Information Technology Reference
In-Depth Information
These results were achieved by applying a cognitive (semantic) analysis conducted
by semantic reasoning modules of parsers in the proposed diagnostic system.
Semantics could be identified by defining semantic actions subordinated to structural
rules of the introduced graph grammars. It seems that this proposed approach has the
characteristics of a scientific novelty, as there are no reports in the literature of similar
algorithmic solutions and analyses of highly complex, i.e. multi-object, images. This
approach could be applied in practice in medical diagnostic support systems.
The research work conducted to demonstrate the legitimacy of cognitive analysis
application in UBIAS systems, image data analysis systems, aims to develop them and
broaden their application capabilities. This type of data analysis and interpretation is
only possible due to image analysis. It is worth adding that the image analysis process
rarely covers the complete image. Very frequently only fragments of the image, impor-
tant from the perspective of the analysis process carried out, are analysed. Elements
acquired from an image by the process of exposing them constitute an important ele-
ment of the correct process of cognitive data analysis. We need to remember that very
frequently, the analysed image contains some (frequently a lot of) interference due to
either hardware defects or the lack of skill of the person taking that image. Such cases
require special precision in acquiring significant information necessary to carry out the
cognitive analysis from the image. If insignificant information were acquired, this
could lead to the wrong interpretation of the analysed data. If medical lesions are inter-
preted, this could threaten the patient's life and health. This could happen during an
attempt at the automatic analysis of image data, while in the case of the cognitive
analysis of data it should be emphasised that this analysis allows such a hazard to be
eliminated. This is because the analysis process itself runs in two trains. An important
stage in the process of cognitively analysing the interpreted data is reasoning, based on
the semantic (meaning) content, that is e.g. the size of the lesion, its location, length,
width, area etc. At the same time, a comparative stage (described here as the cognitive
resonance) is carried out, as a result of which expert knowledge is used to compare the
information coming from the system with the current database of expert knowledge. It
is this structure of the system which helps to minimise mistakes, which can be made
during the varied processes of data analysis and interpretation.
At the current stage, the research work allows us to find that the proposed solution
works right when analysing and interpreting medical data, in particular foot bone
images. It is worth noting that the notion of applying cognitive categorisation proc-
esses to analyse image data can be used to analyse other types of images. However, it
should be added that such an experiment will require the formal linguistic structures
proposed here to be re-established and the correct groups of lesion types to be defined
for other medical deformations found in other image groups. It will also be necessary
to build an expert database about the lesions analysed.
The research conducted by the author based on analyses of images depicting
lesions of foot bones has shown that the cognitive analysis of data in the form of
images may significantly enhance the capabilities of modern information systems and
systems supporting medical diagnostics. In particular, this research has shown that a
correctly built grammar makes it possible to precisely define and then run an analysis
and then describe selected diagnostic cases. Then, significant semantic information
can be extracted from them about the nature of the processes and lesions occurring in
foot bone structures. It is worth emphasising that the results obtained in this research
Search WWH ::
Custom Search