A computer-aided automated methodology for the detection and classification of occlusal caries from photographic color images

Berdouses, Elias; Koutsouri, Georgia; Tripoliti, Evanthia E.; Matsopoulos, George K.; Oulis, Constantine; Fotiadis, Dimitrios I.

doi:10.1016/j.compbiomed.2015.04.016

Cited by 52 publications

(50 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Evaluation of attrition bias: Not all of the studies reported complete results. Berdouses et al [25] detailed all the results analyzed in the present review.…”

Section: Study Quality Assessmentmentioning

confidence: 89%

Dental Caries Diagnosis and Detection Using Neural Networks: A Systematic Review

Prados-Privado

Villalón²,

Martínez-Martínez

et al. 2020

JCM

View full text Add to dashboard Cite

Dental caries is the most prevalent dental disease worldwide, and neural networks and artificial intelligence are increasingly being used in the field of dentistry. This systematic review aims to identify the state of the art of neural networks in caries detection and diagnosis. A search was conducted in PubMed, Institute of Electrical and Electronics Engineers (IEEE) Xplore, and ScienceDirect. Data extraction was performed independently by two reviewers. The quality of the selected studies was assessed using the Cochrane Handbook tool. Thirteen studies were included. Most of the included studies employed periapical, near-infrared light transillumination, and bitewing radiography. The image databases ranged from 87 to 3000 images, with a mean of 669 images. Seven of the included studies labeled the dental caries in each image by experienced dentists. Not all of the studies detailed how caries was defined, and not all detailed the type of carious lesion detected. Each study included in this review used a different neural network and different outcome metrics. All this variability complicates the conclusions that can be made about the reliability or not of a neural network to detect and diagnose caries. A comparison between neural network and dentist results is also necessary.

show abstract

“…Evaluation of attrition bias: Not all of the studies reported complete results. Berdouses et al [25] detailed all the results analyzed in the present review.…”

Section: Study Quality Assessmentmentioning

confidence: 89%

Dental Caries Diagnosis and Detection Using Neural Networks: A Systematic Review

Prados-Privado

Villalón²,

Martínez-Martínez

et al. 2020

JCM

View full text Add to dashboard Cite

show abstract

“…The clinical parameters were further pruned to a compact yet informative subset to investigate informative variables that could distinguish low-risk patients from high-risk patients. This process is known as feature subset selection (FSS) [ 23 - 25 ], and we used the well-known Recursive Feature Elimination procedure based on SVM (RFE-SVM). The optimal subset with acceptable prognostic capabilities was obtained, and the SVM model was tested using a ten-fold cross-validation scheme.…”

Section: Methodsmentioning

confidence: 99%

Development of a ten-signature classifier using a support vector machine integrated approach to subdivide the M1 stage into M1a and M1b stages of nasopharyngeal carcinoma with synchronous metastases to better predict patients' survival

Jiang

You²,

Pei

et al. 2015

Oncotarget

View full text Add to dashboard Cite

The aim of this study was to develop a prognostic classifier and subdivided the M1 stage for nasopharyngeal carcinoma patients with synchronous metastases (mNPC). A retrospective cohort of 347 mNPC patients was recruited between January 2000 and December 2010. Thirty hematological markers and 11 clinical characteristics were collected, and the association of these factors with overall survival (OS) was evaluated. Advanced machine learning schemes of a support vector machine (SVM) were used to select a subset of highly informative factors and to construct a prognostic model (mNPC-SVM). The mNPC-SVM classifier identified ten informative variables, including three clinical indexes and seven hematological markers. The median survival time for low-risk patients (M1a) as identified by the mNPC-SVM classifier was 38.0 months, and survival time was dramatically reduced to 13.8 months for high-risk patients (M1b) (P < 0.001). Multivariate adjustment using prognostic factors revealed that the mNPC-SVM classifier remained a powerful predictor of OS (M1a vs. M1b, hazard ratio, 3.45; 95% CI, 2.59 to 4.60, P < 0.001). Moreover, combination treatment of systemic chemotherapy and loco-regional radiotherapy was associated with significantly better survival outcomes than chemotherapy alone (the 5-year OS, 47.0% vs. 10.0%, P < 0.001) in the M1a subgroup but not in the M1b subgroup (12.0% vs. 3.0%, P = 0.101). These findings were validated by a separate cohort. In conclusion, the newly developed mNPC-SVM classifier led to more precise risk definitions that offer a promising subdivision of the M1 stage and individualized selection for future therapeutic regimens in mNPC patients.

