Comparison of caries diagnostic modalities: A clinical study in 40 subjects

A Deep Learning Approach to Automatic Tooth Caries Segmentation in Panoramic Radiographs of Children in Primary Dentition, Mixed Dentition, and Permanent Dentition

OBJECTIVES

The purpose of this study was to evaluate the effectiveness of dental caries segmentation on the panoramic radiographs taken from children in primary dentition, mixed dentition, and permanent dentition with Artificial Intelligence (AI) models developed using the deep learning method.

METHODS

This study used 6075 panoramic radiographs taken from children aged between 4 and 14 to develop the AI model. The radiographs included in the study were divided into three groups: primary dentition (n: 1857), mixed dentition (n: 1406), and permanent dentition (n: 2812). The U-Net model implemented with PyTorch library was used for the segmentation of caries lesions. A confusion matrix was used to evaluate model performance.

RESULTS

In the primary dentition group, the sensitivity, precision, and F1 scores calculated using the confusion matrix were found to be 0.8525, 0.9128, and 0.8816, respectively. In the mixed dentition group, the sensitivity, precision, and F1 scores calculated using the confusion matrix were found to be 0.7377, 0.9192, and 0.8185, respectively. In the permanent dentition group, the sensitivity, precision, and F1 scores calculated using the confusion matrix were found to be 0.8271, 0.9125, and 0.8677, respectively. In the total group including primary, mixed, and permanent dentition, the sensitivity, precision, and F1 scores calculated using the confusion matrix were 0.8269, 0.9123, and 0.8675, respectively.

CONCLUSIONS

Deep learning-based AI models are promising tools for the detection and diagnosis of caries in panoramic radiographs taken from children with different dentition.

DIAGNOdent versus International Caries Detection and Assessment System in detection of incipient carious lesions: A diagnostic accuracy study

Aim

The present investigation aimed at assessing the diagnostic accuracy of DIAGNOdent compared to the International Caries Detection and Assessment System-II (ICDAS-II) in the detection of facial, smooth surface noncavitated carious lesions.

Settings and Design

Sixty patients were enrolled in the current study according to the eligibility criteria. There were 161 teeth with noncavitated, white spot carious lesions and 32 sound teeth.

Materials and Methods

Before examination, teeth were cleaned and polished and all patients were assessed under standardized operating conditions: preset dental unit position, operating light, and prolonged air drying (approximately 5 s). All teeth were assessed by two calibrated examiners individually without any contact using ICDAS-II and DIAGNOdent.

Statistical Analysis

The diagnostic accuracy of DIAGNOdent device was determined, including sensitivity, specificity, overall accuracy, positive and negative predictive values, and receiver operating characteristic curve analysis. Chi-square test was conducted to compare the distribution of ICDAS-II and DIAGNOdent scores. Inter-observer agreement between assessors was evaluated using Cohen's kappa test.

Results

In the current study, DIAGNOdent had an overall accuracy of 84.45% with sensitivity and specificity of 87.58% and 96.87%, respectively, and +PV and -PV of 97.7% and 83.9%, respectively, when score 0 represented sound tooth surface, while scores 1 and 2 were considered clinically noncavitated carious lesions. Moreover, when only ICDAS score 1 was considered representing first visual change in enamel, DIAGNOdent had an accuracy of 74.15% with sensitivity and specificity of 83.53% and 90.62%, respectively, and +PV and -PV of 93% and 78.6%, respectively. In the present study, when only ICDAS score 2 was considered representing distinct visual change in enamel, DIAGNOdent had an accuracy of 100% with sensitivity and specificity of 100% and 100%, respectively, and +PV and -PV of 100% and 100%, respectively.

Conclusions

The overall performance of DIAGNOdent was equivalent to the visual inspection using ICDAS-II. DIAGNOdent might be considered a useful adjunctive device for detection and monitoring development of noncavitated carious lesions on facial smooth surfaces.

