Skip to main content
Journal of Dental Hygiene Science logoLink to Journal of Dental Hygiene Science
. 2026 Mar 31;26(1):44–52. doi: 10.17135/jdhs.2026.26.1.44

Microscope-Integrated Quantitative Fluorescence Imaging for Assessing Crack Severity in Teeth

Hye-Min Ku 1,“
PMCID: PMC13671813  PMID: 43059638

Abstract

Background

This study investigated the fluorescence characteristics associated with cracks and evaluated the relationship between crack depth and fluorescence-derived variables using quantitative light-induced fluorescence-digital (QLF-D) and microscope-QLF hybrid techniques. Agreement between maximum fluorescence loss (ΔFmax) values obtained from the two modalities was also assessed.

Methods

Extracted teeth with confirmed cracks were examined using QLF-D and a microscope-QLF hybrid technique integrating a fluorescence light source with an operating microscope. Fluorescence-derived variables, including ΔFmax and red fluorescence (ΔR) and maximum red fluorescece (ΔRmax) variables, were quantified using dedicated analysis software. The crack depth was measured using micro-computed tomography as the reference standard. Associations between the crack depth and fluorescence-derived variables were analyzed using Spearman’s rank correlation analysis. Agreement between ΔFmax values obtained from the two imaging modalities was evaluated using Bland-Altman analysis.

Results

Cracks displayed characteristic fluorescence loss and red fluorescence patterns in both imaging modalities. Among the fluorescence-derived variables, ΔFmax demonstrated the strongest correlation with crack depth for both the QLF-D and the microscope-QLF hybrid techniques (p<0.01). The corresponding ΔFmax values obtained from the two modalities were strongly correlated. Bland-Altman analysis indicated a small systematic bias, with the microscope-QLF hybrid technique slightly overestimating ΔFmax; however, all measurements remained within the 95% limits of agreement.

Conclusion

The microscope-QLF hybrid technique showed a measurable and consistent relationship with QLF-D for quantitative fluorescence assessment. Although systematic differences were observed, the hybrid technique may serve as a supportive tool for assessing relative crack severity during endodontic procedures.

Keywords: Cracked tooth, Diagnosis, Fluorescence, Operating microscope, Quantitative light-induced fluorescence

Introduction

Cracked teeth present a significant challenge in clinical dentistry, serving as the primary etiology of persistent pain, endodontic pathologies, and localized destruction of periodontal tissues. These cracks are clinically categorized into various manifestations, including cracked teeth, split teeth, and vertical root fractures (VRF), depending on their extent and orientation1). In the incipient stage, clear crack lines are frequently indiscernible on conventional periapical radiographs2). This diagnostic latency often facilitates the infiltration of pathogenic bacteria along the fracture line, precipitating the formation of deep periodontal pockets and acute or chronic destruction of the surrounding bone tissue3-5). Consequently, these delays lead to an unfavorable prognosis, frequently culminating in tooth extraction6). Therefore, early and precise detection of tooth fractures coupled with quantitative assessment of crack depth is imperative for establishing an evidence-based treatment plan and accurately predicting tooth survival.

In contemporary clinical practice, several modalities are used to diagnose root fractures, including periodontal probing, methylene blue staining via exploratory surgery, and advanced imaging techniques such as cone beam computed tomography (CBCT). However, visual inspection and tactile probing are inherently subjective and highly dependent on clinician experience and acuity7-9). While methylene blue staining offers enhanced visualization, it is a fundamentally invasive procedure requiring surgical access8). Radiographic examinations also have substantial limitations; the detection rate diminishes significantly when the fracture plane does not align precisely with the primary X-ray beam or when the fracture width is microscopic10,11). Furthermore, the presence of radiopaque materials such as metal posts or restorative crowns introduces significant imaging artifacts that obscure the fracture line and compromise diagnostic reliability12). These constraints collectively hinder not only the definitive identification of fractures but also the quantitative evaluation of their depth, which is critical for clinical decision-making.

