Comparative assessment of CBCT-synthesized bitewing radiography for occlusal and interproximal caries depth detection: an in vitro study

Introduction

Dental caries is a health dilemma, and one of the most common chronic diseases worldwide, having a profound impact on quality of life and a huge indirect effect on eco­nomy [1]. Dental caries is defined as a loss of dental hard tissue due to bacterial invasion [2]. Early and accurate detection of occlusal and proximal caries enables prompt treatment, and prevents extensive loss of the tooth structure [3]. Diagnostic tools currently used for detection of carious lesions include fiber-optic transillumination, contrast dye, and a combination of clinical and conventional radiographic examinations [2]. Despite the confirmed variations in interpretation of lesions, radiography is still the most practical paraclinical method utilized in routine clinical practice [4].
Panoramic radiography is not sufficiently accurate for assessment of details, such as caries, due to its low spatial resolution of approximately 5 lp/mm. However, it is suitable for general assessment of both the maxilla and mandible on one cliché [5, 6]. Periapical radiography with modern intraoral sensors provides high spatial resolution, often exceeding 20 lp/mm, ensuring precise detection of changes in the bone and tooth structure.
Bitewing (BW) radiography is divided into intraoral (IBW) and extraoral (EBW). IBW is the most commonly requested radiographic examination for detection of proximal caries, and is more accurate than periapical radiography used for this purpose [1]. Detection of pro­ximal caries on IBW radiographs depends on several factors, including caries depth, tooth position in dental arch, X-ray angle, superimposition of adjacent structures, artifacts, and patient-related factors [3]. Moreover, IBW has the shortest exposure time and the lowest patient radiation dose compared with panoramic radiography and EBW; however, it is not well-tolerated by children, patients with severe gag reflex, and those susceptible to coughing. Also, it has the highest risk of aerosol generation compared with extraoral radiographic modalities [5]. The success of this method highly depends on the experience and expertise of the operator in adjusting correct angle of the tube for optimal visualization of proximal contacts without superimposition. Improper angulation requires retaking, which is not only time-consuming, but also exposes patient to additional radiation [7].
EBW can offer images similar to IBW, and provide data on the periapical area as well. It has been claimed that this approach is suitable for patients who have problems with taking IBW radiographs, and may be able to serve as an alternative to panoramic radiography for detection of proximal caries in challenging cases, such as those with acute gag reflex, patients who cannot sufficiently open their mouth due to trismus or infection, and disabled or mentally retarded patients [3]. This modality has minimal risk of aerosol generation, since the film is not placed in the oral environment. The patient radiation dose in EBW is higher than that in IBW, and lower than that in panoramic radiography. Compared with IBW, the need for radiograph retakes is diminished in EBW. In this modality, 2 EBW radiographs are obtained instead of 4 IBW radiographs [5].
EBW radiographs are acquired in panoramic imaging by adjusting the projection geometry, whereas cone-beam computed tomography (CBCT) enables the synthesis of bitewing radiographs from volumetric data. Although it is recognized that CBCT has a significantly higher radiation dose than intraoral bitewing radiography and its routine use for caries detection is not justifiable under the ALARA principle, previous studies have highlighted its potential utility in cases, where CBCT images were obtained for other clinical purposes, such as implant planning [8]. In our study, CBCT was used exclusively for extracted teeth without metal restorations in a controlled in vitro condition, minimizing radiation concerns and avoiding the interfe­rence of metal artifacts, which are known to affect diagnostic accuracy.
The current study aimed to assess and compare the dia­gnostic accuracy of CBCT-synthesized bitewing and intraoral bitewing radiographic modalities for detection of occlusal and interproximal caries depth in comparison with histopathological gold standard. Utiliz­ing the advanced capabilities of the Carestream CS 9600 CBCT unit, this study highlighted the importance of CBCT-synthesized bitewing radiography (CSBW) as an innovative diagnostic modality that combines high diagnostic accuracy with significant patient comfort. It is particularly useful for those who have difficulty with IBW imaging, providing a practical and efficient alternative. This research fills an important gap in the scientific lite­rature, demonstrating the ability of CBCT-based imaging to improve the accuracy of caries detection while reducing additional radiation exposure in cases, where CBCT scans obtained for other clinical applications are available.