Diagnostic Value of Fluorescence Methods, Visual Inspection and Photographic Visual Examination in Initial Caries Lesion: A Systematic Review and Meta-Analysis

This systematic review and meta-analysis aimed to investigate the efficacy of fluorescence-based methods, visual inspections, and photographic visual examinations in initial caries detection. A literature search was undertaken in the PubMed and Cochrane databases. Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) guidelines were followed, and eligible articles published from 1 January 2009 to 30 October 2019 were included if they met the following criteria: they (1) assessed the accuracy of methods of detecting initial tooth caries lesions on occlusal, proximal, or smooth surfaces in both primary and permanent teeth (in clinical); (2) used a reference standard; (3) reported data regarding the sample size, prevalence of initial tooth caries, and accuracy of the methods. Data collection and extraction, quality assessment, and data analysis were conducted according to Cochrane standards Quality Assessment of Diagnostic Accuracy Studies-2. Statistical analyses were performed using Review Manager 5.3 and STATA 14.0. A total of 12 eligible articles were included in the meta-analysis. The results showed that the sensitivity and specificity of fluorescence-based methods were 80% and 80%, respectively; visual inspection was measured at 80% and 75%, respectively; photographic visual examination was measured at 67% and 79%, respectively. We found that the visual method and the fluorescence method were reliable for laboratory use to detect early-stage caries with equivalent accuracy.

Transillumination and optical coherence tomography for the detection and diagnosis of enamel caries

BACKGROUND

Caries is one of the most prevalent and preventable conditions worldwide. If identified early enough then non-invasive techniques can be applied, and therefore this review focusses on early caries involving the enamel surface of the tooth. The cornerstone of caries detection and diagnosis is a visual and tactile dental examination, although alternative approaches are available. These include illumination-based devices that could potentially support the dental examination. There are three categories of illumination devices that exploit various methods of application and interpretation, each primarily defined by different wavelengths, optical coherence tomography (OCT), near-infrared (NIR), and fibre-optic technology, which incorporates more recently developed digital fibre optics (FOTI/DIFOTI).

OBJECTIVES

To estimate the diagnostic test accuracy of different illumination tests for the detection and diagnosis of enamel caries in children or adults. We also planned to explore the following potential sources of heterogeneity: in vitro or in vivo studies with different reference standards; tooth surface (occlusal, proximal, smooth surface, or adjacent to a restoration); single or multiple sites of assessment on a tooth surface; and the prevalence of caries into dentine.

SEARCH METHODS

Cochrane Oral Health's Information Specialist undertook a search of the following databases: MEDLINE Ovid (1946 to 15 February 2019); Embase Ovid (1980 to 15 February 2019); US National Institutes of Health Ongoing Trials Register (ClinicalTrials.gov, to 15 February 2019); and the World Health Organization International Clinical Trials Registry Platform (to 15 February 2019). We studied reference lists as well as published systematic review articles.

SELECTION CRITERIA

We included diagnostic accuracy study designs that compared the use of illumination-based devices with a reference standard (histology, enhanced visual examination with or without radiographs, or operative excavation). These included prospective studies that evaluated the diagnostic accuracy of a single index test and studies that directly compared two or more index tests. Both in vitro and in vivo studies of primary and permanent teeth were eligible for inclusion. We excluded studies that explicitly recruited participants with caries into dentine or frank cavitation. We also excluded studies that artificially created carious lesions and those that used an index test during the excavation of dental caries to ascertain the optimum depth of excavation.

DATA COLLECTION AND ANALYSIS

Two review authors extracted data independently and in duplicate using a standardised data extraction form and quality assessment based on QUADAS-2 specific to the clinical context. Estimates of diagnostic accuracy were determined using the bivariate hierarchical method to produce summary points of sensitivity and specificity with 95% confidence regions. The comparative accuracy of different illumination devices was conducted based on indirect and direct comparisons between methods. Potential sources of heterogeneity were pre-specified and explored visually and more formally through meta-regression.