To overcome these limitations, quantitative light-induced fluorescence (QLF) technology has emerged as a promising non-destructive diagnostic adjunct. QLF utilizes the principle of autofluorescence, where tooth structures emit specific fluorescence signals when irradiated with visible light at 405 nm13). While sound enamel exhibits a characteristic tooth color, areas characterized by mineral loss or structural disruptions, such as fractures, manifest as “fluorescence loss,” appearing darker than the surrounding sound enamel14). Porphyrin, a metabolic byproduct of anaerobic bacteria residing within the fracture interstices, emits a distinct red fluorescence under a 405 nm light source15,16). This phenomenon allows for the vivid visualization of even minute fracture lines that are otherwise invisible to the naked eye9,17). Prior research has demonstrated that QLF-digital (QLF-D) systems possess high sensitivity and specificity for fracture detection, establishing a significant correlation between fluorescence loss parameters and the actual depth of the fracture17-19).

Despite their diagnostic potential, the routine clinical implementation of dedicated QLF-D devices faces practical hurdles. Practical constraints such as high equipment costs and the necessity for rigorous infection control hinder the widespread adoption of QLF-D in clinical settings. Utilizing an independent imaging system during complex microsurgical interventions can compromise the workflow efficiency and focus required by clinicians. In contrast, the operating microscope has become an indispensable standard in modern endodontics and microsurgery, providing high-magnification visualization and illumination20-22). If the QLF functionality could be integrated directly into the optical pathway of an operating microscope, clinicians would be able to obtain real-time quantitative diagnostic data regarding crack sites without departing from the surgical field. Such integration would not only enhance diagnostic precision but also maximize clinical efficiency by streamlining the procedural workflow.

The objective of this study was to evaluate and compare the diagnostic performance of the conventional QLF-D system and a newly developed ‘microscope-QLF hybrid technique’ using extracted crack teeth. Using micro-computed tomography (CT) as the gold standard for measuring the actual root crack depth, this study sought to verify the correlation between crack depth and the fluorescence parameters produced by both devices, namely, the maximum fluorescence loss and the prevalence of red fluorescence. Through this comparative analysis, we aimed to validate the clinical utility of microscope-based fluorescence technology as a robust tool for quantitative evaluation of cracked teeth in a real-world clinical setting.

Materials and Methods

1. Specimen collection and ethics statement

The study protocol was reviewed and approved by the Institutional Review Board of the Yonsei University Dental Hospital (IRB No. 2-2017-0038). All participating patients provided informed consent before donating their teeth for research purposes. A total of 33 cracked teeth were collected from patients who visited the hospital for the diagnosis of suspected tooth fractures, including cracked teeth, split teeth, and VRF. The definitive diagnosis of a crack was confirmed through clinical and radiological evaluations after diagnostic surgery or intentional replantation procedures.

2. Sample preparation and preservation

Upon extraction, the cracked teeth were immediately stored in 50 ml conical tubes and shielded from ambient light to preserve their optical properties. To ensure a clean surface for fluorescence analysis, extraneous granulation tissue and calculus were meticulously debrided from the root surfaces using a No. 15 scalpel blade (Paragon, Sheffield, UK). The samples were then fixed in 10% neutral-buffered formalin solution (Sigma Chemical Co., St. Louis, MO, USA) for 24 hours, starting from the day of extraction.

3. Crack detection modalities

Tooth surface moisture was gently eliminated using a cotton pellet prior to imaging. The specimens were subsequently examined using the designated crack-detection systems. A total of 33 cracked teeth were imaged under light-controlled conditions with ambient illumination eliminated, and the acquired images were processed and analyzed using dedicated proprietary software. Initial verification of the cracks was conducted using an operating microscope (OPMI PICO; Carl Zeiss, Göttingen, Germany) at magnifications ranging from 16× to 25× under halogen illumination. Images were recorded using a dedicated adapter system. Subsequently, the specimens were imaged using QLF-D. A novel hybrid technique combining a microscope with QLF (microscope-QLF hybrid technique) was also employed to capture fluorescence aspects, followed by micro-CT scanning for precise depth analysis.

4.Quantitative light-induced fluorescence evaluation

Fluorescence signals from the cracked areas were captured using a QLF-D Biluminator 2+ camera (Inspektor Research Systems BV, Amsterdam, Netherlands). This system utilizes two light sources: 12 blue LEDs (405 nm peak wavelength) for fluorescence excitation and four white LEDs for standard imaging. For white light, the camera settings were 1/45s shutter speed, f/13.0 aperture, and International Organization for Standardization (ISO) 1600. For the blue light, the settings were adjusted to 1/10s, f/11.0, and ISO 1600. A specialized high-pass filter (λ>520 nm) was used to isolate the autofluorescence from the tooth structure. All images were captured at a standardized distance of 10 cm.

