Caries Detection around Restorations Using ICDAS and Optical Devices

Accuracy of detection methods for secondary caries around direct restorations: A systematic review and meta-analysis

OBJECTIVE

To evaluate and compare the accuracy of detection methods for the diagnosis of secondary caries around direct restorations in posterior teeth.

DATA

Accuracy parameters including sensitivity, specificity, diagnostic odds ratio (DOR), area under curve (AUC), and partial AUC (pAUC) are generated from studies assessing the accuracy of detection methods for secondary caries.

SOURCES

Publications from PubMed, Web of Science, Scopus, Medline, EMBASE and Cochrane Library databases.

STUDY SELECTION/RESULTS

This review included 25 studies evaluating visual examination (V(laboratory); n = 9 & V(clinical); n = 2), tactile examination (T; n = 3), intra-oral radiography (IR; n = 14), cone-beam computed tomography (CBCT; n = 4), quantitative light-induced fluorescence (QLF; n = 4), laser fluorescence (LF; n = 8) and digital imaging fiber-optic transillumination (DIFOTI; n = 1). The pooled sensitivity [95 % Confidence Interval, CI] and specificity [95 % CI] of detection methods for secondary caries were 0.60[0.45-0.73] and 0.67[0.53-0.78] for V(laboratory); 0.82[0.23-0.99] and 0.77[0.15-0.98] for V(clinical); 0.31[0.25-0.39] and 0.95[0.78-0.99] for T; 0.59[0.52-0.66] and 0.82[0.75-0.88] for IR; 0.61[0.48-0.73] and 0.82[0.64-0.92] for CBCT; 0.71[0.64-0.78] and 0.51[0.40-0.62] for QLF; 0.57[0.43-0.71] and 0.81[0.76-0.85] for LF; and 0.63[0.47-0.76] and 0.95[0.90-0.98] for DIFOTI. DOR values [95 % CI] of the secondary caries detection methods were V(laboratory)-2.88[2.18-3.80]; V(clinical)-16.66[3.84-72.28]; T-6.36[1.12-36.28]; IR-6.55[3.44-12.46]; CBCT-6.18[1.42-26.91]; QLF-2.25[1.39-3.63]; LF-4.86[2.40-9.82]; and DIFOTI-30.00[11.94-75.36], respectively. Respective AUC (pAUC) were V-0.645(0.535); T-0.379(0.315); IR-0.767(0.693); CBCT-0.887(0.820); QLF-0.581(0.633) and LF-0.828(0.590). AUC values were not available for DIFOTI and V(clinical).

CONCLUSIONS

Among the seven types of detection method for secondary caries diagnosis, none of the detection methods demonstrate satisfactory accuracy in detecting secondary caries around direct restorations in posterior teeth.

CLINICAL SIGNIFICANCE

This systematic review provides insights for the clinician and researcher in selecting the clinical detection method for secondary caries diagnosis and facilitates clinical decision making.

Clinical validity of fluorescence-based devices versus visual-tactile method in detection of secondary caries around resin composite restorations: diagnostic accuracy study

OBJECTIVES

To assess the validity of light-induced and laser-induced fluorescence devices compared to the visual-tactile method for detecting secondary caries around resin composite restorations.

MATERIALS AND METHODS

The study included 20 participants with 30 resin-composite restored teeth. Restorations' margins were examined using three diagnostic methods: the visual-tactile method (FDI criteria), the light-induced fluorescence camera (VistaCam iX), and the laser-induced fluorescence device (DIAGNOdent pen), and the reference was visual inspection after removal of defective restorations. The validity of each method was evaluated. Inter-examiner reliability was calculated using Cohen's kappa statistics. The level of significance was set at P = 0.05.

