Diagnosis of occlusal caries using optical coherence tomography in a real clinical situation

Abstract

Purpose: Optical coherence tomography (OCT) is an imaging device that can create cross-sectional images of tooth structure. This study evaluated the diagnostic performance of dental OCT system for occlusal caries in a real clinical situation.

Methods: Twenty-four occlusal locations on 20 teeth were selected from 12 patients (mean age: 39.6 years). Swept-source OCT (SS-OCT) images were taken from the occlusal surfaces to diagnose occlusal enamel caries using the international caries detection and assessment system (ICDAS) grading. Indices of sensitivity and specificity for the detection of superficial enamel demineralization (code 1), distinct enamel demineralization (code 2), and enamel caries (code 3) were calculated from the results of visual inspection and SS-OCT obtained by three dentists with less than 2 years clinical experience.

Results: The sensitivity of visual inspection and SS-OCT for enamel demineralization (code 1) were 0.22 and 0.55; for enamel demineralization (code 2) were 0.50 and 0.61; and for enamel caries (code 3) were 0.22 and 0.44, respectively. The specificity of visual inspection and SS-OCT were 0.52 and 0.78, respectively.

Conclusion: SS-OCT can provide excellent diagnostic performance for occlusal enamel caries in a real clinical situation.

Introduction

Dental caries is a dynamic disease process that involves demineralization and remineralization, for which the imbalance of equilibrium decline during demineralization can cause advanced lesions [1,2]. The application of bioavailable calcium, phosphate and fluoride ions in an oral environment is recognized as an efficient therapy for the remineralization of demineralized enamel [2-4]. As for the recent development of a minimally invasive option for the treatment of caries, an accurate diagnosis is essential to select the optimal therapy for the caries.

For the diagnosis of caries, intraoral radiography is widely used especially for occlusal caries and interproximal caries, where visual inspection is difficult to access. However, dental radiographs exhibit a low sensitivity in early caries because of the superimposition of anatomical structures and artifacts [5,6].

The International Caries Detection and Assessment System (ICDAS) was recently developed to detect the severity and activity of the caries as early as possible by visual inspection. The ICDAS determines optical changes in enamel demineralization after air drying. Thus far, the ICDAS has demonstrated a high accuracy and reproducibility for detecting early enamel caries in permanent teeth [7], and for diagnosis of caries in various stages [8].

Optical coherence tomography (OCT) is a non-invasive imaging method using light to visualize the internal biological structure. Swept-source OCT (SS-OCT) is one of the latest versions of OCT, which demonstrates excellent scanning speed and image resolution for clinical applications. Since OCT does not require any radiation dose and can safely be used on pregnant woman, it is recognized as a sustainable imaging technology for the diagnosis of caries in dentistry. Previous studies showed that a dental OCT system based on SS-OCT has a high diagnostic accuracy for occlusal and proximal surface caries from in vitro experiments using extracted teeth [9-11]. Nevertheless, only a small number of studies have been conducted on OCT used to diagnose caries in real clinical situations [12]. Therefore, this study was designed to investigate the intraoral diagnostic accuracy of dental OCT for early caries in molar occlusal surfaces.

Materials and Methods

Subjects

This study included 12 adults without any systemic diseases aged between 23 and 77 years (mean age: 39.6 years). Twenty-four occlusal surfaces of 20 molars were selected from patients for this study. After tooth surface cleaning using prophylaxis paste (Pressage, Shofu Inc., Kyoto, Japan) and brush cone (Merssage Brush No.1, Shofu Inc.) attached to a low-speed handpiece at a rotating speed below 2,000 rpm, digital camera photographs of the occlusal surfaces were taken both under moist and dry conditions. SS-OCT scanning was then performed to obtain a three-dimensional (3D) dataset of the occlusal surface. Using digital camera photographs and SS-OCT images, ICDAS grading of enamel caries at occlusal surfaces was carried out by the examiners.

