MITOTIC HEP-2 CELL RECOGNIITON USING SUPPORT VECTOR MACHINE UNDER CLASS SKEW
A person with an autoimmune diseases will became hypersensitive to the surrounding that other normal person would usually not consider at all such as an allergy. This reaction happened when our immune system recognise our normal tissue as a dangerous foreign elements and proceed to attack them. The...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Year Project |
| Language: | English |
| Published: |
Universiti Teknologi Petronas
2014
|
| Subjects: | |
| Online Access: | http://utpedia.utp.edu.my/13450/1/22.pdf http://utpedia.utp.edu.my/13450/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | A person with an autoimmune diseases will became hypersensitive to the surrounding that other normal person would usually not consider at all such as an allergy. This reaction happened when our immune system recognise our normal tissue as a dangerous foreign elements and proceed to attack them. The presence of antinuclear autoantibodies (ANA) in a patient serum can be detected by using the Indirect Immunofluorescence (IIF) image. By adding the mitotic cells into the well, the level of accuracy of the results achieved can be increased. The mitotic cells itself plays a crucial role in diagnosing an autoimmune diseases. This paper will focuses on the extracting the features of a mitotic HEp-2 cell in order to determine the presence of an ANA by noting the cells fluorescent-stained pattern, their intensity and also the presence of the mitotic cell itself. A skewed distribution of both mitotic and non-mitotic cells in the samples will also be considered to ensure the practicality of the project. To assist in the objectives, all the techniques used are explain in more detailed in this paper along with the result obtained by simulation from MATLAB for every steps from pre-processing to user interface menu. The procedures for the recognition of mitotic cells are image acquisition, pre-processing, segmentation, feature extraction and classification. The results obtained were tested using HEp-2 cell image datasets from MIVIA and from collaboration with Hospital Universiti Sains Malaysia (HUSM). The feature extractor used is the gray level co-occurrence matrix (GLCM) and classified using support vector machine (SVM) which will be presented in the RESULTS section. |
|---|