Detection and diagnosis of dental caries using a deep learning-based convolutional neural network algorithm.

Background
Proximal dental lesions, limited to dentine, are traditionally treated by invasive (drill and fill) means. Non-invasive alternatives (e.g. fluoride varnish, flossing) might avoid substance loss but their effectiveness depends on patients' adherence. Recently, micro-invasive approaches for treating proximal caries lesions have been tried. These interventions install a barrier either on top (sealing) or within (infiltrating) the lesion. Different methods and materials are currently available for micro-invasive treatments, such as sealing via resin sealants, (polyurethane) patches/tapes, glass ionomer cements (GIC) or resin infiltration.

Objectives
To evaluate the effects of micro-invasive treatments for managing proximal caries lesions in primary and permanent dentition in children and adults.

Search methods
We searched the following databases to 31 December 2014: the Cochrane Oral Health Group Trials Register, the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE via OVID, EMBASE via OVID, LILACs via BIREME Virtual Health Library, Web of Science Conference Proceedings, ZETOC Conference Proceedings, Proquest Dissertations and Theses, ClinicalTrials.gov, OpenGrey and the World Health Organization (WHO) International Clinical Trials Registry Platform. We searched the metaRegister of Controlled Trials to 1 October 2014. There were no language or date restrictions in the searches of the electronic databases.

Selection criteria
We included randomised controlled trials of at least six months' duration that compared micro-invasive treatments for managing non-cavitated proximal dental decay in primary teeth, permanent teeth or both, versus non-invasive measures, invasive means, no intervention or placebo. We also included studies that compared different types of micro-invasive treatments.

Data collection and analysis
Two review authors independently screened search results, extracted data and assessed the risk of bias. We used standard methodological procedures expected by Cochrane to evaluate risk of bias and synthesise data. We conducted meta-analyses with the random-effects model, using the Becker-Balagtas method to calculate the odds ratio (OR) for lesion progression. We assessed the quality of the evidence using GRADE methods.

Main results
We included eight trials, which randomised 365 participants. The trials all used a split-mouth design, some with more than one pair of lesions treated within the same participant. Studies took place in university or dental public health clinics in Brazil, Colombia, Denmark, Germany, Thailand, Greenland and Chile. Six studies evaluated the effects of micro-invasive treatments in the permanent dentition and two studies on the primary dentition, with caries risk ranging from low to high. Investigators measured caries risk in different studies either by caries experience alone or by using the Cariogram programme, which combines eight contributing factors, including caries experience, diet, saliva and other factors related to caries. The follow-up period in the trials ranged from one to three years. All studies used lesion progression as the primary outcome, evaluating it by different methods of reading radiographs. Four studies received industry support to carry out the research, with one of them being carried out by inventors of the intervention.

We judged seven studies to be at high overall risk of bias, primarily due to lack of blinding of participants and personnel. We evaluated intervention effects for all micro-invasive therapies and analysed subgroups according to the different treatment methods reported in the included studies.

Our meta-analysis, which pooled the most sensitive set of data (in terms of measurement method) from studies presenting data in a format suitable for meta-analysis, showed that micro-invasive treatment significantly reduced the odds of lesion progression compared with non-invasive treatment (e.g fluoride varnish) or oral hygiene advice (e.g to floss) (OR 0.24, 95% CI 0.14 to 0.41; 602 lesions; seven studies; I2 = 32%). There was no evidence of subgroup differences (P = 0.36).

The four studies that measured adverse events reported no adverse events after micro-invasive treatment. Most studies did not report on any further outcomes.

We assessed the quality of evidence for micro-invasive treatments as moderate. It remains unclear which micro-invasive treatment is more advantageous, or if certain clinical conditions or patient characteristics are better suited for micro-invasive treatments than others.

Authors' conclusions
The available evidence shows that micro-invasive treatment of proximal caries lesions arrests non-cavitated enamel and initial dentinal lesions (limited to outer third of dentine, based on radiograph) and is significantly more effective than non-invasive professional treatment (e.g. fluoride varnish) or advice (e.g. to floss). We can be moderately confident that further research is unlikely to substantially change the estimate of effect. Due to the small number of studies, it does remain unclear which micro-invasive technique offers the greatest benefit, or whether the effects of micro-invasive treatment confer greater or lesser benefit according to different clinical or patient considerations.

Micro-invasive interventions for managing proximal dental decay in primary and permanent teeth.

Background: Radiographs, adjunct to clinical examination are always valuable complementary methods for dental caries detection. Recently, progressing in digital imaging system provides possibility of software designing for automatically dental caries detection. Objectives: The aim of this study was to develop and assess the function of diagnostic computer software designed for evaluation of approximal caries in posterior teeth. This software should be able to indicate the depth and location of caries on digital radiographic images. Materials and Methods: Digital radiographs were obtained of 93 teeth including 183 proximal surfaces. These images were used as a database for designing the software and training the software designer. In the design phase, considering the summed density of pixels in rows and columns of the images, the teeth were separated from each other and the unnecessary regions; for example, the root area in the alveolar bone was eliminated. Therefore, based on summed intensities, each image was segmented such that each segment contained only one tooth. Subsequently, based on the fuzzy logic, a well-known data-clustering algorithm named fuzzy c-means (FCM) was applied to the images to cluster or segment each tooth. This algorithm is referred to as a soft clustering method, which assigns data elements to one or more clusters with a specific membership function. Using the extracted clusters, the tooth border was determined and assessed for cavity. The results of histological analysis were used as the gold standard for comparison with the results obtained from the software. Depth of caries was measured, and finally Intraclass Correlation Coefficient (ICC) and Bland-Altman plot were used to show the agreement between the methods. Results: The software diagnosed 60% of enamel caries. The ICC (for detection of enamel caries) between the computer software and histological analysis results was determined as 0.609 (95% confidence interval [CI] = 0.159-0.849) (P = 0.006). Also, the computer program diagnosed 97% of dentin caries and the ICC between the software and histological analysis results for dentin caries was determined as 0.937 (95% CI=0.906-0.958) (P < 0.001). Bland-Altman plot showed an acceptable agreement for measuring the depth of caries in enamel and dentin. Conclusions: The designed software was able to detect a significant number of dentin caries and acceptable measuring of the depth of carious lesions in enamel and dentin. However, the software had limited ability in detecting enamel lesions.

