Caries detection in patients can be a challenging process that, in rare cases, may result in misdiagnosis and mismanagement of patients.
The clinical detection of caries involves a visual, tactile, and radiographic examination of the tooth structure. Classically, the diagnosis for occlusal and smooth surface lesions involves visual examination of the well-dried tooth surface and tactile exploration for surface cavitation. Radiographic examination is another important clinical tool for diagnosing caries, especially on surfaces that are not easily visualized [1]. Dental radiographs utilize between 0.6 to 18.0 µSv of ionizing radiation to produce a two-dimensional image of the teeth [2]. Bitewing radiographs are considered the gold standard for clinically detecting interproximal lesions [3] and have a sensitivity of 0.38 and specificity of 0.98 [4]. The low sensitivity suggests that smaller cavities and white spot lesions (WSL) may not be detected by radiographic examination. There is insufficient radiolucency contrast between the healthy enamel and the WSL to be able to detect the lesions. Finally, WSLs are usually very small in volume, and very demarcated area of the enamel, making it even harder to image them using standard dental radiographic imaging.
Other methods for caries diagnosis include quantitative light-induced fluorescence and fiber-optic transillumination [5, 6]. These techniques are non-invasive and apply a beam of light to transilluminate the tooth surface. Areas of demineralization reflect the light differently and result in a shadowed appearance that clinicians can more easily detect than by visual exam alone [6]. The commercially available systems provide a numerical score to help quantify and monitor changes in lesions [7]. A meta-analysis examining fluorescence-based methods for caries detection of pre-cavitated lesions reported a sensitivity value of 0.77 and a specificity value of 0.75 [7]. However, the sensitivity of the transillumination systems is influenced by the presence of plaque, calculus, and/or staining on the tooth surface and the degree of dehydration of tooth tissue [8]. Another limitation is that such approaches do not provide information regarding lesion depth. Unless the enamel is significantly affected on its surface, it is almost impossible to quantify the nature of the lesion. Research on fluorescent-based technologies is still in its early stages, but these methods appear to be a promising adjunct to visual examination for caries detection.
In vitro methods for caries detection
Many caries detection methods have been and are being developed in laboratory environments. They comprise micro- and nano hardness (e.g., Vickers hardness test and nanoindentation), scanning electron microscopy (SEM), transverse microradiography (TMR), optical coherence tomography (OCT), and µ-computed-tomography (microCT) for example. However, while highly sensitive, these methods are usually not practical in a clinical setting due to their invasiveness, sample preparation requirements, or patient safety concerns. Previously, hardness (i.e., micro- and/or nano-hardness) was used to quantify both enamel demineralization/remineralization indirectly, and this approach has been used as a proxy for caries detection [9]. The technique relies on higher hardness of the sound enamel when compared to "weakened" carious or demineralized enamel [10].
A second in vitro method that has been valuable in studying enamel ultrastructure has been scanning electron microscopy (SEM) [11–13]. The advantages of SEM include high resolution and magnification; however, SEM requires samples to be processed (sectioning, dehydrating, metal-coating) and analyzed in a vacuum environment. While element mapping can be performed in an SEM, little to no quantitative information regarding mineral density can be obtained. Unfortunately, SEM is not a technique that can be translated chair-side, and studies can only involve either extracted or exfoliated teeth.
More recent studies have begun to employ microCT and TMR for caries detection and analysis. Both TMR and microCT have the advantages of being able to quantitatively measure demineralization and lesion depth at a high resolution and involve non-destructive sample processing [14–18]. Although MicroCT still requires the sample to be placed in a vacuum for analysis, this technique has long been regarded as the gold standard for detecting the demineralization of teeth in vitro. MicroCT is a modern imaging modality that uses X-ray beams to create cross-sections of an object to produce three-dimensional rendering, whereas TMR can only produce a two-dimensional image. MicroCT images can have a resolution with a voxel size smaller than 10 µm. Studies comparing caries detected by microCT to histological sections of decayed teeth found no significant differences between the two imaging modalities. A meta-analysis reported that for enamel caries detection, the sensitivity value of microCT ranged from 29–84% and the specificity from 88–95% [19].
OCT is a diagnostic imaging technique that has recently begun to gain popularity in dentistry due to its non-destructive and non-ionizing properties [20, 21]. Using a near-infrared laser, OCT can capture live 3D scans (5 to 10µm resolution) of the enamel based on the light scattering properties of the enamel. OCT suffers from the limited depth of penetration of its laser up to 2mm. This is, however, very suitable for any analysis on enamel and WSL. Specifically, enamel surface demineralization causes an increase in the scattering of the incident laser, which in turn produces an increase in reflectivity and depolarization. Resolution of newer OCT instruments can be in the order of 5 µm, and caries analysis findings appear consistent with others [22].
In addition to imaging, protocols for studying the caries process have also evolved over the years. Several demineralization protocols to produce artificial carious lesions have been developed in cariology, as presented elsewhere [23–25]. However, protocols have varied regarding sample preparation, demineralization solutions, and demineralization times. Most published literature uses an acidic buffer typically made from lactic acid to mimic the lactic acid produced by cariogenic bacteria found in dental plaque [26]. While these demineralization protocols have been employed to create artificial carious lesions, these studies are often studying the effect of a remineralization or anti-caries agent. Very few have studied how these artificial lesions compared to naturally occurring demineralized teeth [19] and have not quantified the mineral density. Herein, we aimed to compare artificially created with naturally occurring human enamel lesions using OCT and microCT.