Accuracy of different approaches for detecting proximal root caries lesions in vitro

OBJECTIVES

The objective was to evaluate the diagnostic accuracy of radiographic evaluation (XR), visual-tactile assessment (VT), laser-fluorescence (LF) (DIAGNOdent Pen/KaVo), and near-infrared-light transillumination (NILT) (DIAGNOcam/KaVo) on proximal root caries lesions in vitro.

METHODS

Two-hundred extracted permanent premolars and molars with and without proximal root caries lesions were allocated to 50 diagnostic models simulating the proximal contacts between teeth and mounted in a phantom dummy head. Two independent examiners used the diagnostic approaches to detect any or advanced root caries lesions, with histologic evaluation of the lesions serving as reference. Receiver operating characteristic (ROC) curves were employed, and sensitivity, specificity, and the area under the ROC curve (AUC) are calculated. Significant differences in mean AUCs between approaches were assumed if p < 0.05 (two-sample t-test).

RESULTS

NILT was not applicable for proximal root caries detection. The sensitivity/specificity to detect any lesions was 0.81/0.63 for XR, 0.76/0.88 for VT and 0.81/0.95 for LF, and the sensitivity/specificity to detect advanced lesions was 0.43/0.94 for XR, 0.66/0.99 for VT, and 0.83/0.78 for LF, respectively. For both, any and advanced root caries lesions, mean AUCs for LF and VT were significantly higher compared to XR (p < 0.05). For any root caries lesions, LF was significantly more accurate than VT (p = 0.01), but not for advanced root caries lesions (p = 0.59).

CONCLUSIONS

Under the in vitro conditions chosen, LF and VT were more accurate than XR to detect proximal root caries lesions, with LF being particularly useful for initial lesion stages.

CLINICAL RELEVANCE

LF might be a useful diagnostic aid for proximal root caries diagnosis. Clinical studies are necessary to corroborate the findings.

The association between molar incisor hypomineralization and oral health-related quality of life: a cross-sectional study

Objectives

We aimed to assess the association between molar incisor hypomineralization (MIH) and the oral health-related quality of life (OHRQoL) in a group of 7- to 14-year-old children in Berlin, Germany.

Materials and methods

The cross-sectional study consisted of a consecutive sample of 317 children, aged 7–14 years (49% girls, 51% boys; mean age, 8.71). Data were collected between June 2018 and December 2019. MIH was diagnosed using the criteria of the European Academy of Paediatric Dentistry. OHRQoL was assessed using the German 19-item version of the Child Oral Health Impact Profile (COHIP-G19). Differences in COHIP-19 summary scores between controls without MIH and MIH patients and with regards to MIH severity were tested for statistical significance using t test and analysis of variance (ANOVA), respectively.

Results

Data were obtained for 217 untreated MIH patients and 100 controls. OHRQoL of MIH patients was significantly more impaired than of controls indicated by COHIP-19 mean scores (60.9 ± 10.7 vs. 67.9 ± 7.8; p < 0.001). Patients with severe MIH (59.6 ± 11.0) reported significantly worse OHRQoL than patients with mild MIH (63.6 ± 9.1; p = 0.013).

Conclusions

MIH has a significant negative impact on the children’s OHRQoL. Patients with severe MIH experience a greater negative impact on OHRQoL than those diagnosed with mild MIH.

Clinical significance

MIH is one of the major dental problems of our time; pediatric dentists should be aware of its impact on the OHRQoL of the patient.

Preventing and Arresting Primary Tooth Enamel Lesions Using Self- Assembling Peptide P11-4 In Vitro

Objectives

To evaluate self-assembling peptides (SAP) for caries prevention and arrest in primary tooth enamel in vitro.

Materials and Methods

Overall, 180 extracted primary teeth were used. In the prevention experiment (n = 20 samples per group), self-assembling peptide for prevention (SAPP), fluoride varnish/mouthwash (FV/FMW), casein-phosphopeptide amorphous-calcium phosphate (CPP-ACP), and nanohydroxyapatite (nHA) were applied. Samples were subjected to a demineralizing pH cycling for 14 days. In the arrest experiment (n = 15/group), 60 samples were pre-demineralized; induced lesions were treated using self-assembling peptide for repair (SAPR), FV, CPP-ACP plus fluoride, and resin infiltration (RI) and submitted to pH cycling. Mineral loss and its differences as well as lesion depth were determined using transversal microradiography. Numerical data were tested for normality using Shapiro-Wilk's test and were compared using Kruskal-Wallis test followed by pairwise comparisons utilizing multiple Mann-Whitney U tests with Bonferroni correction. The significance level was set at P < 0.05 within all tests.