Quantitative analysis was performed using QA2 v1.26. (Inspektor Research Systems BV, Amsterdam, Netherlands). The region of interest (ROI) was defined as an area extending 3 mm above the root apex and 3 mm below the coronal portion of the root. Maximum fluorescence loss (ΔFmax [%]) was calculated by comparing the crack site with the surrounding healthy tooth structure. Red fluorescence, associated with bacterial porphyrin, was quantified as red fluorescence (ΔR) and maximum red fluorescece (ΔRmax). Red fluorescence was independently assessed by two examiners.

5.Microscope-quantitative light-induced fluorescence hybrid technique

Autofluorescence was further evaluated using the microscope-QLF hybrid technique, in which a light source (Qray-view; AIOBIO Inc., Seoul, Korea) was positioned parallel to the dental operating microscope lens (OPMI PICO) (Fig. 1). A specific filter set (λ>520 nm) was attached to the objective lens to suppress back-scattered light, allowing for the observation of tooth auto-fluorescence through the eyepieces at 16× to 26× magnifi-cation. The image parameters were analyzed using the same software (QA2 v1.26) for consistency. All fluorescence parameters (ΔFmax, ΔR, and ΔRmax) were calculated for all specimens (n=33) using the same ROI definition, regardless of the presence or absence of visible red fluorescence.

Fig. 1.

Fig. 1

Microscope-quantitative light-induced fluorescence hybrid technique.

6.Micro-computed tomography for depth measurement

The specimens were scanned using a Skyscan 1076 micro-CT device (SkyScan, Aartselaar, Belgium) with the following settings: 100 kV accelerating voltage, 100 µA beam current, a 0.5 mm Al filter, and 18 µm resolution. The scan involved a 360° rotation with 0.5° increments. The deepest crack point was measured in 3D views using DataViewer software (SkyScan, Kontich, Belgium), excluding the 3-mm coronal and apical margins. Measurements were performed in triplicate by a single calibrated operator to ensure reliability.

7. Statistical analysis

Statistical computations were performed using R language (R Foundation for Statistical Computing, Vienna, Austria). Fluorescence parameters (ΔFmax, ΔR) were analyzed as absolute values. Prior to correlation analysis, the normality of the crack depth and fluorescence-derived variables (ΔFmax, ΔR, and ΔRmax) was assessed using the Shapiro-Wilk test. As several variables did not satisfy the assumption of normal distribution (p<0.05), a non-parametric Spearman’s rank correlation analysis was used to evaluate the associations between the crack depth and fluorescence variables. Additionally, Bland-Altman plots were generated to compare the diagnostic agreement between the new microscope-QLF hybrid technique and the standard QLF-D method.

To assess the intra-examiner reliability of the crack-depth measurements obtained from micro-CT, repeated me-asurements were performed at a two-week interval for all specimens. The intra-class correlation coefficient (ICC) was calculated using a two-way mixed-effects model with absolute agreement. Inter-examiner agreement for the presence or absence of red fluorescence was evaluated using Cohen’s kappa coefficient. Kappa values were calculated based on independent binary assessments performed by two examiners. Linear regression analysis was additionally performed to evaluate the predictive relationship between ΔFmax and crack depth for both QLF-D and the Microscope-QLF hybrid technique. The coefficient of determination (R2) and regression equations were reported.

Results

1. Evaluation of fluorescence characteristics

All the specimens included in this study exhibited distinct crack lines when observed under high magnifi-cation using an operating microscope (Fig. 2A). When evaluated using fluorescence-based imaging, the crack lines demonstrate a marked loss of fluorescence and appear darker than the surrounding sound tooth structures. In addition, red fluorescence along the crack line was observed in 57% of the specimens included in this study (Fig. 2B, 2C). The spatial distribution and intensity of red fluorescence detected using the microscope-QLF hybrid technique were consistent with those observed with the conventional QLF-D system. Intra-examiner reliability of crack depth measurements was high, yielding an ICC of 0.90. Inter-examiner agreement for the assessment of red fluorescence presence was also high, with a Cohen’s kappa coefficient of 0.84.

Fig. 2.

