Dental Images' Segmentation Using Threshold Connected Component Analysis.

Vincent Majanga,Serestina Viriri

doi:10.1155/2021/2921508

Vincent Majanga, Serestina Viriri

Open Access

https://doi.org/10.1155/2021/2921508

Copy DOI

Abstract

Recent advances in medical imaging analysis, especially the use of deep learning, are helping to identify, detect, classify, and quantify patterns in radiographs. At the center of these advances is the ability to explore hierarchical feature representations learned from data. Deep learning is invaluably becoming the most sought out technique, leading to enhanced performance in analysis of medical applications and systems. Deep learning techniques have achieved great performance results in dental image segmentation. Segmentation of dental radiographs is a crucial step that helps the dentist to diagnose dental caries. The performance of these deep networks is however restrained by various challenging features of dental carious lesions. Segmentation of dental images becomes difficult due to a vast variety in topologies, intricacies of medical structures, and poor image qualities caused by conditions such as low contrast, noise, irregular, and fuzzy edges borders, which result in unsuccessful segmentation. The dental segmentation method used is based on thresholding and connected component analysis. Images are preprocessed using the Gaussian blur filter to remove noise and corrupted pixels. Images are then enhanced using erosion and dilation morphology operations. Finally, segmentation is done through thresholding, and connected components are identified to extract the Region of Interest (ROI) of the teeth. The method was evaluated on an augmented dataset of 11,114 dental images. It was trained with 10 090 training set images and tested on 1024 testing set images. The proposed method gave results of 93% for both precision and recall values, respectively.

Highlights

Dental caries is the most widespread chronic disease affecting teeth worldwide
Training CNNs on computed tornograph (CT) images helps extract positive intensity patches that belong to organs at risks (OAR) of interest and negative intensity patches that belong to surrounding structures. e patches were passed through CNN layers and image features; namely, edges, corners, and end points were captured and combined into more complex high-order features that describe the OAR efficiently. e trained network was applied to classify voxels in a region of interest (ROI) in the test images to obtain a corresponding OAR as a segmented image result [23]
Ahmed et al [25] present a caries detection method that applies K-means clustering and threshold method for segmentation of CT images. is is in order to construct a 3D view of the carious lesion which is an integral part of the diagnosis of dental caries

Summary

Introduction

Dental caries is the most widespread chronic disease affecting teeth worldwide. Recently, there has been a decline in the rates of large cavity lesions, but still early lesions can be identified in most people [1]. Most of conventional caries detection methods rely on inspecting teeth visually. E introduction of dental radiography is to detect hidden or inaccessible lesions that could not be otherwise done through conventional methods. Detection of dental caries lesions is an important determinant of treatment measures and is a beneficiary of the introduction of new tools [3]. Most dentists use bitewing radiographs to aid in the location of dental caries. Locating of dental caries is a challenging task and sometimes even the experienced dentists miss the carious lesions when they are just presented with bitewing radiographs [5]. Detection of dental caries relied on visual-tactile methods [6]. Sensitivity of visual-tactile methods is limited and especially done on posterior proximal teeth surfaces. Radiographic methods tend to have high sensitivity but require ionizing radiation [7]

Methods

Results

Conclusion