Results

FV (median: -46.3 [interquartile range: 175.52] vol% × µm) and FMW (-33.35 [124.65] vol% × µm) prevented caries significantly more effectively than all other groups (P < 0.001), which did not show significant preventive effects. RI (median: 4949.70 [1637.20] vol% × µm) and FV (median = 6076.05 [5190.08] vol% × µm) arrested lesions, whereas SAPR and CPP-ACPF did not show such arrest.

Conclusions

FV and FMW showed the largest caries-preventive effect, whereas RI and FV arrested lesion progression in primary tooth enamel in vitro.

Dental Students' Knowledge, Attitudes and Beliefs Regarding Molar Incisor Hypomineralization (MIH): A Survey in Vienna, Austria

Purpose

Molar incisor hypomineralization (MIH) is a frequently encountered dental condition worldwide. The aim of this study was to evaluate the knowledge of Austrian dental students about MIH.

Methods

A validated paper-based survey was distributed among 100 final year dental students at the Medical University of Vienna to assess their knowledge, perception and believes regarding MIH. One hundred dental students at the University Dental Clinic of Vienna were included in the study. At the time of the survey, they were in their 11th or 12th semester (final year). Information about the knowledge on diagnosis and prevalence, and attitudes and beliefs around etiology and management was collected.

Results

All students replied, 94 questionnaires could be included in the analysis (50% female). 98% of the students were familiar with MIH and 86% were aware with the clinical features of MIH. Only 13% reported clinical ability to identify MIH. Regarding the etiology of MIH, most of the students (69%) named "genetic factors" as the most frequent etiological factor.

Conclusion

Dental students showed great interest in the topic of MIH, as the majority of all respondents were in favor of deepening their knowledge regarding this topic. The students should be offered opportunities to do this both during and after their studies.

Knowledge, attitudes, and beliefs regarding molar incisor hypomineralization (MIH) amongst German dental students

BACKGROUND

Knowledge of molar incisor hypomineralization (MIH) has relevance for paediatric dentists.

AIM

To assess final-year German dental students' knowledge, attitudes, and beliefs regarding MIH.

MATERIALS AND METHODS

A previously validated questionnaire was posted to the 31 German dental schools. Demographic covariates as well as knowledge regarding diagnosis and prevalence, and attitudes and beliefs around aetiology and management were collected.

RESULTS

Twenty-two (71%) dental schools responded and a total of 877 students participated. Most (97%) were familiar with MIH and 88% were aware of the diagnostic criteria for MIH; however, only 42% knew how to implement them. One-third were able to identify MIH and 16% reported diagnostic confidence when doing so; 90% assumed the MIH prevalence to be <10%. Two-thirds of the respondents implicated genetic components as the main aetiological factor of MIH. Resin composite (60%) and preformed metal crowns (46%) were the dental materials most often suggested for restorative management. Almost all (98%) respondents were interested in receiving more clinical training.

CONCLUSION

German students were familiar with MIH; however, they reported low levels of knowledge and confidence regarding its prevalence and diagnosis. Standardized nationwide, up-to-date curricula should be implemented to educate future dentists in Germany.

Generalizability of Deep Learning Models for Caries Detection in Near-Infrared Light Transillumination Images

OBJECTIVES

The present study aimed to train deep convolutional neural networks (CNNs) to detect caries lesions on Near-Infrared Light Transillumination (NILT) imagery obtained either in vitro or in vivo and to assess the models' generalizability.

METHODS

In vitro, 226 extracted posterior permanent human teeth were mounted in a diagnostic model in a dummy head. Then, NILT images were generated (DIAGNOcam, KaVo, Biberach), and images were segmented tooth-wise. In vivo, 1319 teeth from 56 patients were obtained and segmented similarly. Proximal caries lesions were annotated pixel-wise by three experienced dentists, reviewed by a fourth dentist, and then transformed into binary labels. We trained ResNet classification models on both in vivo and in vitro datasets and used 10-fold cross-validation for estimating the performance and generalizability of the models. We used GradCAM to increase explainability.

RESULTS

The tooth-level prevalence of caries lesions was 41% in vitro and 49% in vivo, respectively. Models trained and tested on in vivo data performed significantly better (mean ± SD accuracy: 0.78 ± 0.04) than those trained and tested on in vitro data (accuracy: 0.64 ± 0.15; p < 0.05). When tested in vitro, the models trained in vivo showed significantly lower accuracy (0.70 ± 0.01; p < 0.01). Similarly, when tested in vivo, models trained in vitro showed significantly lower accuracy (0.61 ± 0.04; p < 0.05). In both cases, this was due to decreases in sensitivity (by -27% for models trained in vivo and -10% for models trained in vitro).