Dental OCT system and imaging

Figure 1 shows the Dental OCT System (Octina, the Yoshida Dental Mfg. Co., Ltd., Tokyo, Japan), which was used in the present study. This system is a SS-OCT, which employs a high-speed frequency-swept laser light with a center wavelength of 1,310 nm and a scan range of 140 nm at 50 KHz. The z-axis resolution of the system is 11 µm in air. A handheld probe for intraoral imaging attached to the system was used to obtain 3D dataset for the occlusal surfaces (Fig. 1).

Scoring of occlusal surfaces using ICDAS and dental OCT

Before scoring by the examiners, the occlusal surfaces were carefully observed by four calibrated dentists with over 8 years of experience in cariology at Tokyo Medical and Dental University (NH, TT, TS, and YS) using both digital camera photographs and SS-OCT images. Using digital camera photographs, taken under both moist and dry conditions, together with SS-OCT images, the calibrated dentists performed ICDAS grading for the 24 occlusal surfaces as follows: Code 0, Sound surface; Code 1, First visual change in enamel seen only after prolonged air drying or restricted to the confines of a pit or fissure; Code 2, Distinct visual change in enamel that is within or wider than the confines of a pit or fissure and seen on a wet or dry surface; Code 3, Localized enamel breakdown without visual signs on dentin exposure; and Code 4, Underlying dark shadow from dentin.

The findings obtained from the investigated sites were then compared by the calibrated dentists to achieve a consensus score for each site. The scoring results by the calibrated dentists were employed as the definitive diagnosis and used for the following experiment.

Three dentists who had less than two years of clinical experience participated in this study, carrying out the ICDAS grading of occlusal surface blind to each other. The same criteria as for the definitive diagnosis was employed for the diagnosis by the three dentists, but using either digital camera photographs for visual inspection or SS-OCT images. For visual inspection, digital camera photographs both under the moist and dry conditions were used. For SS-OCT imaging, 2D image of occlusal surface was selected from the 3D dataset and displayed in the monitor, where the three dentists evaluate the presence of enamel caries.

Fig. 1 Clinical image using a dental OCT system

OCT imaging of molar occlusal surface was performed using an intraoral probe attached to the OCT system (red arrow). Obtained 3D preview image was displayed on the monitor.

Table 1 Sensitivity, specificity and diagnostic accuracy for visual inspection and SS-OCT using ICDAS code

		Visual inspection	SS-OCT
SP		0.52	0.78
SE	code 1	0.22	0.55
	code 2	0.50	0.61
	code 3	0.22	0.44
κ		0.33	0.61

SE, sensitivity; SP, specificity; κ, diagnostic accuracy

Statistical analysis

The results obtained by the three dentists were statistically analyzed using a statistical software package (Dr SPSS-2 for Windows, SPSS Inc., Chicago, IL, USA). The results of ICDAS grading for visual inspection and SS-OCT image by three dentists were compared with the results of definitive diagnosis performed by the calibrated dentist, and indices of sensitivity and specificity for visual inspection and SS-OCT were calculated. The diagnostic accuracy as the agreement of visual inspection or SS-OCT imaging with the definitive diagnosis was calculated using weighted kappa.

Results

The results of sensitivity, specificity and diagnostic agreement for visual inspection and SS-OCT are shown in Table 1. The sensitivity values of visual inspection and SS-OCT for code 1 enamel demineralization were 0.22 and 0.55, whereas code 2 enamel demineralization values were 0.50 and 0.61, respectively. The sensitivity values of visual inspection and SS-OCT for code 3 enamel caries were 0.22 and 0.44, respectively. The specificity values of visual inspection and SS-OCT were 0.52 and 0.78, respectively. The diagnostic agreement for visual inspection and SS-OCT were 0.33 and 0.61, respectively.

Representative images of visual inspection and SS-OCT obtained from in vivo occlusal surfaces are shown in Figs. 2-5.

Fig. 2 Representative images of intact enamel (red arrows, ICDAS code 0)

a: occlusal view under dry condition. b: occlusal view under moist condition. No visible changes of occlusal fissure were observed before and after the air drying. c: en face intensity image generated by 3D OCT. d: extracted 2D image of the area along the red line in c. Whole thickness of enamel was identified with the location of DEJ (white open arrow).