MAIN RESULTS

We included 24 datasets from 23 studies that evaluated 16,702 tooth surfaces. NIR was evaluated in 6 datasets (673 tooth surfaces), OCT in 10 datasets (1171 tooth surfaces), and FOTI/DIFOTI in 8 datasets (14,858 tooth surfaces). The participant selection domain had the largest number of studies judged at high risk of bias (16 studies). Conversely, for the index test, reference standard, and flow and timing domains the majority of studies were judged to be at low risk of bias (16, 12, and 16 studies respectively). Concerns regarding the applicability of the evidence were judged as high or unclear for all domains. Notably, 14 studies were judged to be of high concern for participant selection, due to selective participant recruitment, a lack of independent examiners, and the use of an in vitro study design. The summary estimate across all the included illumination devices was sensitivity 0.75 (95% confidence interval (CI) 0.62 to 0.85) and specificity 0.87 (95% CI 0.82 to 0.92), with a diagnostic odds ratio of 21.52 (95% CI 10.89 to 42.48). In a cohort of 1000 tooth surfaces with a prevalence of enamel caries of 57%, this would result in 142 tooth surfaces being classified as disease free when enamel caries was truly present (false negatives), and 56 tooth surfaces being classified as diseased in the absence of enamel caries (false positives). A formal comparison of the accuracy according to device type indicated a difference in sensitivity and/or specificity (Chi²(4) = 34.17, P < 0.01). Further analysis indicated a difference in the sensitivity of the different devices (Chi²(2) = 31.24, P < 0.01) with a higher sensitivity of 0.94 (95% CI 0.88 to 0.97) for OCT compared to NIR 0.58 (95% CI 0.46 to 0.68) and FOTI/DIFOTI 0.47 (95% CI 0.35 to 0.59), but no meaningful difference in specificity (Chi²(2) = 3.47, P = 0.18). In light of these results, we planned to formally assess potential sources of heterogeneity according to device type, but due to the limited number of studies for each device type we were unable to do so. For interpretation, we presented the coupled forest plots for each device type according to the potential source of heterogeneity. We rated the certainty of the evidence as low and downgraded two levels in total due to avoidable and unavoidable study limitations in the design and conduct of studies, indirectness arising from the in vitro studies, and imprecision of the estimates.

AUTHORS' CONCLUSIONS

Of the devices evaluated, OCT appears to show the most potential, with superior sensitivity to NIR and fibre-optic devices. Its benefit lies as an add-on tool to support the conventional oral examination to confirm borderline cases in cases of clinical uncertainty. OCT is not currently available to the general dental practitioner, and so further research and development are necessary. FOTI and NIR are more readily available and easy to use; however, they show limitations in their ability to detect enamel caries but may be considered successful in the identification of sound teeth. Future studies should strive to avoid research waste by ensuring that recruitment is conducted in such a way as to minimise selection bias and that studies are clearly and comprehensively reported. In terms of applicability, any future studies should be undertaken in a clinical setting that is reflective of the complexities encountered in caries assessment within the oral cavity.

Fluorescence devices for the detection of dental caries

BACKGROUND

Caries is one of the most prevalent and preventable conditions worldwide. If identified early enough then non-invasive techniques can be applied, and therefore this review focusses on early caries involving the enamel surface of the tooth. The cornerstone of caries detection is a visual and tactile dental examination, however alternative methods of detection are available, and these include fluorescence-based devices. There are three categories of fluorescence-based device each primarily defined by the different wavelengths they exploit; we have labelled these groups as red, blue, and green fluorescence. These devices could support the visual examination for the detection and diagnosis of caries at an early stage of decay.

OBJECTIVES

Our primary objectives were to estimate the diagnostic test accuracy of fluorescence-based devices for the detection and diagnosis of enamel caries in children or adults. We planned to investigate the following potential sources of heterogeneity: tooth surface (occlusal, proximal, smooth surface or adjacent to a restoration); single point measurement devices versus imaging or surface assessment devices; and the prevalence of more severe disease in each study sample, at the level of caries into dentine.

SEARCH METHODS

Cochrane Oral Health's Information Specialist undertook a search of the following databases: MEDLINE Ovid (1946 to 30 May 2019); Embase Ovid (1980 to 30 May 2019); US National Institutes of Health Ongoing Trials Register (ClinicalTrials.gov, to 30 May 2019); and the World Health Organization International Clinical Trials Registry Platform (to 30 May 2019). We studied reference lists as well as published systematic review articles.

SELECTION CRITERIA

We included diagnostic accuracy study designs that compared a fluorescence-based device with a reference standard. This included prospective studies that evaluated the diagnostic accuracy of single index tests and studies that directly compared two or more index tests. Studies that explicitly recruited participants with caries into dentine or frank cavitation were excluded.

DATA COLLECTION AND ANALYSIS

Two review authors extracted data independently using a piloted study data extraction form based on the Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2). Sensitivity and specificity with 95% confidence intervals (CIs) were reported for each study. This information has been displayed as coupled forest plots and summary receiver operating characteristic (SROC) plots, displaying the sensitivity-specificity points for each study. We estimated diagnostic accuracy using hierarchical summary receiver operating characteristic (HSROC) methods. We reported sensitivities at fixed values of specificity (median 0.78, upper quartile 0.90).

