Caries detection and characterization in pediatric patients using iTero 5D near-infrared technology

Introduction

Near-infrared imaging (NIRI) has been proposed as an alternative to radiographs and uses nonionizing radiation in the near-infrared spectrum to differentially scatter light off tooth surfaces and generate images allowing interproximal caries detection. The new iTero 5D Element Scanner (Align Technology, Santa Clara, Calif) has integrated NIRI capture and viewing technology but has not been specifically studied in a pediatric population. Therefore, this study aimed to assess clinicians’ abilities to detect and characterize caries in pediatric patients using this instrument.

Methods

Bitewing (BW) radiographs and an intraoral scan were captured on 17 pediatric patients (344 surfaces were analyzed). Data were randomized and graded by 5 calibrated clinicians individually with 2 different rounds of grading.

Results

The reliability of lesion characterization (ie, grade) among examiners was poor to fair in both systems, whereas the reliability of caries detection was moderate. Both systems had a high specificity and low sensitivity. The reliability of the characterization of the combined dataset was moderate to substantial, whereas, for detection, it was substantial.

Conclusions

When using either BW or NIRI analysis, reliability is relatively poor, and clinicians are more likely to correctly identify a healthy tooth surface when compared with a carious surface. There is a small difference in error rate between BW and NIRI systems that is not likely to be clinically significant. When NIRI and BW data are combined, clinician agreement for both lesion characterization and detection increases significantly.

Highlights

•

Evaluation of the caries detection capability of Itero 5D.
•

Assessing the reliability of caries detection of iTero 5D relative to bitewing radiographs.
•

Caries detection comparison was completed using both adult and primary teeth.

Dental caries is the most common chronic disease in children aged 6-19 years and can have several negative sequelae. Specifically, caries has been linked to lower oral health–related quality of life in children and pain, impacting sleeping, eating, schooling, and socializing. ^, Identification of lesions in the early stages of disease progression allows clinicians to employ minimally invasive techniques to delay or prevent placement of restorations and conserve tooth structure.

The traditional method of interproximal caries detection combines radiographic and visual-tactile analysis. This method has been shown to be successful at detecting caries that have progressed into dentin. Radiographs are necessary as it has been shown that a clinical examination alone is inadequate and results in undetected lesions. ^, A main shortcoming of this method is the reality of the small but real risks of using ionizing radiation for routine caries monitoring. It is the consensus of the medical community that the as low as diagnostically acceptable principle is a sound method for minimizing potential risks to patients. The American Dental Association recommends posterior bitewing (BW) radiographic assessments in children every 6-24 months depending on factors such as caries risk, presence of disease indicators, and clinical judgment. Compared with the current standard, the ideal caries detection instrument should have a high specificity, accuracy, precision, ease of use, repeatability, an ability to detect caries at all stages of disease progression, and importantly, pose no risk to patient’s health.

Nonionizing radiation for detecting dental caries has been studied since the 1960s and is a potential alternative to the current standard. Enamel and dentin structure and composition cause unique scattering effects of light in the region of the electromagnetic spectrum between 0.7 to 2.0 micrometers, known as the near-infrared (NIR) region. The crystalline structure of enamel weakly scatters light in this region, whereas the more complex microstructure of dentin strongly scatters light. When enamel and/or dentin become affected by the caries process, the structure of each tissue undergoes a change that subsequently affects the light-scattering properties and allows for differentiation between healthy and carious tissues. ^, These unique properties allow for detecting dental caries using Near-infrared imaging (NIRI) technology.

NIRI may offer several advantages over BW radiographs for interproximal caries detection. Perhaps most notably, more frequent caries monitoring is possible as the technology uses nonionizing radiation and poses no health risks to patients. This allows clinicians to capture as many images as they would like in each setting as opposed to a practitioner who may accept nonideal radiographs instead of exposing a patient to additional radiation. These advantages may lead to an enhanced ability to detect caries. NIRI technology has been available with supporting clinical evidence since the 1990s. It has been shown to be effective, specific, and accurate for interproximal caries detection in-vivo in various clinical studies. There is also evidence that NIR images can detect differences in demineralization, staining, developmental defects, and cavitations. However, most dental professionals have no experience with NIRI technology and are neither trained to read images nor taught how to apply the information clinically.

