Dental caries ranks among the most prevalent concerns in dentistry, posing challenges for early detection in clinical practice. Given the variability in caries depth, various dental restoration techniques are employed, underscoring the importance of precise measurement for determining optimal treatment plans and restoration modalities.^[1] Intraoral radiographs stand out as a popular choice among the array of caries detection methods and are extensively utilized.^[2]

Initially employed for caries detection, bitewing radiographs have evolved to become the prevailing radiographic method.^[2,3] Nonetheless, employing intraoral radiographs can present challenges in specific scenarios, such as when patient cooperation or tolerance is limited. In such instances, extraoral radiographs are gaining prominence, particularly among children, disabled patients, or individuals with severe gag reflexes.^[4] Furthermore, the quality of intraoral films hinges on the expertise of the radiologist. Consequently, the need for low-quality radiographic retakes not only compromises diagnostic accuracy but also escalates the patient’s radiation exposure.^[5]

Cone-beam computed tomography (CBCT) stands as a widely adopted digital radiography technique, celebrated for its multifaceted advantages. Notably, it offers a comprehensive three-dimensional (3D) depiction of craniofacial structures across axial, coronal, and sagittal planes, thereby enhancing treatment planning.^[6,7] Moreover, in various dental specialties such as restorative dentistry and endodontics, CBCT radiography contributes to minimizing treatment complexities alongside other advancements in the field.^[8-12]

Two primary types of artifacts are encountered in CBCT radiography, namely scatter and beam hardening, particularly in the presence of metal such as dental amalgam – a common metallic restorative material – which can disrupt the caries detection process.^[13,14] Scatter arises when X-ray photons deviate from their intended path following interaction with matter. Conversely, beam hardening occurs when lower-energy photons are preferentially absorbed compared to higher-energy photons as the X-ray beam traverses through an object, resulting in two distinct artifact manifestations: streaking (dark bands) and cupping artifacts (distortion of the metal structure).^[15]

Şeker et al.’s study^[16] and Kayipmaz et al.’s findings highlighted the superiority of CBCT imaging over phosphor storage plate (PSP)-based digital and conventional radiographs in detecting occlusal dental caries in the absence of amalgam restoration around the carious lesion.^[17] Furthermore, Zhang et al.’s investigation revealed no significant disparities in diagnostic accuracy between 3D CBCT systems (e.g., Kodak 9000 and Promax) and traditional intraoral PSPs when detecting noncavitated dental carious lesions.^[18] Both Esmaeili et al.^[19] and Isman et al.^[20] demonstrated that the beam-hardening effect induced by metal objects, such as orthodontic brackets, can produce artifacts that impede the diagnostic efficacy of CBCT imaging systems. These artifacts, similarly induced by amalgam restorations within the oral cavity, pose challenges for diagnosing dental caries.^[13]

This study investigated occlusal surface caries among individuals who had undergone CBCT imaging for various therapeutic objectives. The investigation stemmed from the expanding utilization of CBCT in dentistry, the conflicting outcomes of previous research regarding its efficacy in caries diagnosis, and the impact of metallic artifacts on the diagnostic process.

Table 1 presents the histopathological outcomes of the occlusal surfaces of 96 teeth. Intraobserver kappa coefficients ranged from 0.759 to 0.844 for the first observer and from 0.716 to 0.867 for the second observer. In addition, the interobserver kappa coefficient, calculated based on the mean values for each parameter from each observer, fell within the range of 0.631–0.769. A high intraobserver kappa coefficient signifies robust intraobserver agreement; hence, calculations were conducted using each observer’s initial readings.

The Cochran test findings revealed significant disparities between the histopathological assessments of the teeth and the CBCT images obtained before and after amalgam restorations. Subsequently, McNemar’s test was conducted, with the results outlined in Table 2. The McNemar test indicated no statistically significant difference between CBCT radiography before amalgam restoration placement and the pathological view of the teeth (Pv = 0.296, κ =0.004, and Pv = 0.294) [Figure 1].