MAIN RESULTS

We included a total of 133 studies, 55 did not report data in the 2 x 2 format and could not be included in the meta-analysis. 79 studies which provided 114 datasets and evaluated 21,283 tooth surfaces were included in the meta-analysis. There was a high risk of bias for the participant selection domain. The index test, reference standard, and flow and timing domains all showed a high proportion of studies to be at low risk of bias. Concerns regarding the applicability of the evidence were high or unclear for all domains, the highest proportion being seen in participant selection. Selective participant recruitment, poorly defined diagnostic thresholds, and in vitro studies being non-generalisable to the clinical scenario of a routine dental examination were the main reasons for these findings. The dominance of in vitro studies also means that the information on how the results of these devices are used to support diagnosis, as opposed to pure detection, was extremely limited. There was substantial variability in the results which could not be explained by the different devices or dentition or other sources of heterogeneity that we investigated. The diagnostic odds ratio (DOR) was 14.12 (95% CI 11.17 to 17.84). The estimated sensitivity, at a fixed median specificity of 0.78, was 0.70 (95% CI 0.64 to 0.75). In a hypothetical cohort of 1000 tooth sites or surfaces, with a prevalence of enamel caries of 57%, obtained from the included studies, the estimated sensitivity of 0.70 and specificity of 0.78 would result in 171 missed tooth sites or surfaces with enamel caries (false negatives) and 95 incorrectly classed as having early caries (false positives). We used meta-regression to compare the accuracy of the different devices for red fluorescence (84 datasets, 14,514 tooth sites), blue fluorescence (21 datasets, 3429 tooth sites), and green fluorescence (9 datasets, 3340 tooth sites) devices. Initially, we allowed threshold, shape, and accuracy to vary according to device type by including covariates in the model. Allowing consistency of shape, removal of the covariates for accuracy had only a negligible effect (Chi2 = 3.91, degrees of freedom (df) = 2, P = 0.14). Despite the relatively large volume of evidence we rated the certainty of the evidence as low, downgraded two levels in total, for risk of bias due to limitations in the design and conduct of the included studies, indirectness arising from the high number of in vitro studies, and inconsistency due to the substantial variability of results.

AUTHORS' CONCLUSIONS

There is considerable variation in the performance of these fluorescence-based devices that could not be explained by the different wavelengths of the devices assessed, participant, or study characteristics. Blue and green fluorescence-based devices appeared to outperform red fluorescence-based devices but this difference was not supported by the results of a formal statistical comparison. The evidence base was considerable, but we were only able to include 79 studies out of 133 in the meta-analysis as estimates of sensitivity or specificity values or both could not be extracted or derived. In terms of applicability, any future studies should be carried out in a clinical setting, where difficulties of caries assessment within the oral cavity include plaque, staining, and restorations. Other considerations include the potential of fluorescence devices to be used in combination with other technologies and comparative diagnostic accuracy studies.

Combined Near-Infrarred Light Transillumination and Direct Digital Radiography Increases Diagnostic In Approximal Caries

The objective of this study was to evaluate the clinical ability of Near-Infrared Light-Transillumination (NILT) for approximal dentinal caries detection and to compare with direct digital-radiography (DDR), as well as to determine whether the combination of both techniques improves the diagnostic capacity of the lesions. From 88 patients (over 18 years), 138 posterior teeth (76 molars and 62 premolars), that had approximal caries reached into dentine determined by DDR, were included. Lesion extension and DDR images were scored as follows: D0 = sound surface, D1/D2 = caries restricted to the outer/inner-half of the enamel, and D3/4 = caries restricted to the outer/inner-half of the dentin. Opening of the approximal surface using 0.5 mm-in-diameter diamond-bar was used as gold-standard. The lesion extension was then determined by the following criteria: no dentinal caries (D0/1/2) or dentinal caries (D3/4). Seventy-one lesions were D3 and 67 lesions were D4. Sensitivities of overall/D3/D4 were 98.0/95.7/100.0 (NILT) and 100/100/100 (DDR), respectively. Correlations with gold-standard were 0.92 (NILT) and 0.42 (DDR), respectively. The correlation increased to 0.97 (p = 0.045) on combining NILT and DDR. There was no difference in sensitivity between the methods (p > 0.05); but was differences in the correlation (p < 0.01). It can be concluded that NILT showed sensitivity similar to that of DDR and higher correlation than DDR for approximal dentinal caries detection. Accordingly, it may be used to monitor the progression of caries without exposing the patient to ionizing radiation, this being of particular interest in growing patients and in pregnant women. In this respect, NILT can be an effective diagnostic tool adjunctive to bitewing radiographs in detecting approximal dentinal caries. The combination of NILT and DDR represents an increase in the diagnosis of approximal lesions The proposed diagnostic protocol comprises visual examination, followed by NILT and DDR only if the former technique detects approximal caries.