Objectives

The objective of this study was to investigate the dia­gnostic accuracy of CSBW radiography compared with IBW radiography and the histopathological gold standard, for detecting occlusal and interproximal caries depth. Since histopathological sections are not applicable in clinical practice, IBW, as one of the most commonly used methods for caries detection, was included as a comparative reference to assess the relative performance of CSBW in clinical applications. Additio­nally, the study aimed to assess the clinical applicability of CSBW radiography in challenging patient scenarios, where intraoral imaging is not feasible.

Material and methods

This in vitro experimental study was approved by the Ethics Committee of Hamadan University of Medical Sciences (approval number: IR.UMSHA.REC.1402.124). Sample size was calculated to be 160 human extracted posterior teeth using G*Power software, assuming 0.05 level of significance, area under the receiver operating curve (ROC; AUC) of 0.7, and margin of error of 0.071 with an 80% power. Accordingly, 160 extracted poste­rior teeth (including first and second premolars, and first and second molars) were employed to assess occlusal and interproximal caries depth. Inclusion criteria were human extracted posterior teeth with no restoration or endodontic treatment, and dry human skulls with no fracture. Two sound dry human skulls were coated with three layers of red dental wax buccally and lingually (Cavex dental waxes, Holland) for soft tissue simulation [9]. At each imaging stage (five stages), 16 teeth (8 in the maxilla and 8 in the mandible) were placed in each skull, resulting in a total of 32 teeth per stage across both skulls. To enhance accuracy and prevent repeated evaluations of the same teeth, after each stage, these 32 teeth were removed from the skulls and replaced with a new set of 32 teeth. This process was repeated five times to ensure that all 160 teeth were evaluated under identical conditions. In each stage, two CSBW and four IBW radiographs were obtained from each skull, resulting in a total of 20 CSBW and 40 IBW radiographs for teeth assessment. In case of poor adaptation of teeth with extraction sockets in the skulls, the root form was changed by a cylindrical diamond bur (Mani, Inc., Tokyo, Japan). To prevent overlapping, the teeth were mounted with proximal surfaces being parallel to each other. The mandible was then fixed to the skull with wax and glue such that the teeth were in occlusion. Next, the teeth underwent digital IBW radio­graphy using a MINRAY® intraoral X-ray unit (MINRAY, Soredex, Tuusula, Finland), operated at 60 kVp, 7 mA, and an exposure time of 0.32-0.40 seconds. A size 2 photostimulable phosphor plate (PSP) was used with a DIGORA® Optime scanner (Soredex, Tuusula, Finland) for image processing. PSP plates used in this study were specifically designed for the DIGORA® Optime system. IBW was included as a reference method (not as the gold standard) to provide a comprehensive comparison of clinical applications of CSBW under real-world conditions. This was done, because histopathological sections, despite being the gold standard, are not applicable for caries diagnosis in clinical practice. By incorporating IBW, it was aimed to better evaluate the diagnostic performance of CSBW and its potential as an alternative imaging modality. Figure 1 illustrates IBW radiographs taken from the first skull with the first arrangement pattern of teeth and PSP sensor. A PSP plate was used with Rinn (Pekan, China) film holder. CSBW radiograph (Figure 2) was obtained from the first skull with the first arrangement pattern of teeth, using CS 9600 CBCT scanner (Care Stream, Kodak, France) with exposure settings of 73 kVp, 8 mA, and 8-second time. Extraoral bitewing images were synthesized from the CBCT volume using Carestream CS imaging software (Carestream Dental, Rochester, NY, USA). The process was fully automated, and did not require any manual selection of region of interest (ROI) or segmentation. The system autonomously handled all image reconstruction parameters, and final images were generated according to the default processing protocols of the software.
To obtain CSBW radiographs, colored markers were used around the skull to standardize the object position and eliminate the effect of variations in object position on the results. The device’s laser beam matched the color lines on the skull in different planes [10]. Radiographs with poor image quality (blurring, magnification, asymmetry, distortion, etc.) were excluded. All the radiographs were evaluated by a trained second-year postgraduate student of oral radiology and subsequently by an oral radiologist with 7 years of experience, with a one-week interval and blinded to results. The images were observed on a 20-inch monitor (LG, Seoul, Korea) under semi-dark conditions, and data were recorded in a checklist. After one week, 20% of the radiographs were evaluated again by the radiologist and postgraduate student to calculate intra- and interobserver agreements. The depth of caries was scored with a code merging system presented in Table 1 [11]. Each group included 40 teeth, ensuring an equal distribution among caries classifications: sound, incipient enamel caries, moderate caries, and severe caries extending to the inner dentin or pulp, making a total of 160 teeth (Table 1).
Next, the teeth were removed from the sockets to obtain histopathological sections to serve as the gold standard for caries depth. The teeth were individually sectioned in mesiodistal direction with a sectioning machine (Krupp Dental Dentarapid, Fride Krupp GMBH, Krupp WIDIA, N:759 DR2, Germany), using a 0.51 mm thick diamond blade. Subsequently, each section was inspected under a stereomicroscope (Carl Zeiss AG, Oberkochen, Germany) at 8X magnification, and the depth of caries on IBW and CSBW radiographs was compared with the gold standard (Figures 3 and 4). All data were analyzed using SPSS version 26.0 and STATA version 17.0 at 0.05 level of significance.

