Biomedical Engineering Reference
In-Depth Information
framework for other segmentation algorithm to determine the optimum region
size for algorithm implementation.
4. Incorporate quality assurance module in the segmentation framework to evalu-
ate the appropriateness to serve as input for subsequent stages of computerized
bone age assessment. The step is important to eliminate over-segmented regions
of hand bone and restore under-segmented regions to further improve the quality
of segmented hand bone. This concept provides an insight that a procedure for
imperfect segmentation framework system that is capable of analyzing the cur-
rent output and patch up the incompleteness accordingly is needed.
In conclusion to contributions, this topic provides the contrary perception to
conventional concept that prone to complicating the segmentation algorithm to
seek for enhancement in segmentation performance. Instead, the proposed seg-
mentation framework pioneers the insight by postulating that combinations of sev-
eral customized modules are capable of achieving result that tantamount to result
achieved by complicated algorithm or algorithm that demands scarce and limited
resources. The conception lies in the strategy to identify, target and analyze the
principal adversities that impede the performance of existing methods; then, based
on the analyzed result, the contribution is claimed to be advancing the concept of
adapting the input information to the central segmentation algorithm (in this topic,
it refers to unsupervised clustering). This concept is of contrast to conventional
segmentation framework that tends to complicate the fundamental segmentation
algorithm to adapt the input information. The strength of each module and the
idea of concatenating each of them by utilizing the by-product of each module are
breaking new ground in the field of automated image processing. These instruc-
tive insights embrace the potential to spark attentions and generate new research
grounds that will in turn contribute in other fields of applications.
1.7 Book Organization
The organization of the topic is summarized as follows.
This chapter introduces the background of research topic, presents the motiva-
tion of the research, states the research problems, declares the objectives of the
research, specifies the scope and identifies the significance of the research and
contributions of the research.
Chapter 2 presents the literature review by providing the critical appraisal of
various existing segmentation techniques which include the discussion on the
basic concept, mathematical formulation and their strengths and weaknesses in
the context of hand bone segmentation for computer-aided skeletal age scoring
system. This chapter concludes that the existing segmentation techniques are still
technically impractical for fully automated hand bone segmentation.
Chapter 3 presents the design and implementation of the proposed seg-
mentation framework that consist of several modules. Firstly, we describe the
Search WWH ::




Custom Search