RESULTS

DIAGNOdent pen showed the highest sensitivity (100%) followed by VistaCam (98.82%) and the visual-tactile method (98.82%) at the enamel threshold. DIAGNOdent pen and VistaCam had lower specificity values than the visual-tactile method (81.69%, 76.06%, and 88.73% respectively). At the dentin threshold, DIAGNOdent pen yielded the highest sensitivity (89.36%), whereas VistaCam had the lowest (8.51%). The sensitivity of the visual-tactile method was low (57.45%) whereas all diagnostic methods had high specificity. There was perfect agreement in inter-examiner reliability for all assessment methods (Kappa 0.858-0.992).

CONCLUSIONS

Both fluorescence-based devices and the visual-tactile method are reliable for detecting secondary caries around resin composite restorations. DIAGNOdent pen is accurate in enamel and dentin, while VistaCam and the visual-tactile method can detect secondary caries in enamel only.

CLINICAL RELEVANCE

Fluorescence-based devices could be used as a valuable aid to supplement or as a second opinion after the visual-tactile method.

TRIAL REGISTRATION

The study was listed on www.

CLINICALTRIALS

gov with registration number (NCT04426604) on 11/06/2020.

Fluorescence and Near-Infrared Light for Detection of Secondary Caries: A Systematic Review

BACKGROUND

Early detection of secondary caries near dental restorations is essential to prevent further complications. This systematic review seeks to evaluate the sensitivity of fluorescence and near-infrared (NIR) imaging techniques for detecting secondary caries and to provide insight into their clinical utility.

METHODS

A comprehensive search strategy was used to select studies from seven databases, emphasizing diagnostic accuracy studies of secondary caries detection using fluorescence and NIR imaging techniques. The Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) instrument assessed bias risk and practicality. Two evaluators performed data extraction, screening, and quality assessment independently.

RESULTS

From 3110 initial recordings, nine studies were selected for full-text analysis. Wide variations in sensitivity (SE) and specificity (SP) values were reported across the studies. These studies exhibited variable SE and SP values, and the findings highlighted the importance of method selection based on clinical context. This systematic review underlines the potential for fluorescence and NIR imaging to detect secondary caries. However, results from different studies vary, indicating the need to consider additional variables such as restoration materials.

CONCLUSIONS

Although these technologies exhibit potential for detecting caries, our research underscores the complex procedure of identifying secondary caries lesions. It is a continuous necessity for progress in dental diagnostics to promptly identify secondary caries lesions, particularly those in proximity to tooth-colored ones.

Detection of Caries Around Resin-Modified Glass Ionomer and Compomer Restorations Using Four Different Modalities In Vitro

The aim of this study was to evaluate the ability of visual examination (International Caries Detection and Assessment System—ICDAS II), light-emitting diodes (LED) fluorescence (SPECTRA), laser fluorescence (DIAGNODent, DD), photothermal radiometry and modulated luminescence (PTR-LUM, The Canary System, CS) to detect natural decay beneath resin-modified glass ionomer (RMGIC) and compomer restorations in vitro. Twenty-seven extracted human molars and premolars, consisting of 2 control teeth, 10 visually healthy/sound and 15 teeth with natural cavitated lesions, were selected. For the carious teeth, caries was removed leaving some carious tissue on one wall of the preparation. For the sound teeth, 3 mm deep cavity preparations were made. All cavities were restored with RMGIC or compomer restorative materials. Sixty-eight sites (4 sites on sound unrestored teeth, 21 sound sites and 43 carious sites with restorations) were selected. CS and DD triplicate measurements were done at 2, 1.5, 0.5, and 0 mm away from the margin of the restoration (MOR). SPECTRA images were taken, and two dentists provided ICDAS II scoring for the restored surfaces. The SPECTRA data and images were inconclusive due to signal interference from the restorations. Visual examinations of the restored tooth surfaces were able to identify 5 of the 15 teeth with caries. In these situations, the teeth were ranked as having ICDAS II 1 or 2 rankings, but they could not identify the location of the caries or depth of the lesion. CS and DD were able to differentiate between sound and carious tissue at the MOR, but larger variation in measurement, and poorer accuracy, was observed for DD. It was concluded that the CS has the potential to detect secondary caries around RMGIC and compomer restorations more accurately than the other modalities used in this study.