Near-infrared multispectral endoscopic imaging of deep artificial interproximal lesions in extracted teeth

BACKGROUND AND OBJECTIVE

A safer alternative method to radiographic imaging is needed. We present a multispectral near-infrared scanning fiber endoscope (nirSFE) for dental imaging which is designed to be the smallest imaging probe with near-infrared (NIR) imaging (1200-2000 nm).

MATERIALS AND METHODS

The prototype nirSFE is designed for wide-field forward viewing of scanned laser illumination at 1310, 1460, or 1550 nm. Artificial lesions with varying sizes and locations were prepared on proximal surfaces of extracted human teeth to examine capability and limitation of this new dental imaging modality. Nineteen artificial interproximal lesions and several natural occlusal lesions on extracted teeth were imaged with nirSFE, OCT, and microCT.

RESULTS

Our nirSFE system has a flexible shaft as well as a probe tip with diameter of 1.6 mm and a rigid length of 9 mm. The small form factor and multispectral NIR imaging capability enables multiple viewing angles and reliable detection of lesions that can extend into the dentin. Among nineteen artificial interproximal lesions, the nirSFE reflectance imaging operating at 1460-nm and OCT operating at 1310-nm scanned illumination exhibited high sensitivity for interproximal lesions that were closer to occlusal surface. Diagnosis from a non-blinded trained user by looking at real-time occlusal-side nirSFE videos indicate true positive rate of 78.9%. There were no false positives.

CONCLUSIONS

This study demonstrates that nirSFE may be used for detecting occlusal lesions and interproximal lesions located less than 4 mm under the occlusal surface. Major advantages of this imaging system include multiple viewing angles due to flexibility and small form factor, as well as the ability to capture real-time video. The multispectral nirSFE has the potential to be employed as a low-cost dental camera for detecting dental lesions without exposure to ionizing radiation. Lasers Surg. Med. 51:459-465, 2019. © 2019 Wiley Periodicals, Inc.

[Ex vivo assessment of the potency of optical coherence tomography for the detection of early occlusal caries]

OBJECTIVE

This study aimed to evaluate the potency of optical coherence tomography (OCT) to detect early occlusal caries compared with clinical visual examination.

METHODS

Approximately 97 sites of occlusal fissures on 77 extracted accessional human teeth were scored by three examiners using conventional visual examination and OCT. Results of histological examination on these sites obtained by polarimicroscope served as a gold standard to analyze the sensitivity (SE), specificity (SP), positive predictive value (PPV), and negative predictive value (NPV). Results of the area under receiver operating characteristic (ROC) curve (AUC) by visual examination and OCT were also analyzed. The Spearman's rank correlation coefficient with histology and the inter-examiner reliability were compared.

RESULTS

For sites of enamel demineralization limited to the outer 1/2 of the enamel layer, the detection rate of OCT (14/25) was obviously higher than that of the clinical and visual examination (3/25). SE, SP, PPV and NPV of OCT during diagnosis of the early occlusal caries (0.83, 0.64, 0.87 and 0.57) were higher than that of the visual examination (0.79, 0.60, 0.85 and 0.50). AUC (95%CI) of OCT and the visual examination were 0.737 (0.569-0.822) and 0.696 (0.614-0.859), respectively. No statistically significant difference was observed between the results. Results of OCT correlated well with histology (r=0.559, P<0.05). The inter-examiner reliability of OCT was medium.

CONCLUSIONS

OCT can accurately detect early occlusal lesions atraumatically with high sensitivity and effectiveness. OCT exhibits the potential of auxiliary clinical diagnosis enhancing detection rate and finally implementing early diagnosis and early intervention of early occlusal lesions in clinical practice.