Fig. 3 Representative images of first enamel change by demineralization (red arrows, ICDAS code 1)

a: occlusal view under dry condition. Central pit of occlusal surface was stained dark (red arrow). b: occlusal view under moist condition. Staining of central pit was not obvious under the moist condition (red arrow). c: en face intensity image generated by 3D OCT. d: extracted 2D image of the area along the red line in c. The brightness of occlusal fissure was increased and imaged white (red arrow). The depth of bright zone was confined within the half thickness of enamel. Location of DEJ was identified (white open arrow).

Fig. 4 Representative images of distinct enamel change by demineralization (red arrows, ICDAS code 2)

a: occlusal view under dry condition. Distal pit of occlusal surface was stained dark (red arrow). b: occlusal view under moist condition. Distal pit was staining dark even under the moist condition (red arrow). c: en face intensity image generated by 3D OCT. d: extracted 2D image of the area along the red line in c. The brightness of occlusal fissure was increased and imaged white (red arrow). The depth of bright zone was confined within the half thickness of enamel. However, the area was increased more than Fig. 3d. Location of DEJ was identified (white open arrow).

Fig. 5 Representative images of localized enamel breakdown without visual signs on dentin exposure (red arrows, ICDAS code 3)

a: occlusal view under dry condition. Almost the whole occlusal fissure was stained dark suspicious of enamel caries. Distal pit was visibly concave and became chalky (red arrow). b: occlusal view under moist condition. Almost whole occlusal fissure was stained dark. The chalky appearance observed under the dry condition was not obvious under the moist (red arrow). c: en face intensity image generated by 3D OCT. d: Extracted 2D image of the area along the red line in c. The contour of occlusal fissure was concave (red arrow). The brightness of occlusal fissure was increased and imaged white. The depth of bright zone was confined within the enamel. Location of DEJ was identified (white open arrow).

Discussion

The SS-OCT system used in this study involves intraoral scanning using an oral probe, for which the imaging of molar occlusal surfaces showed no adverse events in this study. This result was in accordance with previous studies performed on populations with special health care needs using a prototype OCT system [13]. In this study, 1 of the 14 patients, who had a history of temporomandibular joint disorder, found the SS-OCT imaging of the lower second molar uncomfortable because of difficulties keeping their mouth open. However, the lower first molar of this patient could be safely scanned using SS-OCT. The acquisition time of the 3D dataset in the SS-OCT system was approximately 4 s. While patients were required to keep their mouth open during the 3D scanning, the extension of acquisition time to take several images should be avoided in patients with temporomandibular joint disorder. As a result, one premolar and 19 molars of 10 first molars and 9 second molars from 12 healthy volunteer patients were used in this study.

In SS-OCT, the entire enamel layer of the molar occlusal surface, as well as dentinoenamel junction (DEJ) with underlying dentin, was displayed on the monitor (Fig. 1). It has been reported that the demineralization of enamel increases the scattering coefficient of the structure, which in turn increases the brightness of OCT image [14-16]. The results of this study also show that OCT images of enamel demineralization for ICDAS codes 1 and 2 are associated with an increase in brightness and are significantly different from the surrounding healthy enamel (Figs. 3, 4). Due to the increased brightness of OCT, there were also variations in depth and area images between ICDAS code 1 and code 2 fissures (Figs. 3, 4). In code 3, due to the distinct changes in occlusal fissure in the enamel, image alterations of SS-OCT were also evident, demonstrating the morphological deterioration and increased scattering of the roughened enamel (Fig. 5).

Although ICDAS is based on visual evaluation, initial enamel demineralization is differentiated into code 1 and code 2 since moisture affects the optical appearance of teeth [17,18]. In enamel demineralization, water usually clogs the pores of demineralization and obscures the appearance of lesions in a similar refractive index. Code 1 is the enamel demineralization that can only be found when enamel is dehydrated, where demineralization is confined to the superficial enamel (Fig. 3). Code 2 is the enamel demineralization that can be identified even under the presence of moisture, where demineralization penetrates into deeper enamel (Fig. 4). In this experiment, the results of SS-OCT from the examiners with less than two years clinical experience showed excellent agreement with the results of definitive diagnosis performed by calibrated dentists.