In Vitro Detection of Caries Around Amalgam Restorations Using Four Different Modalities

Objective:

The aim of this study was to evaluate the ability of PTR-LUM (The Canary System, CS), laser fluorescence (DIAGNOdent, DD), LED fluorescence (Spectra), and visual inspection (ICDAS II) to detect natural decay around bonded amalgam restorations in vitro.

Methods:

Seventeen extracted human molars and premolars, consisting of visually healthy (n=5) and natural cavitated (n=12) teeth were selected. For the carious teeth, caries was removed leaving some decayed tissue on the floor and or wall of the preparation. For sound teeth, 3 mm. deep cavity preparations were made and teeth were restored with bonded-amalgam restorations. Thirty-six sites (13 sound sites; 23 carious sites) were selected. CS and DD scans were performed in triplicate at 2, 1.5, 0.5, and 0 mm away from the margin of the restoration (MOR). Spectra images were captured for the entire surface, and dentists blinded to the samples provided ICDAS II scoring.

Results:

Canary Numbers (Mean±SE) for healthy and carious sites at 2, 1.5, 0.5, and 0 mm from the MOR ranged from 12.9±0.9 to 15.4±0.9 and 56.1±4.0 to 56.3±2.0, respectively. DD peak values for healthy and carious sites ranged from 4.7±0.5 to 13.5±2.99, and 16.7±3.7 to 24.5±4.4, respectively. For CS and DD, sensitivity/specificity for sites at 2.0, 1.5, 0.5, 0 mm ranged from 0.95-1.0/0.85-1.0, and 0.45-0.74/0.54-1.0, respectively. For ICDAS II, sensitivity and specificity were 1.0 and 0.17, respectively. For Spectra, data and images were inconclusive due to signal intereference from the amalgam restoration.

Conclusions:

Using this in-vitro model, CS and DD were able to differentiate between sound and carious tissue at the MOR, but larger variation, less reliability, and poorer accuracy was observed for DD. Therefore, CS has the potential to detect secondary caries around amalgam restorations more accurately than the other investigated modalities.

In vitro Validation of Quantitative Light-Induced Fluorescence for the Diagnosis of Enamel Fluorosis in Permanent Teeth

This study aimed to validate quantitative light-induced fluorescence (QLF) as a diagnostic tool for mild and moderate enamel fluorosis in permanent teeth, comparing it to visual diagnosis and histological assessment completed using polarized light microscopy (PLM). The buccal surfaces of 139 teeth were visually classified using the Thylstrup and Fejerskov Index (TFI) into sound (TFI 0; n = 17), mild (TFI 1-2; n = 69), and moderate (TFI 3-4; n = 43) fluorosis. Fluorosis was then assessed with QLF (variables ΔF, A, and ΔQ at 5-, 15-, and 30-radiance thresholds) using as reference areas the entire surface and a region of interest (ROI), identified as the most representative region of a fluorosis lesion. PLM images of longitudinal thin sections including the ROI were assessed for histological changes. Correlations among TFI, PLM, and QLF were determined. A receiver-operating characteristic curve was conducted to determine QLF's diagnostic accuracy when compared to the TFI and PLM assessments. This was used to assess the probability that the images were correctly ranked according to severity as determined by PLM and TFI. A positive correlation was found between QLF and PLM, and between QLF and TFI. QLF showed the highest sensitivity and specificity for the diagnosis of mild fluorosis. There was also a strong agreement between TFI and PLM. The selection of a ROI resulted in a stronger correlation with TFI and PLM than when the entire surface was used. The study results indicate that defining an ROI for QLF assessments is a valid method for the diagnosis of mild and moderate enamel fluorosis.