Results

As shown in Table 2, a significant difference existed in detection of different occlusal and proximal caries depths on IBW radiographs compared with the gold standard. The highest sensitivity (100%) for depth detection was found in a severe carious lesion (C) in the distal surface of the teeth, and the lowest (12%) sensitivity for depth detection was observed in an incipient enamel lesion (A) in the distal surface. The diagnostic accuracy for caries depth detection was the highest for severe caries (C) in the mesial surface (0.98), and the lowest (0.67) in a sound occlusal surface. Table 3 shows the comparative accuracy of CSBW and IBW for different caries depths. It is important to note that while histopatholo­gical sections served as the gold standard in this study, IBW was included for comparative analysis to evaluate the relative diagnostic performance of CSBW in clinical applications. As demonstrated in Table 3, there was a significant difference in detection of different occlusal and proximal caries depths on CSBW compared with IBW, and the accuracy was higher for proximal caries than occlusal caries for all depths. The sensitivity was also higher for proximal caries than occlusal caries for all depths, except for incipient caries (A).
As reported in Table 4, a significant difference was found in detection of different occlusal and proximal caries depths on CSBW compared with the gold standard. The sensitivity of caries depth detection was higher in sound teeth (0) and moderate caries (B) of the proximal surfaces as well as incipient caries (A) and severe caries (C) of the occlusal surfaces. Furthermore, the accuracy was higher in the proximal than occlusal surface at all depths, except for moderate caries (B). The ROC curves of CSBW and IBW compared with the gold standard for caries detection separately in the mesial, distal, and occlusal surfaces are presented in Figures 5, 6, and 7, respectively. The AUC values were 0.958 and 0.924 in the mesial, 0.884 and 0.926 in the distal, and 0.854 and 0.819 in the occlusal surface, respectively, indicating high sensitivity and spe­cificity for detection of mesial, distal, and occlusal caries.
Table 5 presents intraobserver k coefficients for the agreement between CSBW and IBW relative to each other and the gold standard for each observer. The results indicated very strong agreement (range, 0.81-100) for the mesial and distal surfaces, and strong (range, 0.61-0.80) to very strong agreement for the occlusal surfaces. Both IBW and CSBW demonstrated reliable performance in detecting caries depth compared with the gold standard, with slightly lower agreement in the occlusal surface. The agreement between IBW and CSBW was consistently very strong at all surfaces.
Tables 6 and 7 present the weighted k coefficients for interobserver agreement in extraoral CSBW and IBW radiography. The agreement was reported separately for the mesial, distal, and occlusal surfaces. Table 6 shows the results of the first reading (evaluation of all images), while Table 7 provides the results of the second reading (reevaluation of 20% of the images after one-week interval). Interobserver agreement was consistently very strong (> 0.8) in both the first and second readings.