show abstract

“…We also found two studies in which the authors detected different ICDAS scores on photographic images. Unfortunately, the results of one of them are not comparable to ours, since this study aimed exclusively at comparing the caries lesion detection performance of newly developed software to that of an ICDAS-experienced dentist using the same photographic images [ 26 ]. No comparison with clinical examination data was reported in this previously published article [ 26 ].…”

Section: Discussionmentioning

confidence: 99%

“…Unfortunately, the results of one of them are not comparable to ours, since this study aimed exclusively at comparing the caries lesion detection performance of newly developed software to that of an ICDAS-experienced dentist using the same photographic images [ 26 ]. No comparison with clinical examination data was reported in this previously published article [ 26 ]. The authors of the other study used images of the buccal aspect of teeth from patients that had recently completed orthodontic treatment, with caries lesions in different stages of progression [ 13 ].…”

Section: Discussionmentioning

confidence: 99%

Is it feasible to use smartphone images to perform telediagnosis of different stages of occlusal caries lesions?

et al. 2018

View full text Add to dashboard Cite

The purpose of this study was to compare the performance of two different models of smartphone and a conventional camera with that of direct clinical examination in detecting caries lesions at different stages of progression in deciduous molars. The photographic equipment consisted of two smartphones (iPhone and Nexus 4) and a conventional macro camera setup. First, in the laboratory phase of the study, we compared the images of 20 exfoliated primary teeth having caries lesions at different stages. Then, in the clinical phase of the study, the images of 119 primary molars from fifteen children (3 to 6 years old) were used. All of the photographic images were taken using the previously described devices. In both groups, two examiners, blinded to the photographic equipment used, assessed the images independently on a computer screen, and classified them according to the International Caries Detection and Assessment System (ICDAS). The teeth were then examined directly by two other experienced examiners, and the consensus reached was considered the reference standard. Parameters of validity, such as percentage of correct answers, agreement with the reference standard, sensitivity, specificity and inter-examiner agreement (using the weighted kappa test) were calculated. The examiners performed similarly in both in vitro and in vivo studies. Inter-examiner reliability was approximately 0.7 for all the devices in the laboratory setting, and for the macro camera photography system in the clinical setting, but it was approximately 0.9 for the iPhone and Nexus images taken in vivo. With regard to the percentage of correct answers, the highest values were observed for sound and extensive caries lesions in both laboratory and clinical settings. The percentage of correct answers for initial and moderate lesions was particularly low in the clinical evaluation, irrespective of the camera devices used. Therefore, we concluded that photographic diagnosis using smartphone images is feasible and accurate for distinguishing sound tooth surfaces from extensive caries lesions; however, photographic images are not a good method for accurately detecting initial and moderate caries lesions.

show abstract

A computer-aided automated methodology for the detection and classification of occlusal caries from photographic color images

Cited by 52 publications

References 32 publications

Dental Caries Diagnosis and Detection Using Neural Networks: A Systematic Review

Dental Caries Diagnosis and Detection Using Neural Networks: A Systematic Review

Development of a ten-signature classifier using a support vector machine integrated approach to subdivide the M1 stage into M1a and M1b stages of nasopharyngeal carcinoma with synchronous metastases to better predict patients' survival

Is it feasible to use smartphone images to perform telediagnosis of different stages of occlusal caries lesions?

Contact Info

Product

Resources

About