL of bidistilled water was prepared.
This medium was used to dissolve NaCAS to obtain
a solution at 3%.
2.3 Image Acquisition and Analysis
Transmission images of gels were obtained using a
conventional inverted microscopy (Union Optical)
with an objective 100x and a digital camera (Canon
PowershotA640) with a zoom 9.1x, for the different
GDL/NaCAS ratios.
Images of internal structure of gels stained with
Rhodamine B were obtained using a confocal
microscope (Nikon EZ-C1.) at (14.25 ± 0.05) µm on
the glass slide (inner the gel).
The effect on the formation and structure of
protein gels induced by different R was assessed
using analysis of microscopic digital images.
Previous to the analysis, the images were
transformed to numerical 8 bit formats RGB (Red,
Green, Blue) where each of them corresponds to
level color. All images were normalized to grayscale
by following transformation:
BGRY 114.0587.0299.0 +
+
=
(2)
Then, the values were normalized by mean of the
transformation:
minmax
min
)1(
YY
YY
LN
=
(3)
where L is the maximum gray level. It can see in
Figures 2 and 3.
In this numerical representation a simplest
statistical approaches were used. From normalized
gray level it obtained its histogram
()
i
Np
, where
11
=
Li
and L is the maximum gray level. As
well as texture estimations of obtained images, the
Shannon entropy S, smoothness R and uniformity U
were studied and they were defined by the following
equation (Gonzalez and Woods, 2002; Haralick,
1979):
()
=
=
1
0
2
))((log
L
i
ii
NpNpS
(4)
)(1
1
1
2
N
R
σ
+
=
(5)
=
=
1
0
2
)(
L
i
i
NpU
(6)
where
)(
2
N
σ
is the variance. Because the
(
)
i
Np
have values in the range form 0 to 1 and their sum is
TEXTURE ANALYSIS OF MILK PROTEIN GELS USING DIGITAL IMAGE ANALYSIS
323
equals 1, measure U is maximum for an image in
which all gray level are equal. Instead entropy is a
measure of variability and is 0 for a uniform image.
3 RESULTS AND DISCUSSIONS
Figure 1 shows the pictures registered at different
times after GDL addition for the visualization of
gelification kinetics.
Figure 2 and 3 show respectively the
transmission and confocal images of gels obtained
for R = 0.35, 0.5, 0.7 and 1.0. From these images, it
is possible to observe differences in the internal
microstructure of gels. These observations can be
quantified by the texture parameters as shown in
tables 1 and 2.
15 minutes 30 minutes
Figure 1: Visualization of gelification kinetics as a
function of GDL/NaCAS ratios (0.35, 0.5, 0.7 and 1.0).
(a) R = 0.35 (b) R = 0.5 (c) R = 0.7 (d) R = 1
Figure 2: Images of gels obtained using a conventional inverted microscopy (Union Optical) with an objective 100x and a
digital camera (Canon PowershotA640) with a zoom 9.1x, for the different GDL/NaCAS rates.
(a) R = 0.35 (b) R = 0.5 (c) R = 0.7 (d) R = 1
Figure 3: Images obtained by confocal microscopy inner the gels at (14.25 ± 0.05) µm on the slide.
Table 1: Texture parameters for the different GDL/NaCAS ratios. Digital images obtained by conventional inverted
microscopy of acid gels on the slides.
R Entropy Smoothness Variance Uniformity
0.35 7,68 ± 0,01 0,046 ± 0.001 3100 ± 100 0,0050 ± 0,0001
0.5 7,58 ± 0,01 0,034 ± 0.001 2300 ± 100 0,0058 ± 0,0001
0.7 7,47 ± 0,09 0,029 ± 0,004 2000 ± 200 0,0065 ± 0,0006
1 7,39 ± 0,03 0,026 ± 0,001 1750 ± 50 0,0069 ± 0,0002
Table 2: Texture parameters for the different GDL/NaCAS ratios. Digital images obtained by confocal microscopy of inner
structure of acid gel measured at 14 μm on glass slides.
R Entropy Smoothness Variance Uniformity
0.35 5.2 ± 0.1 0.012 ± 0.001 800 ± 90 0.038 ± 0.004
0.5 5.4 ± 0.2 0.014 ± 0.003 940 ± 100 0.030 ± 0.003
0.7 4.6 ± 0.1 0.014 ± 0.001 830 ± 100 0.051 ± 0.004
1 4.4 ± 0.1 0.015 ± 0.001 980 ± 90 0.056 ± 0.005
BIOINFORMATICS 2011 - International Conference on Bioinformatics Models, Methods and Algorithms
324
4 CONCLUSIONS
Results show that the structure compaction, texture
and size of internal interstices depend on the
gelification rate, which is related to the GDL added
to the solution.
The entropy decrease and the uniformity
increases with the added GDL in the images
obtained using both microscopes, being the entropy
the parameter that has more high precision. Results
confirm that the texture depends on gelification
speed, which is directly related to the amount of
added GDL. Therefore, the present digital image
analysis technique is suitable and very useful to
characterize the texture of NaCAS acid gels formed
by different GDL ratios.
The present analysis technique of digital images
obtained by conventional and confocal microscopy
is suitable for a structural study of acid protein gels
and can be useful to evaluate the texture of dairy
products.
REFERENCES
Braga, A., Menossi, M., Cunha, R. (2006). The Effect of
the Glucono-[delta]-Lactone/Caseinate Ratio on
Sodium Caseinate Gelation. International Dairy
Journal, 16, (5), 389-398.
de Kruif, C. G. (1997). Skim Milk Acidification. Journal
of Colloid and Interface Science, 185, (1), 19-25.
Gonzalez, R., Woods, E. (2002). Digital Imaging
Processing. 2nd edition, Prentice Hall.
Haralick R. (1979). Statistical and Structural Approach to
Texture. Proceeding of IEEE, Vol. 67 (5).
Jensen, J.R. (1996). Introductory Digital Image
Processing, Prentice Hall.
Lucey, J. (2002). Formation and Physical Properties of
Milk Protein Gels, J. Dairy Sci. 85, 281-294.
Mulvihill, D., Fox, P. (1989). Caseins and Manufactured.
In Development in Dairy Chemistry, Ed., P. F. F., Ed.
Elsevier Applied Science: London & New York,;
(Vol. 4, pp. 97-130).
Nespolo, C., Reggiardo, A., Mancilla Canales, M.,
Wagner, J., Alvarez, E., Brandelli, A, Risso, P. (2010).
Effect of Calcium on Ovine Caseinate Functional
Properties. Journal of Chemical and Engineering
Data, in press.
Ruis, H., Venema, P., van der Linden, E. (2007). Relation
Between pH-Induced Stickiness and Gelation
Behaviour of Sodium Caseinate Aggregates as
Determined by Light Scattering and Rheology. Food
Hydrocolloids, 21, (4), 545-554.
Walstra, P. (1984). Dairy Chemistry and Physics. New
York, USA; John Wiley & Sons; (pp. 1-225).
Yan, L., Yue, W., Jiye, C. (2008). Atomic Force
Microscopic Investigations on Pectin Gelling Made
Under Optimum Conditions. Chemical Journal on
Internet 10, (9), P. 45
TEXTURE ANALYSIS OF MILK PROTEIN GELS USING DIGITAL IMAGE ANALYSIS
325