New advances in digital intraoral dental scanners have allowed the incorporation of NIRI into these instruments. The iTero Element 5D scanner (Align Technology, Santa Clara, Calif) is equipped with NIRI technology and will likely become the first widespread instrument in dental offices with this capability. The scanner uses light at a wavelength of 850 nm and collects real-time, 3-dimensional NIR during routine intraoral scanning. The dental professionals who use this scanner and future NIRI-equipped intraoral scanners will acquire many diagnostic NIR images. Given that most dentists are unfamiliar with this technology, it is critical that research be conducted using this specific intraoral scanner to make evidence-based recommendations available for clinicians.

The goal of this research is to assess the accuracy and reliability of clinicians in detecting supragingival interproximal caries lesions using BW radiographs compared with the NIRI index test (captured by the iTero Element 5D scanner) in a pediatric population. We subsequently aim to determine if the NIRI technology of the iTero Element 5D scanner can be used as a nonionizing alternative to characterize interproximal caries in children with a previously established lesion classification system to determine the reliability and reproducibility of given grades. The hypothesis is that the diagnostic testing parameters of accuracy and examiner reproducibility of the NIRI index test perform better than or equal to the clinical reference standard BW radiographs in detecting supragingival interproximal caries lesions in pediatric patients.

Material and methods

This multicenter, observational case-control clinical study was approved by the University of the Pacific Institutional Review Board (no. IRB2021-34). Carious and noncarious surfaces were identified and used for this study from March 2021 to June 2022 from a private dental office and the Arthur A. Dugoni School of Dentistry, University of the Pacific.

We used previously published studies to assess the range of surfaces, standard deviation, and effect sizes needed for our study. ^, We set the parameters of our analysis at an α of 0.05 and β of 0.2. Equivalence studies typically require larger samples, and we sought to use as many surfaces as we could obtain in the period of study. We set our threshold of equivalence to a clinically significant 5% difference in error rates and applied it to our current sample of 344 surfaces. A post-hoc power analysis using α of 0.05, β of 0.2, and a stringent error rate of 1% as parameters confirmed our sample of 344 surfaces was effectively well powered (sample power >0.999) to detect a difference of 5% error rate ( Table I ).

Table I

Post-hoc power calculation

Alpha	Power	n	P ₀value	P _avalue	Delta
0.05	0.2642	100	0.0004	0.01	0.0096
0.05	0.5954	200	0.0004	0.01	0.0096
0.05	0.8024	300	0.0004	0.01	0.0096
0.05	0.9095	400	0.0004	0.01	0.0096

Note. Power calculations were performed using an α of 0.05 and β of 0.2.

Inclusion criteria included patients with mixed or permanent dentition aged ≤18 years who had BW radiographs taken as part of routine dental care and fully erupted first permanent molars. BWs were captured before selection for the study as part of an examination only when judged necessary by the treating clinician according to the standard of care. Once a lesion was identified, patients/parents were asked to consent to the study, after which an iTero scan of both maxillary and mandibular arches was obtained ( Fig 1 ). No radiographs were taken for the sole purpose of this study. The inclusion criteria for tooth surfaces were any healthy posterior primary or permanent teeth or posterior primary or permanent teeth with lesions extending up to the inner proximal one third of dentin. Exclusion criteria included patients with syndromic or genetic conditions affecting enamel or dentin structure, lesions extending into the inner one third of dentin, or those into the pulp. Previously restored tooth surfaces, nonproximal surfaces, and anterior teeth were excluded from this study ( Figs 2 and 3 ).

BW images were grouped by patient, labeled using an external ID for deidentification, and transferred to university computers. Radiographs were viewed using MiPACs software (Medicor Imaging, Charlotte, NC). Images were graded according to the International Caries Classification and Management System (ICCMS). In this system, the following grades are given: 0 (no radiolucency), 1 (radiolucency in the outer half of the enamel), 2 (radiolucency in the inner half of the enamel ± dentinoenamel junction [DEJ]), 3 (radiolucency limited to the outer one third of the dentin), and 4 (radiolucency reaching the middle one third of the dentin or beyond). Scores of 5-6 represent extensive lesions and were excluded from this study. Scores of 1-2 indicate caries confined to enamel, whereas scores 3-4 indicate the lesion extended into dentin. Lesion scores (BW and NIRI) were marked on a provided grading form ( Fig 4 ) and recorded for later interpretation.

