Automatic cephalometric analysis /certified fixed orthodontic courses by Indian dental academy

INDIAN DENTAL ACADEMY
Leader in continuing dental education
www.indiandentalacademy.com




literature review focused on a single
question which tries to
identify, appraise, select and synthesize all

high quality research evidence relevant to
that question




Highest level of medical evidence



An understanding of systematic reviews

and how to implement them in practice is
becoming mandatory for all professionals
involved in the delivery of health care.






Broadbent & Hofrath – cephalometer 1931
contributed to the analysis of malocclusion
standardized diagnostic method in

orthodontic practice and research




Two approaches
 a manual approach
 a computer- aided approach




uses manual identification of landmarks,
 based either on an overlay tracing of the

radiograph to identify anatomical or
 constructed points followed by the transfer of the
tracing to a digitizer linked to a computer, or
 a direct digitization of the lateral skull
radiograph using a digitizer linked to a computer
and then locating landmarks on the monitor.


the computer software completes the
cephalometric analysis automatically

Landmarks digitized directly
from patient. – the DIGIGRAPH

JCO Volume 1990 Jun(360 - 367): The DigiGraph Work Station Part 1
Basic Concepts - SPIRO J. CHACONAS, DDS, MS; GARY A.
ENGEL, AB, MS; ANTHONY A. GIANELLY, DM



Cohen 1984



a scanned or digital cephalometric
radiograph is stored in the computer and
loaded by the software.



The software then automatically locates
the landmarks and performs the
measurements for cephalometric analysis



landmark detection



calculations have already been
automated with success




Image filtering plus knowledge-based
landmark search



Model- based approaches



Soft-computing approaches



Hybrid approaches




Resolution pyramid



Edge enhancement



Knowledge-based extraction



Gray level value difference


Resolution pyramid
Pyramid or 'pyramid representation' is a
type of multi-scale signal representation
developed by the computer vision, image
processing and signal processing
communities, in which a signal or an image
is subject to repeated smoothing and
subsampling
 Historically, pyramid representation is a
predecessor to scale space representation
and multiresolution analysis



Edge detection


in image processing and computer
vision, particularly in within the areas of
feature detection and feature
extraction, to refer to algorithms which
aim at identifying points in a digital
image at which the image brightness
changes sharply or more formally has
discontinuites




In computing, a grayscale or greyscale
digital image is an image in which the
value of each pixel is a single
sample, that is, it carries the full (and
only) information about its intensity.
Images of this sort are composed
exclusively of shades of neutral
gray, varying from black at the weakest
intensity to white at the strongest.

Advantages


Easy to implement



Image filtering techniques are well studied

and a large number are available


By encoding proper anatomical knowledge

better accuracy is achievable


Disadvantages
Can fail to capture morphological variability
in the radiographs
 Filtering results are highly dependent on
image quality and intensity level
 Sensitive to noise in the image
 Not all landmarks lie on edge
and, moreover, the edges or curves are
often unclear





Pattern matching



Spatial spectroscopy



Active shape models



Active contours with similarity function



Active appearance model


Pattern matching


In computer science, pattern matching is
the act of checking for the presence of the
constituents of a given pattern. In contrast
to pattern recognition, the pattern is rigidly
specified. Such a pattern concerns
conventionally either sequences or tree
structures. Pattern matching is used to test
whether things have a desired structure, to
find relevant structure, to retrieve the
aligning parts, and to substitute the
matching part with something else

Active Shape Models
(ASMs)


are statistical models of the shape of
objects which iteratively deform to fit to
an example of the object in a new
image. The shapes are constrained by
the PDM (Point Distribution Model)
Statistical Shape Model to vary only in
ways seen in a training set of labelled
examples. The shape of an object is
represented by a set of points
(controlled by the shape model)



segmentation refers to the process of
partitioning a digital image into multiple
regions (sets of pixels). The goal of
segmentation is to simplify and/or
change the representation of an image
into something that is more meaningful
and easier to analyze.[1] Image
segmentation is typically used to locate
objects and boundaries
(lines, curves, etc.) in images.

Active Appearance Model
(AAM)
Computer Vision algorithm for matching a
statistical model of object shape and
appearance to a new image. They are built
during a training phase. A set of images
together with coordinates of
landmarks, that appear in all of the images
is provided by the training supervisor.
 The approach is widely used for matching
and tracking faces and for Medical Image
Interpretation.



