Fluorescence Pattern Analysis-How To Achieve
a Better Quality-Standard?
Institute of Computer Vision and Applied Computer Sciences, IBaI, Germany
Submission: October 04, 2017; Published: October 13, 2017
*Corresponding author: Petra Perner, Institute of Computer Vision and Applied Computer Sciences, IBaI, Germany, Email: firstname.lastname@example.org
How to cite this article: Priyanka B, Meeta L, Banwari L. Exterminating Attribute of Microbial Community in Oil and Gas Pipeline Network. Curr Trends
Biomedical Eng & Biosci. 2017; 9(4): 555769. DOI: 10.19080/CTBEB.2017.09.555769.
Quantitative imaging of fluorescent proteins and patterns is accomplished with a variety of techniques, including wide-field, confocal and multiphoton microscopy, ultrafast low-light level digital cameras and multi tracking laser control systems. These microscopic images can be of 2-dimensional or 3-dimensional nature, or even 4-dimensional nature such as videos recording the life cycle of a cell. To make fluorescent pattern analysis feasible in daily practice in cellular and molecular biology as well as in medicine, agriculture or other applications standardization is necessary to obtain authentically and reproducible results. We describe in this article the important steps that are necessary to support standardization.
In the rapidly expanding fields of cellular and molecular biology, fluorescence illumination and observation is becoming one of the techniques of choice to study the localization and dynamics of proteins, organelles, and other cellular compartments, as well as a tracer of intracellular protein trafficking. Quantitative imaging of fluorescent proteins and patterns is accomplished with a variety of techniques, including wide-field, confocal and multiphoton microscopy, ultrafast low-light level digital cameras and multi tracking laser control systems. These microscopic images can be of 2-dimensional or 3-dimensional nature, or even videos recording the life cycle of a cell.
To make fluorescent pattern analysis feasible in daily practice in cellular and molecular biology as well as in medicine, agriculture or other applications standardization is necessary to obtain authentically and reproducible results. Standardization has many aspects (Figure 1). It has to do with sample preparation, imaging techniques, knowledge acquisition, and image interpretation. It is an iterative process and cannot be solved from scratch. It is an interdisciplinary subject that requires input from different disciplines. Some of the aspects of Standardization of Fluorescence Pattern Analysis we will work out throughout this paper.
For research purposes, are usually studied only a few images in order to establish the imaging method. In this case it is feasible to do manually the image-interpretation of the images. The resulting manually obtained image descriptions are given to only a limited and elitare group of researchers for discussion purposes. Over time they will have developed their own image description language that is accepted in this community.
But the human interpretation causes a lot of problems. The interpretation of visual information/pattern requires a lot of experience. Humans are usually good in expressing emotion, giving driving direction that everyone can understand, but to describe what they see in an image so that any person gets the same impression is not a daily task for human and not regularly thought in school. That makes image interpretation difficult. The early research in image interpretation faced exactly on this problem and developed methodologies on visual knowledge acquisition that made progress in defect classification such as
wafer inspection, non-destructive testing or printing process
inspection. As a result, ontologies were born for specific
image interpretation tasks. The ontology is the basis for the
development of an automatically image-interpretation-systems.
Standardization requires the knowledge about prototypes of
pattern and visual descriptions. Methodically collected image
catalogues with showing and naming of examples (Table 1)
based on commonly excepted image vocabulary are necessary.
An automatic image interpretation system would allow
producing results that are reproducible and objective.
Reproducible since the system creates from the same image
always the same output as long as the automatic image processing
procedures work without any system failure nonetheless if
the image is processed today or ten days later. A human might
not able to do that! His daily performance heavily influences
the results. He might interpret an image differently today than
he did yesterday and as long as he does not calculate features
from the image by automatic image processing procedures his
measurements are qualitative and not quantitative. He cannot
give objective results rather his decisions are subjective. One
expert´s answer might be different from the answer of another
expert. Therefore, an automatic image interpretation system
will always be a big step towards standardization of the desired
image inspection tasks.
However to build such a system is difficult since not only
a human has problems to describe the visual content it is also
difficult to develop automatic image processing procedures that
can map the numerical representation of an image to the desired
The problem related to the automatic processing of
multimedia content resulted in MPG-3 standard that grouped
conventional image processing methods to visual symbolic
low-level terms that should allow a user to pick the right image
processing method in order to obtain the desired information
he wants to extract and describe for image retrieval or other
In defect classification, medical image interpretation, nondestructive
testing as well as in fluorescent pattern analysis the
visual terms are usually more complex and cannot be described
by a single visual symbolic low-level term. To give you an
example: How to describe fuzzy margin of an object by low-level
terms? This is a term very often appearing in medical image
interpretation across applications as well as in wafer inspection.
The recent developments in multimedia processing are therefore
not sufficient for many new arising visual image interpretation
tasks and we need to further develop new methods.
After an image method has passed research and goes into
industrial applications then they should usually work on-line
in a process. There is a growing use of these techniques in
industry for pharmacological aspects or diagnostic purposes
in medicine and agriculture. The huge amount of data created
in these processes cannot anymore be handled manually. They
require automatic image interpretation system. These image
interpretation systems should allow to interpret these images
automatically, and also to detect automatically new knowledge
to study the cellular and molecular processes.