CONCLUSIONS

Using in vitro setups for generating NILT imagery and training CNNs comes with low accuracy and generalizability.

CLINICAL SIGNIFICANCE

Studies employing in vitro imagery for developing deep learning models should be critically appraised for their generalizability. Applicable deep learning models for assessing NILT imagery should be trained on in vivo data.

Detecting white spot lesions on dental photography using deep learning: A pilot study

OBJECTIVES

We aimed to apply deep learning to detect white spot lesions in dental photographs.

METHODS

Using 434 photographic images of 51 patients, a dataset of 2781 cropped tooth segments was generated. Pixelwise annotations of sound enamel as well as fluorotic, carious or other types of hypomineralized lesions were generated by experts and assessed by an independent second reviewer. The union of the reviewed annotations were used to segment the hard tissues (region-of-interest, ROI) of each image. SqueezeNet was employed for modelling. We trained models to detect (1) any white spot lesions, (2) fluorotic lesions and (3) other-than-fluorotic lesions. Modeling was performed on both the cropped and the ROI images and using ten-times repeated five-fold cross-validation. Feature visualization was applied to visualize salient areas.

RESULTS

Lesion prevalence was 37 %; the majority of lesions (24 %) were fluorotic. None of the metrics differed significantly between the models trained on cropped and ROI imagery (p > 0.05/t-test). Mean accuracies ranged between 0.81-0.84, without significant differences between models trained to detect any, fluorotic or other-than-fluorotic lesions (p > 0.05). Specificities were 0.85-0.86; sensitivities were lower (0.58-0.66). Models to detect any lesions showed positive/negative predictive values (PPV/NPV) between 0.77-0.80, those to detect fluorotic lesions 0.67 (PPV) to 0.86 (NPV), and those to detect other-than-fluorotic lesions 0.46 (PPV) to 0.93 (NPV). Light reflections were the main reason for false positive detections.

CONCLUSIONS

Deep learning showed satisfying accuracy to detect white spot lesions, particularly fluorosis. Some models showed limited stability given the small sample available.

CLINICAL SIGNIFICANCE

Deep learning is suitable for automated classification of retro- or prospectively collected imagery and may assist practitioners in discriminating white spot lesions. Future studies should expand the scope into more granular multi-class detections on a larger and more generalizable dataset.

Cost-effectiveness of Artificial Intelligence for Proximal Caries Detection

Artificial intelligence (AI) can assist dentists in image assessment, for example, caries detection. The wider health and cost impact of employing AI for dental diagnostics has not yet been evaluated. We compared the cost-effectiveness of proximal caries detection on bitewing radiographs with versus without AI. U-Net, a fully convolutional neural network, had been trained, validated, and tested on 3,293, 252, and 141 bitewing radiographs, respectively, on which 4 experienced dentists had marked carious lesions (reference test). Lesions were stratified for initial lesions (E1/E2/D1, presumed noncavitated, receiving caries infiltration if detected) and advanced lesions (D2/D3, presumed cavitated, receiving restorative care if detected). A Markov model was used to simulate the consequences of true- and false-positive and true- and false-negative detections, as well as the subsequent decisions over the lifetime of patients. A German mixed-payers perspective was adopted. Our health outcome was tooth retention years. Costs were measured in 2020 euro. Monte-Carlo microsimulations and univariate and probabilistic sensitivity analyses were conducted. The incremental cost-effectiveness ratio (ICER) and the cost-effectiveness acceptability at different willingness-to-pay thresholds were quantified. AI showed an accuracy of 0.80; dentists’ mean accuracy was significantly lower at 0.71 (minimum–maximum: 0.61–0.78, P < 0.05). AI was significantly more sensitive than dentists (0.75 vs. 0.36 [0.19–0.65]; P = 0.006), while its specificity was not significantly lower (0.83 vs. 0.91 [0.69–0.98]; P > 0.05). In the base-case scenario, AI was more effective (tooth retention for a mean 64 [2.5%–97.5%: 61–65] y) and less costly (298 [244–367] euro) than assessment without AI (62 [59–64] y; 322 [257–394] euro). The ICER was −13.9 euro/y (i.e., AI saved money at higher effectiveness). In the majority (>77%) of all cases, AI was less costly and more effective. Applying AI for caries detection is likely to be cost-effective, mainly as fewer lesions remain undetected. Notably, this cost-effectiveness requires dentists to manage detected early lesions nonrestoratively.