In this experiment, there are several limitations. If caries have demineralized enamel with no cavitation, they can be treated noninvasively using appropriate remineralization therapy; caries-related changes in enamel with ICDAS codes from 1 to 3 can be treated by specific teeth brushing and remineralization therapy. Therefore, in this study, the patient's teeth were not cut histologically to validate the real penetration depth of caries.

Moreover, only enamel caries were diagnosed in this study since ICDAS can outperform the diagnostic accuracy of enamel caries via optical changes in enamel [17,19]. However, it is necessary to evaluate the diagnostic accuracy of SS-OCT for caries, including dentin caries. In a previous experiment carried out by the authors using extracted teeth, the accuracy of OCT diagnosis by dentists with over 9 years of clinical experience was higher than that carried out by dentists with less than 3 years of experience [10]. Therefore, it is highly probable that dentists can improve the diagnostic accuracy of caries using SS-OCT system through clinical experience. Moreover, dental OCT has the potential to provide caries management and monitoring, determining the remineralization state of enamel demineralization using tomographic images [16]. Further evidence is necessary to evaluate enamel demineralization and remineralization in order to develop an accurate diagnosis for occlusal enamel caries.

Ethical Approval

This study design was approved by the Institutional Review Board of Tokyo Medical and Dental University (approval number D2013-013 and D2013-0014) and conducted on the patients who visited the university hospital at Tokyo Medical and Dental University. The study protocol had been explained to the subjects before the involvement in the study and written informed consent for the participation in the study was obtained.

Consent to Participate

The subjects in this clinical study gave their informed consent by signing a written consent form for participation in the research project.

Author Contributions

Conceptualization: TT, NH, TS, YS; Methodology: TT, NH, TS, KK, AT, RF, KD, RN, HI, TG, YT, XC, GI; Validation: TT, NH, YS; Formal analysis, TT, NH; Investigation: TT, NH, KK, AT, RF, KD, RN, HI, TG, YT, XC, GI; Data curation: TT, NH; Writing—original draft preparation: TT, NH, YS; Writing—review and editing: TT, NH, TS, YS; Supervision: TS, YS; Project administration: TS, GI, YS; Funding acquisition: YS

Conflict of Interest

Yasushi Shimada received a research grant from the Yoshida Dental Mfg. Co., Ltd. The other authors declare that they have no conflicts of interest to disclose.

Data Availability Statement

All data generated during this study are available from the corresponding author on reasonable request.

Funding

This research received no external funding.

ORCID iD

• 1) YS*: [e-mail], https://orcid.org/0000-0002-0751-4819

  1) TT: [e-mail], https://orcid.org/0000-0001-9423-2424

  2) HN: [e-mail], https://orcid.org/0000-0003-3364-0031

  1) KK: [e-mail], NA

  1) AT: [e-mail], NA

  1) RF: [e-mail], NA

  1) KD: [e-mail], NA

  1) RN: [e-mail], NA

  1) HI: [e-mail], NA

  1) TG: [e-mail], NA

  1) YT: [e-mail], NA

  1) XC: [e-mail], https://orcid.org/0000-0002-8357-0863

  2) TS: [e-mail], https://orcid.org/0000-0001-5515-7345

  1) GI: [e-mail], https://orcid.org/0000-0002-7933-0832

Acknowledgments

The authors are grateful to the Yoshida Dental Mfg. Co., Ltd. for their support in using the dental OCT system.