However, a statistically significant difference was observed between CBCT radiography and the pathological view following amalgam restoration placement (Pv = 0.001, κ = 0.207, and Pv = 0.024) [Figure 2].

Table 3 presents the specificity, sensitivity, NPV, PPV, FNR, and FPR values for CBCT readings obtained before and after amalgam restoration. CBCT images exhibited higher sensitivity and lower specificity both before and after amalgam restoration.

Based on the receiver operating characteristic curve (ROC curve) analysis, the sensitivity of CBCT radiography in detecting caries is higher before amalgam restoration, but the specificity is higher after restoration [Table 3 and Figure 3].

Despite extensive research on CBCT’s efficacy in dentistry, several challenges remain to be addressed.^[21] Consequently, the current study aimed to assess the influence of amalgam fillings on CBCT-based occlusal caries detection accuracy, particularly considering deviations from strict adherence to the as low as reasonably achievable principle.^[22]

This research revealed that severe carious lesions exhibited low specificity values, indicating CBCT’s limitations in detecting noncarious occlusal lesions around amalgam restorations. In addition, reduced sensitivity values suggested an inability to detect demineralized enamel areas on the approximal surface in the presence of amalgam restorations. Furthermore, the ROC curve underscored the CBCT imaging’s inaccuracy in caries detection when amalgam restorations were present. Similarly, in Kulczyk et al.’s study, CBCT sensitivity for caries detection on surfaces adjacent to amalgam fillings ranged from 0.27 to 0.30 for enamel and from 0.47 to 0.56 for dentin. Specificity values for enamel proximal and distal lesions were 0.48 and 0.53, respectively, and for proximal and distal dentin lesions, they ranged from 0.33 to 0.38. Intraobserver reliability was 0.84, and interobserver reliability was 0.49. These findings corroborate the current study’s conclusions, suggesting that CBCT radiography may not be highly accurate in diagnosing dental caries.^[23] Indeed, various studies have produced conflicting findings regarding the accuracy of CBCT imaging in caries detection. In this study, CBCT images demonstrated low sensitivity and specificity values for detecting occlusal caries. Zhang et al. analyzed the detection accuracy of proximal caries using film, PSP, and CBCT. They compared the results with histological examination, considered the gold standard, and utilized film, PSP, ProMax 3D, and Kodak 9000 3D imaging systems. The findings indicated similar detection accuracy of proximal noncavitated carious lesions between CBCT imaging and film- and PSP-based intraoral imaging.^[18]

Haiter-Neto et al. reported sensitivity values of 13%–18% for NewTom and 21% for Accuitomo CBCT imaging systems across various FOV sizes (6, 9, and 12 inches). This diminished sensitivity could be attributed to subtle dental caries and a high ratio of sound-to-carious surfaces, mirroring findings in the present study. Their research also indicated lower caries detection accuracy with the NewTom third-generation CBCT compared to the 3DX Accuitomo CBCT and intraoral modalities.^[6] However, in contrast to our findings, Tarim Ertas et al. demonstrated that CBCT imaging surpassed other modalities in the detection of deep occlusal caries.^[24]

In their study, Nabha et al. concluded that metal objects such as amalgam could generate artifacts in 3D imaging, thereby impeding the diagnosis and prognosis of dental diseases.^[25] They also highlighted that MOD amalgam restorations could result in notable artifacts in CBCT images.

One limitation of this study is that it was conducted in in vitro conditions and the imaging was done without motion and in the absence of soft tissues around the mouth and teeth. Future research endeavors are advised to explore additional investigations with expanded sample sizes within a clinical environment. While the present study focused on occlusal caries, it is suggested that forthcoming studies explore the evaluation of carious lesions on proximal surfaces. In vivo studies with larger sample sizes hold the potential to yield more precise and comprehensive results.

This research uncovered the inadequate efficacy of CBCT imaging in detecting occlusal caries surrounding amalgam restorations.