A necessity for good image interpretation is images with
good image quality. Protocols for sample preparation as well
as robust and solidly-working imaging devices are important
to ensure that the objects in the image get imaged with high
contrast and constant brightness. This is another step towards
Standardization in Fluorescent Pattern Analysis.
Although there has been achieved a lot in that direction it is
not standard that the same image quality can be ensured over
the whole process. Researcher in the image processing field
are therefore more and more engaged with the development of
methods for assessing the image quality for different imaging
techniques for application in medicine, chemistry and biology
that allow to select the best image during the imaging process.
In general we can describe automatic image interpretation
as the process of mapping the numerical representation of an image into a logical representation such as suitable for image
description (Figure 2). An image interpretation system must
be able to extract symbolic features from the pixels of an image
(e.g., irregular structure inside the object, colocalization of
mitochondria, sharp margin). This is a complex process; the
image passes through several general processing steps until
the final symbolic description is obtained. These include image
preprocessing, image segmentation, image analysis, and image
interpretation (Figure 2). Interdisciplinary knowledge from
image processing, syntactical and statistical pattern-recognition
and artificial intelligence is required to build such systems. The
primitive (low- level) image features will be extracted at the
lowest level of an image interpretation system. Therefore, the
image matrix acquired by the image acquisition component
must first undergo image pre-processing to remove noise,
restore distortions, undergo smoothing, and sharpen object
contours. In the next step, objects of interest are distinguished
from background and uninteresting objects, which are removed
from the image matrix.
After having found the objects of interest in an image, we
can then describe the objects using primitive image features.
Depending on the particular objects and focus of interest, these
features can be lines, edges, ribbon, etc. A geometric object such
as a block will be described, for example, by lines and edges.
Typically, these low-level features have to be mapped to highlevel
features. A symbolic feature such as fuzzy margin will be
a function of several low- level features. Lines and edges will be
grouped together by perceptual criteria such as collinearity and
continuity in order to describe a block.
Image classification is usually referred to as the mapping of
features to predefined classes. Sometimes image interpretation
requires only image classification. However, image classification
is frequently only a first step of image interpretation. Low-level
features or part of the object description are used to classify
the object into different object classes in order to reduce the
complexity of the search space. The image interpretation
component identifies an object by finding the object that it
belongs to (among the models of the object class). This is done
by matching the symbolic description of the object in the scene
to the model of the object stored in the knowledge base. When
processing an image using an image interpretation system, an
image’s content is transformed into multiple representations
that reflect different abstraction levels. This incrementally
removes unnecessary detail from the image. The highest
abstraction level will be reached after grouping the image’s
features. It is a product of mapping the image pixels contained
in the image matrix into a logical structure. This higher level
representation ensures that the image interpretation process
will not be affected by noise appearing during image acquisition,
and it also provides an understanding of the image’s content. A
bottom-up control structure is shown for the generic system in
Figure 2. This control structure allows no feedback to preceding
processing components if the result of the outcome of the
current component is unsatisfactory. A mixture of bottom-up
and top-down control would allow the outcome of a component
to be refined by returning to previous component.
Assuming the prototypical pattern or scenes are known
as standard then it is possible to develop the necessary image
processing algorithm as a standard for analysing fluorescent
paper. In order to do that in a more systematic way, a
categorization of the tasks in the application area of Fluorescent
Pattern Analysis is necessary. The observation of prototypical
pattern or scenes empirically done by a human is usually a
time consuming process. Much more preferable would it be to
discover the prototypical appearance of pattern automatically.
In many high-content analysis project in drug discovery
are therefore recently calculated a lot of image features based
on conventional image processing algorithm from fluorescent
images. These features are more or less the features on which the
MEP3-standard is based on. They are not specially constructed to
describe the visual appearances of the objects in microscopic cell
fluorescent images. The experts usually try to summarize these
features by descriptive statistics or simple classifier to discover
some knowledge from the data. The problem with the described
feature description based on the low-level features still exists
here. Besides that, the large amounts of features are hardly to
overlook by the statistical data summarization methods.
New methods on data mining are necessary that can
automatically discover the final information needed for the
respective process. In general, we need to identify groups of
objects or events or map an image description into the final
decision (bacteria…, mitochondria). New clustering methods
based on conceptual clustering, incremental classification
method based on decision tree induction, case-based reasoning
and prototype- based classification have been developed so far
for this task.
As shown above, Standardization has to do with the
standardized preparation of the right probe, with the
establishment of robust and stable imaging devices as well as
with knowledge acquisition according to the established visual
knowledge elicitation methodology. An accepted and understood
ontology is necessary to describe the visual content in order to be
able share the knowledge and as basis for building an automatic
image interpretation system. The image interpretation system
should have application-oriented image pre-processing, image
segmentation and interpretation methods that allow adapting
the system to different application in the class of application.
To understand what the class of application is we need some
categorization of different application.
We have established a new forum to discuss this task in
more detail and show new research results to the community.
The forum is running under the umbrella of the community of
Mass Data Analysis of Images and Signals with Applications in
Medicine, r/g/b Biotechnology, Food Industries and Dietetics,
Biometry and Security, Agriculture, Drug Discover, and System
Biology (www.mda-signals.de). The aim of our new forum should
be to establish a forum of experts and practioners where we can
work on these topics and make progress toward Standardization
in Fluorescence Pattern Analysis.