Outcome and comparator choice in molar incisor hypomineralisation (MIH) intervention studies: a systematic review and social network analysis

OBJECTIVES

Outcome and comparator choice strongly determine the validity and implementation of clinical trial results. We aimed to assess outcome and comparator choice in intervention studies on molar incisor hypomineralisation (MIH) using systematic review and social network analysis (SNA).

DESIGN AND DATA SOURCES

Medline, Embase, Cochrane Central, Google Scholar, opengrey.eu as well as DRKS.de and Clinicaltrials.gov were searched for MIH intervention studies. The search covered the period from 1980 to 2019.

ELIGIBILITY CRITERIA

Clinical single-arm/multiarm, controlled/uncontrolled studies reporting on the management of MIH were included. Reported outcomes and comparators were extracted and categorised. SNA was used to evaluate comparator choice and the resulting trial networks.

DATA EXTRACTION

Of the 7979 identified records, 100 were evaluated in full text and 35 studies (17 randomised controlled trials, 14 prospective and 4 retrospective cohort studies) were included.

RESULTS

In total, 2124 patients with a mean age of 11 years (min/max 6/70 years) were included. Outcomes fell in one of 11 different outcome categories: restoration success, aesthetic improvement, pain/hypersensitivity/discomfort, mineral gain, space management, anaesthesia effectiveness, preventive success, efficiency, quality of life, gingival and periodontal health and patient satisfaction. Comparators were mainly restorative interventions (17 studies), remineralisation (3), treatment of hypersensitivity (10), aesthetic interventions (5) and orthodontic interventions (3). Two highly clustered comparator networks emerged; many interventions were not robustly linked to these networks.

CONCLUSIONS

MIH intervention studies recorded both clinically centred and patient-centred outcomes. Core outcome set development should consider these and supplement them with outcomes on, for example, applicability. The high number of compared interventions tested in only few studies and our SNA results implicate that the current evidence may not be robust.

Deep Learning for the Radiographic Detection of Apical Lesions

INTRODUCTION

We applied deep convolutional neural networks (CNNs) to detect apical lesions (ALs) on panoramic dental radiographs.

METHODS

Based on a synthesized data set of 2001 tooth segments from panoramic radiographs, a custom-made 7-layer deep neural network, parameterized by a total number of 4,299,651 weights, was trained and validated via 10 times repeated group shuffling. Hyperparameters were tuned using a grid search. Our reference test was the majority vote of 6 independent examiners who detected ALs on an ordinal scale (0, no AL; 1, widened periodontal ligament, uncertain AL; 2, clearly detectable lesion, certain AL). Metrics were the area under the receiver operating characteristic curve (AUC), sensitivity, specificity, and positive/negative predictive values. Subgroup analysis for tooth types was performed, and different margins of agreement of the reference test were applied (base case: 2; sensitivity analysis: 6).

RESULTS

The mean (standard deviation) tooth level prevalence of both uncertain and certain ALs was 0.16 (0.03) in the base case. The AUC of the CNN was 0.85 (0.04). Sensitivity and specificity were 0.65 (0.12) and 0.87 (0.04,) respectively. The resulting positive predictive value was 0.49 (0.10), and the negative predictive value was 0.93 (0.03). In molars, sensitivity was significantly higher than in other tooth types, whereas specificity was lower. When only certain ALs were assessed, the AUC was 0.89 (0.04). Increasing the margin of agreement to 6 significantly increased the AUC to 0.95 (0.02), mainly because the sensitivity increased to 0.74 (0.19).

CONCLUSIONS

A moderately deep CNN trained on a limited amount of image data showed satisfying discriminatory ability to detect ALs on panoramic radiographs.

Visual and radiographic caries detection: a tailored meta-analysis for two different settings, Egypt and Germany

Background

Diagnostic meta-analyses on caries detection methods should assist practitioners in their daily practice. However, conventional meta-analysis estimates may be inapplicable due to differences in test conduct, applied thresholds and assessed population between settings. Our aim was to demonstrate the impact of tailored meta-analysis of visual and radiographic caries detection to different settings using setting-specific routine data.

Methods

Published systematic reviews and meta-analyses on the accuracy of visual and radiographic caries detection were used. In two settings (a private practice in Germany and a public health clinic in Egypt), routine data of a total of 100 (n = 50/practice) consecutive 12–14 year-olds were collected. Test-positive rates of visual and radiographic detection for initial and advanced carious lesions on occlusal or proximal surfaces of molars were used to tailor meta-analyses. If prevalence data were available, these were also used for tailoring.