Advantages
Is invariant to scale, rotation, and
translation (the structure can be located
even if it is smaller or bigger than the given
model)
 Accommodates shape variability



Disadvantages






Needs models that must be created by
averaging the variations in shape of each
anatomical structure in a given set of
radiographs
Model deformation must be constrained and is
not always precise
Cannot be applied to partially hidden regions
Sensitive to noise in the image



PCNN (pulse coupled neural networks)



Support vector machines



Genetic algorithms



Fuzzy neural networks


Pulse-coupled networks or Pulse-Coupled
Neural Networks (PCNNs) are neural
models proposed by modeling a cat’s
visual cortex and developed for highperformance biomimetic image processing.
 Over the past decade, PCNNs have been
utilized for a variety of image processing
applications, including: image
segmentation, feature generation, face
extraction, motion detection, region
growing, noise reduction, and so on





'Support vector machines (SVMs)' are a set
of related supervised learning methods
used for classification and regression. They
belong to a family of generalized linear
classifiers. They can also be considered a
special case of Tikhonov regularization. A
special property of SVMs is that they
simultaneously minimize the empirical
classification error and maximize the
geometric margin; hence they are also
known as maximum margin classifiers.

genetic algorithm (GA)
Search technique used in computing to
find exact or approximate solutions to
optimization and search problems.
 Genetic algorithms are a particular class
of evolutionary algorithms (also known
as evolutionary computation) that use
techniques inspired by evolutionary
biology such as
inheritance, mutation, selection, and
crossover (also called recombination).





A neuro-fuzzy network is a fuzzy inference
system in the body of an artificial neural
network. Depending on the FIS type, there
are several layers that simulate the
processes involved in a fuzzy inference like
fuzzification, inference, aggregation and
defuzzification. Embedding an FIS in a
general structure of an ANN has the benefit
of using available ANN training methods to
find the parameters of a fuzzy system.



Accommodates shape variability



Tolerant to noise



Techniques are well studied



Large selection of software tools available




Results depend on the training set



Difficult to interpret some results



A number of network parameters, such as

topology and number of neurons, must be
determined empirically




To describe the techniques used for
automatic landmarking of cephalograms,



highlighting the strengths and
weaknesses of each one



reviewing the percentage of success in
locating each cephalometric point




The literature survey was performed by
searching
 Medline

 Institute of Electrical and Electronics

Engineers
 ISI Web of Science Citation Index databases




The survey covered the period from
January 1966 to August 2006. Abstracts
that appeared to fulfill the initial selection
criteria were selected by consensus.



The original articles were then retrieved.
Their references were also handsearched for possible missing articles.




Report of mean error between real position

and estimated position of landmark for
each point


Data in millimeter



Articles in English



Articles published from January 1966 to

August 2006




Review articles, abstracts and letters



Data in pixel



Total mean error of the method for a large set of landmarks



Descriptive methods



Computer assisted method



Only graphic data on accuracy of landmark location



Recognition rate presented as percentage of success



Automatic measurements not landmarks



Cephalometric points not stated



Not every landmark detection is a cephalometric point

Soft-computing or learning
approach


Why increased demand?


Advances



Affordability
in digital radiographic imaging


this literature review……..


many studies seemed to be
methodologically unsound
 inclusion criteria of patient radiographs

 the number of radiographs used,
 the error level to create a comparison with

the absence of any standard deviation of the

mean error

marked difference in
results…
Heterogeneity in the performance of
techniques to detect the same landmark
 Sella Point : Hybrid approaches >
model-based approach


 can be due to the high variability of the

shape of sella


marked difference in
results…
Porion, gonion and anterior nasal spine higher precision by the hybrid approach
 Nasion - nearly the same
 hybrid techniques –


 better results,
 accuracy close to the one suitable for clinical

practice


Discussion


Recommended total error
 x coordinate - 0.59 mm
 y coordinate - 0.56 mm



Euclidian value of error should be 0.81 mm



amazing values for standard errors and
standard deviations that are far from
standard errors for landmark identification



2 mm difference between the location of

landmark, obtained by some automatic
method and that obtained by the human
operator, has been considered by most
people to be successful
4 mm distance acceptable
 Conclusions drawn from the studies –
optimistic than reality





if one considers that two cephalometric
points are needed to trace a reference
plane or line, the resulting special
position of the line will be affected by the
errors of two points, not a single
one, and thus the error will be
increased.




studies presenting an agreement
between manual and computer-assisted
methods in millimeters, most consider
the Euclidian value, and do not refer to
the x-axis and




Automatic landmarking is the first and
last step in the development of a
completely automatic cephalometric

analysis.




Four broad categories
 image filtering plus knowledge- based

landmark search
 model-based approaches
 soft-computing approaches
 hybrid approaches




The systems described in the literature are
not accurate enough to allow their use for
clinical purposes as errors in landmark

detection were greater than those
expected with manual tracing




The ability to automatically identify
landmarks is fair for many landmarks, but
for routine clinical use it must be reliable



It should be emphasized that if automatic
land marking shall be used, it has to be
with respect to validity, reliability, and costs


Thank you
For more details please visit


Automatic cephalometric analysis /certified fixed orthodontic courses by Indian dental academy

Empfohlen

Empfohlen

Weitere ähnliche Inhalte

Andere mochten auch

Andere mochten auch (10)

Mehr von Indian dental academy

Mehr von Indian dental academy (20)

Kürzlich hochgeladen

Kürzlich hochgeladen (20)

Automatic cephalometric analysis /certified fixed orthodontic courses by Indian dental academy