An iTero Element 5D intraoral scanner was used to capture NIRI data at a wavelength of 850 nm. A 3-dimensional (3D) model of the teeth and color images are captured simultaneously during the scan. The scanner operator (K.C.) was either a dental professional previously trained by the practicing orthodontist or a second-year orthodontic resident.

Graders viewed the resulting data on a computer using the viewing interface available on Myitero.com ( Fig 1 ). This displays the 3D scan and color images alongside the NIR images without identifiable patient information. For all graders, this was their first experience interpreting NIRI data. Each grader was calibrated and trained during an in-person session using the clinical guide provided by Align Technology for iTero Element 5D users. Example images were displayed, and scoring was performed and discussed during the calibration session. After individual practice grading, a group discussion was held to review the example lesions. All judges were faculty and had been previously calibrated for bitewings with the ICCMS standard as part of university requirements. Grading was completed in a randomized order, and a modified version of the ICCMS grading scale was used to classify lesions on the basis of previous studies. ^, The scale included 0 (sound surface), 1 (first visible signs restricted to the enamel), 2 (enamel caries with an isolated spot reaching the DEJ), 3 (dentin caries penetrating the DEJ with scattering limited to the outer one third of dentin), and 4 (dentin caries with scattering reaching the middle one third of the dentin). Extensive dentin caries and caries into the pulp comprised lesion grades 5 and 6 and were excluded from this study.

Five graders (general and pediatric dentists) with at least 10 years of clinical practice experience were trained and calibrated as described previously. The NIRI and BW images were initially viewed independently in a randomized order with no possible connection between datasets from the same patient. Interexaminer reliability was determined after grading was completed via weighted κ calculations. After the initial round of randomized grading, NIRI and BW images from 8 patients were redistributed in a combined dataset, and examiners were asked to grade surfaces for each patient using both sources of information. Because of an unforeseen circumstance, only 4 of the 5 graders participated in this second round of examination. For assessing accuracy parameters between BW and NIRI, the median score among all graders was taken for each surface. The accuracy of lesion characterization (grade 0-4) and caries detection (carious vs noncarious surfaces) were analyzed. In the detection evaluation, a value of 0 was considered a noncarious surface, whereas a value of ≥1 was considered a carious surface.

Statistical analysis

Data were inputted and cleaned on Excel (Microsoft, Redmond, Wash), then transferred to Stata (TX) for statistical analysis. Interrater reliability was measured using linear weighted Cohen’s κ. Values were graded at ≤0 (no agreement), 0.01-0.20 (poor), 0.21-0.40 (fair), 0.41-0.60 (moderate), 0.61-0.80 (substantial), and 0.81-1.00 (almost perfect agreement). Overall accuracy was calculated as a percentage of the total number of true positive and true negative values in relation to the total number of diagnostic findings. The independence of NIRI and BW data was assessed using a paired chi-square test. Specificity, sensitivity, and negative and positive predictive values were also calculated to compare NIRI images and bitewings.

Results

Seventeen patients aged ≤18 years were included in the study, from which 352 surfaces were available for analysis. Eight surfaces were discarded because of significant interproximal overlap on BW images. A total of 344 nonrestored posterior surfaces with diagnostic quality imaging were included in the study. Of these surfaces, 283 were permanent tooth surfaces, and 61 were primary tooth surfaces. Caries prevalence for the sample was 11.34%.

Interrater reliability among the different judges to each other for BW and NIRI is shown in Table II . When lesion grade was analyzed, there was generally poor to fair agreement among the judges in both systems. The reliability of the median grades of the 5 graders for BW and NIRI is shown in Tables III and IV , respectively. A grade of 0 had fair reliability in both systems, whereas other grades generally had poor agreement. The exception was a grade of 2 for BW and a grade of 4 for NIRI, which both had fair agreement. When caries detection was evaluated, the κ value was 0.4774 for BW and 0.4001 for NIRI, which were statistically significant and indicated moderate reliability. Finally, interrater reliability values for the combined dataset, in which graders were given both BW and NIRI data for the same patient together, are shown in Table V . Reliability of lesion characterization was moderate, with substantial agreement for a grade of 0 (noncarious) and 4 (caries into one third of dentin and beyond). The reliability of lesion detection of the combined data was substantial, with an κ of 0.6463.

Table II

Only gold members can continue reading. Log In or Register to continue