Discussion

Dental caries is one of the most common chronic diseases affecting males and females of all ages [4, 12]. It is characterized by dissolution and demineralization of tooth structure subsequent to microbial infection, and is considered as an infectious microbial disease [8]. Early and precise caries detection in dental practice is highly important, enabling dental clinicians to prevent further loss of the tooth structure by instant operative procedures [8, 12]. Subjective methods [13], such as visual and tactile sense thorough clinical examination and radiographic examination as a supplemental modality, are commonly employed for caries detection [4]. Visual and tactile clinical examination is the standard method used during first patient viewing, while radiography is indicated as additional diagnostic evaluation, especially in high-risk patients and advanced cases with fissure caries and interproximal caries, and may be recommended [12].
The present results revealed a significant difference in detection of caries depth among occlusal and proximal caries with different depths on IBW radiographs compared with the gold standard, CSBW scans compared with IBW, and CSBW compared with the gold-standard. The diagnostic accuracy of IBW was the highest for severe caries (C) compared with other caries in the mesial surface, and the lowest for sound occlusal surface (0) compared with the gold standard. In CSBW, the accuracy for all caries depths (0, A, B, and C) was higher in the proximal than occlusal surface, compared with IBW. In CSBW compared with the gold standard, the accuracy for all caries depths, except for moderate caries (B), was higher in the proximal than occlusal surface. These results are in contrast with the following three studies. Osama et al. [14] in 2020 compared the diagnostic accuracy of CBCT, EBW, and IBW in vitro, and showed similar diagnostic accuracy of CBCT and EBW with IBW for detection of interproximal caries, indicating that with sufficient training and experience, CBCT and EBW can be comparable with IBW for detection of interproximal caries. Similarly, Abu El-Ela et al. [4] in their ex vivo study found no significant difference in the diagnostic accuracy of IBW and EBW for detection of enamel caries, irrespective of their capability in differentiation of interproximal contacts. EBW showed comparable results with IBW, with PSP and CMOS sensors for detection of proximal enamel caries. Only IBW with PSP sensor was evaluated in the present study. Kamburoglu et al. [3] in their ex vivo study in 2012 reported higher diagnostic accuracy of IBW than EBW and panoramic radiography for detection of proximal caries in premolar and molar teeth.
In the current study, the ROC curve and AUC values revealed high sensitivity and specificity of proximal and occlusal caries detection by both the observers for both imaging modalities compared with the gold standard. This finding was in contrast to the results of Chan [15] in 2015, who compared IBW and EBW for detection of proximal caries and alveolar bone loss, showing higher detection of caries and bone loss by using EBW compared with IBW. Although EBW radiography is efficient and comfortable for patients, it appears to have a higher rate of false positive results; therefore, there is need for clinical examination and reevaluation by scheduling follow-up sessions [15]. Oktavian et al. [16] in 2023 reported that the sensitivity, specificity, and AUC of IBW were higher than those of EBW, but EBW showed optimal accuracy. Therefore, it may serve as an alternative to IBW for patients, who have difficulty in taking an IBW scan, and also during the COVID-19 pandemic peaks [16].
The present results indicated higher sensitivity of CSBW compared with the gold standard for detection of sound teeth (0) and moderate caries (B) of the pro­ximal surfaces, and incipient (A) and severe caries (C) of the occlusal surfaces.
The assessment of intra- and interobserver agreements in the first and second readings by calculation of the weighted k coefficient revealed that in all scenarios, except for IBW compared with CSBW, the highest agreement was found in the proximal compared with occlusal surfaces in the second reading of the first observer.

Conclusions

CSBW radiography demonstrates high diagnostic accuracy in detecting the depth of occlusal and proximal caries, and is emerging as a promising alternative to IBW imaging. By integrating advanced CBCT techno­logy, CSBW not only increases diagnostic accuracy but also provides significant patient comfort, particularly for individuals with anatomical complexities or those unable to tolerate IBW procedures. Furthermore, in cases where CBCT imaging is already recommended for other clinical purposes, such as implant planning, CSBW ensures a comprehensive diagnostic assessment by reducing additional radiation exposure. This innovative approach overcomes the fundamental limitations of conventional methods, highlighting the potential of CBCT-synthesized imaging in modern dental diagnostics.

Disclosures

1. Institutional review board statement: This study was approved by the Ethics Committee of the Hamadan University of Medical Sciences (approval number IR.UMSHA.REC.1402.124).
2. Assistance with the article: None.
3. Financial support and sponsorship: None.
4. Conflicts of interest: The authors declare no potential conflicts of interest concerning the research, author­ship, and/or publication of this article.

References