In our lab we collect a set of Software tools to approach the
quantitative measurement of molecular imaging data. For example,
ImageJ [17] is one of the most popular software used for biological
imaging analysis (Fractal Dimension and Lacunarity, Color Histo-
gram, Roundness, Cell counter, etc.). Another relevant Software
tool is CellProfiler [18]. These softwares are made available, via the
COSYSBI architecture already described, to all researchers attend-
ing the lab.2.2.1 An Example of Data
Analysis Workflow for Cell
Cultures
Innovative research in experimental biology is enabled by the avail-
ability of advanced molecular imaging tools. In this review, to
characterize cytoskeleton proprieties, we have used the confocal
images from in-vitro cells cultures experiments, through multiple
analyses and in combination with the morphological information
on optical microscopy. The COSYSBI biological system repository
is an essential management tool for gathering information on cells
(cell lines, morphological characteristics, and information useful for
image analysis), to be linked to external databases on cells or
previous experiments, performed also by other research groups.
The easy accessibility of data stemming from multiple sources is a
key factor in deriving characteristic parameters, to be fed into the
suitable algorithms.
Confocal microscopy images are extremely useful for the quan-
titative analysis of experiments and model construction. In particu-
lar, visualization of the cytoskeleton structure is useful to
understand cell motility, stiffness, and more generally the cell phe-
notype. Cytoskeleton properties can be investigated, among other
methods, by the calculation of the fractal dimension, starting from
microscopy data. Such an analysis has been performed on confocal
images cells, to observe the changes of cell proprieties due to the
experimental protocol. Clear meaningful relationships are identi-
fied from the data experimented on different treatment condition
and are assessed graphically and statistically interpreted.
Figure2 shows the protocol used for the quantitative evalua-
tion of the shape parameters.
The first step is to snake the cell membrane. This is a difficult
task to do at computational level. Often the biologist needs to
manually administer this task. He is able to identify the profile of
the cell membrane with all its protrusions, based on his own expe-
rience, even when the membrane is not on the microscope’s focus
plane. This task is particularly decisive for the correct evaluation of
cell morphology.
The next step is to subtract the background, due to the nutrient
medium and microenvironment. Even this step is particularly cru-
cial for a good fit. The ability to use automatic threshold algorithms
is almost always impossible: in this case too, the biologist supervi-
sion is required.346 Garima Verma et al.