Clinical evaluation of near-infrared light transillumination in approximal dentin caries detection

The objective of this clinical study was to compare conventional caries detection techniques, pen-type laser fluorescence device, and near-infrared light transillumination method in approximal dentin caries lesions. The study included 157 patients, aged 12-18, without any cavity in the posterior teeth. Two calibrated examiners carried out the assessments of selected approximal caries sites independently. After the assessments, the unopened sites were excluded and a total of 161 approximal sites were included in the study. When both the examiners arrived at a consensus regarding the presence of dentin caries, the detected lesions were opened with a conical diamond burr, the cavity extent was examined and validated (gold standard). Sensitivity, specificity, negative predictive value, positive predictive value, accuracy, and area under the ROC curve (Az) values among the caries detection methods were calculated. Bitewing radiography and near-infrared (NIR) light transillumination methods showed the highest sensitivity (0.83-0.82) and accuracy (0.82-0.80) among the methods. Visual inspection showed the lowest sensitivity (0.54). Laser fluorescence device and visual inspection showed nearly equal performance. Near-infrared light transillumination can be used as an alternative method to approximal dentin caries detection. Visual inspection and laser fluorescence device alone should not be used for approximal dentin caries.

Reliability and validity of FluoreCam for white-spot lesion detection: An in vitro study

AIM

In the present study, we tested the reliability and validity of a new light fluorescence device, the FluoreCam.

METHODS

Twenty-five human teeth were sectioned mesiodistally into halves. Group 1 (n=30) included specimens with either sound enamel or natural white-spot lesions (WSL). Group 2 (n=20) included specimens with sound enamel used to create artificial WSL. Using the FluoreCam, baseline scans of enamel surfaces were obtained under standardized conditions. Group1 was scanned again the next day. Group 2 specimens were covered with an acid-resistant nail polish, leaving a 2 × 6-mm enamel window exposed, immersed in Queiroz-solution (64 hours, 37°C/pH5 with agitation), and then scanned again. Random error of the repeated measurements (reliability) was determined using method errors and intraclass correlations (ICC). Systematic error and the ability to detect demineralization (validity) were tested using Wilcoxon signed-rank test.

RESULTS

Method errors for Group 1 replicates were .39 mm² (area), .72 pixels (intensity), and 5.69 pixels.mm² (impact). Group 2 method errors were consistently slightly higher than those of group 1. FluoreCam showed highly reliable measurements for group 1 (ICC: 0.93-0.98) and group 2 (ICC: 0.87-0.97). There were no statistically-significant systematic errors for either group (P>.05). In group 2, enamel demineralization was statistically significant in area (P<.001), intensity (P=.001), and impact (P<.001).

CONCLUSION

FluoreCam is highly reliable and valid for in vitro assessments of enamel demineralization.

Comparative study of two chemical protocols for creating white spot lesions: An in vitro FluoreCam evaluation

AIM

The aim of the present study was to compare the effectiveness and practicality of two commonly-used protocols for white spot lesion creation.

METHODS

Sound posterior human teeth were sectioned into halves and randomly allocated into two groups (n=20). Using the FluoreCam system, the enamel surfaces were imaged under standardized conditions, and baseline data (area, intensity, and impact) were recorded. Specimens were covered with an acid-resistant polish, leaving a 2×6-mm enamel window. Group 1 was immersed in Queiroz solution (64 hours/37°C/pH5 with agitation); Group 2 was immersed in Buskes solution (14 days/37°C/pH5 with agitation). The specimens then were imaged again. Within-group changes that occurred over time were evaluated using the Wilcoxon signed-rank test. Between-group differences were analyzed using the Mann-Whitney U-test. Polarized light microscopy (PLM) was used to calculate the mean lesion depth of representative specimens.

RESULTS

Each protocol showed significant changes in the surface area, intensity and impact of demineralization over time (P<.05). There were no statistically-significant differences (P=.53, P=.2, P=.74) between the two groups. PLM showed that the mean lesion depths of representative specimens were 51.4 and 73.0 μm for groups 1 and 2, respectively.

CONCLUSIONS

While both protocols produce similar amounts of demineralization, the Queiroz solution is more practical, as it requires less time.