Results

From the original reviews, 210 and 100 heterogeneous studies on visual and radiographic caries detection were included in our meta-analyses. For radiographic detection, sensitivity and specificity estimates derived from conventional and tailored meta-analysis were similar. For visual detection of advanced occlusal carious lesions, the conventional meta-analysis yielded a sensitivity and specificity (95% CI) of 64.6% (57–71) and 90.9% (88–93), whereas the tailored estimates for Egypt were 75.1% (70–81) and 84.9% (82–89), respectively, and 43.7% (37–51) and 96.5% (95–97) for Germany, respectively.

Conclusion

Conventional test accuracy meta-analyses may yield aggregate estimates which are inapplicable to specific settings. Routine data may be used to produce a meta-analysis estimate which is tailored to the setting and thereby improving its applicability.

Electronic supplementary material

The online version of this article (10.1186/s12903-018-0561-z) contains supplementary material, which is available to authorized users.

Oral manifestations, dental management, and a rare homozygous mutation of the PRDM12 gene in a boy with hereditary sensory and autonomic neuropathy type VIII: a case report and review of the literature

BACKGROUND

Hereditary sensory and autonomic neuropathy type VIII is a rare autosomal recessive inherited disorder. Chen et al. recently identified the causative gene and characterized biallelic mutations in the PR domain-containing protein 12 gene, which plays a role in the development of pain-sensing nerve cells. Our patient's family was included in Chen and colleagues' study. We performed a literature review of the PubMed library (January 1985 to December 2016) on hereditary sensory and autonomic neuropathy type I to VIII genetic disorders and their orofacial manifestations. This case report is the first to describe the oral manifestations, and their treatment, of the recently discovered hereditary sensory and autonomic neuropathy type VIII in the medical and dental literature.

CASE PRESENTATION

We report on the oral manifestations and dental management of an 8-month-old white boy with hereditary sensory and autonomic neuropathy-VIII over a period of 16 years. Our patient was homozygous for a mutation of PR domain-containing protein 12 gene and was characterized by insensitivity to pain and thermal stimuli, self-mutilation behavior, reduced sweat and tear production, absence of corneal reflexes, and multiple skin and bone infections. Oral manifestations included premature loss of teeth, associated with dental traumata and self-mutilation, severe soft tissue injuries, dental caries and submucosal abscesses, hypomineralization of primary teeth, and mandibular osteomyelitis.

CONCLUSIONS

The lack of scientific knowledge on hereditary sensory and autonomic neuropathy due to the rarity of the disease often results in a delay in diagnosis, which is of substantial importance for the prevention of many complications and symptoms. Interdisciplinary work of specialized medical and dental teams and development of a standardized treatment protocols are essential for the management of the disease. There are many knowledge gaps concerning the management of patients with hereditary sensory and autonomic neuropathy-VIII, therefore more research on an international basis is needed.

Managing molar-incisor hypomineralization: A systematic review

OBJECTIVES

We systematically reviewed treatment modalities for MIH-affected molars and incisors.

DATA

Trials on humans with ≥1 MIH molar/incisor reporting on various treatments were included. Two authors independently searched and extracted records. Sample-size-weighted annual failure rates were estimated where appropriate. The risk of bias was assessed using the Newcastle-Ottawa scale.

SOURCES

Electronic databases (PubMed, Embase, Cochrane CENTRAL, Google Scholar) were screened, and hand searches and cross-referencing performed.

STUDY SELECTION

Fourteen (mainly observational) studies were included. Ten trials (381 participants) investigated MIH-molars, four (139) MIH-incisors. For molars, remineralization, restorative or extraction therapies had been assessed. For restorative approaches, mean (SD) annual failure rates were highest for fissure sealants (12[6]%) and glass-ionomer restorations (12[2]%), and lowest for indirect restorations (1[3]%), preformed metal crowns (1.3 [2.1]%) and composite restorations (4[3]%). Ony study assessed extraction of molars in young patients (median age 8.2 years), the majority of them without malocclusions, but third molars in development. Spontaneous alignment of second molars was more frequent in the maxilla (55%) than the mandible (47%). For incisors, desensitizing agents successfully managed hypersensitivity. Micro-abrasion and composite veneers improved aesthetics.

CONCLUSIONS

Few, mainly moderate to high-risk-studies investigated treatment of MIH. Remineralization or sealants seem suitable for MIH-molars with limited severity and/or hypersensitivity. For severe cases, restorations with composites or indirect restorations or preformed metal crowns seem suitable. Prior to tooth extraction as last resort factors like the presence of a general malocclusion, patients' age and the status of neighboring teeth should be considered. No recommendations can be given for MIH-incisors.

CLINICAL SIGNIFICANCE

Dentists need to consider the specific condition of each tooth and the needs and expectations of patients when deciding how to manage MIH. Strong recommendations are not possible based on the current evidence.