A hybrid Fuzzy C-Means and Neutrosophic for jaw lesions segmentation

It is really important to diagnose jaw tumor in its early stages to improve its prognosis. A differential diagnosis could be performed using X-ray images; therefore, accurate and fully automatic jaw lesions image segmentation is a challenging and essential task. The aim of this work was to develop a...

Full description

Saved in:
Bibliographic Details
Published in:Ain Shams Engineering Journal Vol. 9; no. 4; pp. 697 - 706
Main Author: Alsmadi, Mutasem K.
Format: Journal Article
Language:English
Published: Elsevier B.V 01.12.2018
Elsevier
Subjects:
Automatic segmentation
Fuzzy C-Means clustering algorithm
Jaw panoramic X-ray image
Neutrosophy
Jaw panoramic X-ray image
Neutrosophy
Automatic segmentation
Fuzzy C-Means clustering algorithm
ISSN:2090-4479
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:It is really important to diagnose jaw tumor in its early stages to improve its prognosis. A differential diagnosis could be performed using X-ray images; therefore, accurate and fully automatic jaw lesions image segmentation is a challenging and essential task. The aim of this work was to develop a novel, fully automatic and effective method for jaw lesions in panoramic X-ray image segmentation. The hybrid Fuzzy C-Means and Neutrosophic approach is used for segmenting jaw image and detecting the jaw lesion region in panoramic X-ray images which may help in diagnosing jaw lesions. Area error metrics are used to assess the performance and efficiency of the proposed approach from different aspects. Both efficiency and accuracy are analyzed. Specificity, sensitivity and similarity analyses are conducted to assess the robustness of the proposed approach. Comparing the proposed approach with the Hybrid Firefly Algorithm with the Fuzzy C-Means, and the Artificial Bee Colony with the Fuzzy C-Means algorithm, the proposed approach produces the most identical lesion region to the manual delineation by the Oral Pathologist and shows better performance (FP rate is 6.1%, TP rate is 90%, specificity rate is 0.9412, sensitivity rate is 0.9592 and similarity rate is 0.9471).
ISSN:2090-4479
DOI:10.1016/j.asej.2016.03.016