Sie befinden Sich nicht im Netzwerk der Universität Paderborn. Der Zugriff auf elektronische Ressourcen ist gegebenenfalls nur via VPN oder Shibboleth (DFN-AAI) möglich. mehr Informationen...
2018 13th International Conference on Computer Engineering and Systems (ICCES), 2018, p.313-318
Ort / Verlag
IEEE
Erscheinungsjahr
2018
Quelle
IEEE Xplore
Beschreibungen/Notizen
The main purpose of this research is to establish a Computer Aided Diagnostic (CAD) approach for the detection and classification of Oral Epithelial Dysplasia. The disturbances that occur in the epithelial layers is diagnosed as premalignant dysplasia. The epithelial dysplasia diagnosis, in-terms of accuracy, is pathologically difficult and contributes to main challenges to oral pathologists due to the multiple dysplastic criteria of the disease such as the loss of polarity of the basal cells and other cellular and nuclear changes. A new approach has been developed based on different selections and magnifications of stained microscopic images. The approach extracts a set of features that would automatically diagnose the image supplying its condition and the category it has reached so far. The resulted analysis from our research will enable the pathologists in classifying cells abnormalities. Feature extracted using Oriented FAST and Rotated BRIEF (ORB) algorithm with the Support Vector Machine (SVM) as a classification algorithm. The proposed approach achieved 92.8% of accuracy in classification Oral Epithelial Dysplasia. The system was trained and tested on a total of forty-six cases of magnification 100× levels of 70% and 30% respectively. This research presents for the first time a diagnostic approach for grading oral epithelial dysplasia according to sixteen extracted features with the given experimented accuracy rates on different magnification levels.