Designing of a Computer Software for Detection of Approximal Caries in Posterior Teeth

Secondary caries is the most common cause of dental restoration failures. This study aimed to compare the diagnostic accuracy of conventional and digital intraoral radiography and cone beam computed tomography (CBCT) for detection of recurrent caries around composite restorations. mesio-occluso-distal (MOD) cavities were prepared using bur on 45 extracted sound human molar teeth. The teeth were divided into 3 groups. In the control group, cavities were restored with composite resin after etching and bonding (n=15). In Group 2, 500-μm thick wax was placed over the buccal, lingual and gingival walls and the cavities were restored with composite resin. Group 3 specimens were subjected to pH cycling and artificial caries were created on the buccal, lingual and gingival walls. The cavities were restored with composite. Conventional and digital photo-stimulable phosphor (PSP; Optime) radiographs and two CBCTs images (NewTom 3G and Cranex 3D) were obtained from them. Presence or absence of caries in the cavity walls was assessed on these images. Data were analyzed using Kappa statistic. The diagnostic accuracy of CBCT was significantly higher than that of digital and conventional intraoral radiography (p<0.05). The accuracy was 0.83, 0.78, 0.55 and 0.49 for CBCT Cranex 3D, CBCT NewTom 3G, conventional and digital intraoral radiography, respectively. CBCT has a higher diagnostic accuracy than digital and conventional intraoral radiography for detection of secondary caries around composite restorations.

/pdf/comparison-of-cone-beam-computed-tomography-and-intraoral-10xb86glnp.pdf

Comparison of Cone-Beam Computed Tomography and Intraoral Radiography in Detection of Recurrent Caries under Composite Restorations.

Objective:To compare the diagnostic accuracy of digital intraoral and extraoral bitewing (EO BW) radiography in the detection of enamel proximal caries regardless of their ability to separate contacts.Methods:Artificial caries with different degrees of decalcification was induced in 80 human sound premolars and molars using formic acid. Intraoral radiographs were taken with photostimulable phosphor plate (PSP) and complementary metal oxide semiconductor (CMOS), using the paralleling bitewing technique. Extraoral bitewing radiographs were obtained using Sirona digital panoramic X-ray unit (Sirona Siemens, Bensheim, Germany). In total, 160 proximal surfaces were assessed by 2 observers twice. Area under the receiver operating characteristic curve (Az) values for each image type, observer and reading were compared using z-tests, with a significance level of p ≤ 0.05. Sensitivity, specificity, positive-predictive value and negative-predictive value for each observer and reading were calculated.Results:Spearma...

/pdf/intraoral-versus-extraoral-bitewing-radiography-in-detection-3uzh18cpoq.pdf

Intraoral versus extraoral bitewing radiography in detection of enamel proximal caries: an ex vivo study.

Background Radiography plays an important role in the detection of interproximal caries. Objectives The aim of the present study was to determine diagnostic accuracy of chargecoupled devices (CCD), Photo Stimulable Phosphor (PSP) and film radiography in detecting non-cavitated caries. Patients and methods Seventy-two non-cavitated approximal surfaces of extracted human posterior teeth were radiographed under standardized conditions using three intraoral modalities: CCD Dixi3 (Planmeca, Finland), PSP Digora PCT (Soredex, Finland),and E-speed film (Kodak, USA). Radiographs were interpreted by four observers and caries lesions were classified as sound (R0), restricted to enamel (R1), reaching the dentinoenamel junction (DEJ) and the outer half of the dentin (R2) and the inner half of the dentin (R3). The teeth were subsequently sectioned for histological analysis which served as the gold standard for radiographic examination. Results Microscopic examinations showed that the distribution of caries were 63.9% sound, 18.1% enamel, 9.7% DEJ and outer half of the dentin and 8.3% into the inner half of the dentin. The sensitivity and specificity of film, CCD and PSP for the detection of enamel caries were 38% and 98%; 15% and 96 %; and 23% and 98%, respectively. The sensitivity and specificity of film, CCD and PSP for the detection of both dentin and enamel caries were 55% and 100%; 45% and 100% ; and 55% and 100%, respectively. Conclusions The results demonstrated that the diagnostic accuracy of digital images is similar to that of conventional film radiography in the detection of non-cavitated approximal caries.

/pdf/diagnostic-accuracy-of-digital-and-conventional-radiography-38rqshcmxt.pdf

Nima Haghighat

Papers

Diagnostic accuracy of digital and conventional radiography in the detection of non-cavitated approximal dental caries.