References

• 1. Featherstone JD. The continuum of dental caries - evidence for a dynamic disease process. J Dent Res 2004; 83(Spec C): C39-42. doi: 10.1177/154405910408301s08
• 2. FDI World Dental Federation. FDI policy statement on minimal intervention dentistry (MID) for managing dental caries: adopted by the general assembly: September 2016, Poznan, Poland. Int Dent J 2017; 67: 6-7. doi: 10.1111/idj.12308
• 3. Takagi S, Liao H, Chow LC. Effect of tooth-bound fluoride on enamel demineralization/ remineralization in vitro. Caries Res 2000; 34: 281-8. doi: 10.1159/000016603
• 4. Chow LC, Takagi S, Carey CM, Sieck BA. Remineralization effects of a two-solution fluoride mouthrinse: an in situ study. J Dent Res 2000; 79: 991-5. doi: 10.1177/00220345000790041601
• 5. Wenzel A, Fejerskov O. Validity of diagnosis of questionable caries lesions in occlusal surfaces of extracted third molars. Caries Res 1992; 26: 188-94. doi: 10.1159/000261441
• 6. Bader JD, Shugars DA, Bonito AJ. Systematic reviews of selected dental caries diagnostic and management methods. J Dent Educ 2001; 65: 960-8.
• 7. Rodrigues JA, Hug I, Diniz MB, Lussi A. Performance of fluorescence methods, radiographic examination and ICDAS II on occlusal surfaces in vitro. Caries Res 2008; 42: 297-304. doi: 10.1159/000148162
• 8. Jablonski-Momeni A, Stucke J, Steinberg T, Heinzel-Gutenbrunner M. Use of ICDAS-II, fluorescence-based methods, and radiography in detection and treatment decision of occlusal caries lesions: an in vitro study. Int J Dent 2012; 2012: 371595. doi: 10.1155/2012/371595
• 9. Shimada Y, Sadr A, Burrow MF, Tagami J, Ozawa N, Sumi Y. Validation of swept-source optical coherence tomography (SS-OCT) for the diagnosis of occlusal caries. J Dent 2010; 38: 655-65. doi: 10.1016/j.jdent.2010.05.004
• 10. Nakagawa H, Sadr A, Shimada Y, Tagami J, Sumi Y. Validation of swept source optical coherence tomography (SS-OCT) for the diagnosis of smooth surface caries in vitro. J Dent 2013; 41: 80-9. doi: 10.1016/j.jdent.2012.10.007
• 11. Nakajima Y, Shimada Y, Sadr A, Wada I, Miyashin M, Takagi Y, et al. Detection of occlusal caries in primary teeth using swept source optical coherence tomography. J Biomed Opt 2014; 19: 16020. doi: 10.1117/1.JBO.19.1.016020
• 12. Shimada Y, Nakagawa H, Sadr A, Wada I, Nakajima M, Nikaido T, et al. Noninvasive cross-sectional imaging of proximal caries using swept-source optical coherence tomography (SS-OCT) in vivo. J Biophotonics 2014; 7: 506-13. doi: 10.1002/jbio.201200210
• 13. Espinoza K, Hayashi J, Shimada Y, Tagami J, Sadr A. Optical coherence tomography for patients with developmental disabilities: a preliminary study. Sensors (Basel) 2021; 21: 7940. doi: 10.3390/s21237940
• 14. Jones RS, Darling CL, Featherstone JD, Fried D. Imaging artificial caries on the occlusal surfaces with polarization-sensitive optical coherence tomography. Caries Res 2006; 40: 81-9. doi: 10.1159/000091052
• 15. Jones RS, Fried D. Remineralization of enamel caries can decrease optical reflectivity. J Dent Res 2006; 85: 804-8. doi: 10.1177/154405910608500905
• 16. Hariri I, Sadr A, Nakashima S, Shimada Y, Tagami J, Sumi Y. Estimation of the enamel and dentin mineral content from the refractive index. Caries Res 2013; 47: 18-26. doi: 10.1159/000342416
• 17. Gugnani N, Pandit IK, Srivastava N, Gupta M, Sharma M. International caries detection and assessment system (ICDAS): a new concept. Int J Clin Pediatr Dent 2011; 4: 93-100. doi: 10.5005/jp-journals-10005-1089
• 18. Pitts NB, Ekstrand KR. International caries detection and assessment system (ICDAS) and its international caries classification and management system (ICCMS) - methods for staging of the caries process and enabling dentists to manage caries. Community Dent Oral Epidemiol 2013; 41: e41-52. doi: 10.1111/cdoe.12025
• 19. Carvalho RN, Letieri ADS, Vieira TI, Santos TMPD, Lopes RT, Neves AA, et al. Accuracy of visual and image-based ICDAS criteria compared with a micro-CT gold standard for caries detection on occlusal surfaces. Braz Oral Res 2018; 32: e60. doi: 10.1590/1807-3107bor-2018.vol32.0060