Fig. 2

Representative images of a cracked tooth obtained using different imaging modalities. (A) Operating microscope image showing the crack line (red arrow). (B) Fluorescence image acquired using quantitative light-induced fluorescence (QLF)-digital, demonstrating fluorescence loss and red fluorescence along the crack line (red arrow). (C) Fluore-scence image obtained using the microscope-QLF hybrid technique, showing a similar fluorescence pattern (red arrow).

2.Correlation between crack depth and fluorescence variables

Fluorescence variables obtained using the microscope-QLF hybrid technique demonstrated statistically significant correlations with crack depth. The correlation coefficients were comparable to those observed with QLF-D, with the highest correlation value between ΔFmax and crack depth (Table 1). Spearman’s rank correlation analysis revealed significant correlations between crack depth and all fluorescence variables measured using QLF-D (Table 1). Among these variables, ΔFmax exhibited the strongest correlation with crack depth (r=0.89, p<0.01).

Table 1.

Spearman’s Rank Correlation Coefficients (r) between Crack Depth and Fluorescence Variables

Variables QLF-D (r) Microscope-QLF hybrid technique (r)
∆Fmax 0.89* 0.82*
∆R 0.65* 0.62*
∆Rmax 0.57* 0.57*

QLF-D: quantitative light-induced fluorescence-digital, ∆Fmax: maximum fluorescence loss, ∆R: red fluorescence, ∆Rmax: maximum red fluorescence.

*Spearman’s rank correlation analysis, p<0.01.

Fluorescence variables from the microscope-QLF hybrid technique also showed significant correlations with crack depth. ΔFmax had the highest correlation coefficient (r=0.82, p<0.01) among the fluorescence-derived variables, followed by ΔR and ΔRmax, which is consistent with the findings from the QLF-D analysis (Table 1).

3.Correlation of fluorescence variables between quantitative light-induced fluorescence-digital and the microscope-quantitative light-induced fluorescence hybrid technique

Spearman’s rank correlation analysis demonstrated significant correlations between the corresponding fluorescence parameters measured using QLF-D and the microscope-QLF hybrid technique (Fig. 3). Among these fluorescence variables, ΔFmax showed the strongest correlation with the two modalities (r=0.91, p<0.01). In contrast, the correlation between ΔR values measured by QLF-D and the microscope-QLF hybrid technique was relatively lower (r=0.41, p<0.05).

Fig. 3.

Fig. 3

Correlation matrix showing Spearman’s rank correlation coefficients between fluorescence variables obtained using QLF-D (Q) and the microscope-QLF hybrid technique (M). Color intensity represents the magnitude of the correlation coefficient. QLF-D: quantitative light-induced fluorescence-digital. *p<0.05, **p<0.01.

4.Linear regression analysis for crack depth prediction

Linear regression analysis demonstrated a significant predictive relationship between ΔFmax and crack depth for both imaging modalities.

For the microscope-QLF hybrid technique, the regression equation was: Crack depth (µm)=58.75×ΔFmax+1390.48 (R2=0.5988, p<0.001). For QLF-D, the regression equation was: Crack depth (µm)=72.13×ΔFmax+1390.48 (R2=0.7380, p<0.001).

5.Agreement between ΔFmax obtained from quantitative light-induced fluorescence-digital and the microscope-quantitative light-induced fluorescence hybrid technique

Based on the strong correlations observed between the ΔFmax values obtained using the two modalities, a Bland-Altman analysis was conducted to assess the agreement between the ΔFmax values from the QLF-D and the microscope-QLF hybrid technique for crack line evaluation. The mean difference indicated that the microscope-QLF hybrid technique slightly overestimated ΔFmax compared to QLF-D, with a mean difference of 3.5%. The 95% limits of agreement ranged from –19.1% to 26.1%, and all measurements fell within these limits (Fig. 4).

Fig. 4.

Fig. 4

Bland-Altman plot showing the agreement between maximum fluorescence loss (∆Fmax) values obtained from the microscope-quantitative light-induced fluore-scence (QLF) hybrid technique and QLF-digital (QLF-D). The solid horizontal line represents the mean difference, while the dashed lines indicate the upper and lower 95% limits of agreement (mean difference±1.96 standard deviation [SD]).

Discussion

This study investigated the fluorescence characteristics of crack lines and evaluated the relationship between crack depth and fluorescence-derived variables using QLF-D and a microscope-QLF hybrid technique. In addition, agreement between ΔFmax measurements obtained from the two modalities was assessed. The principal findings were that ΔFmax showed the strongest association with crack depth regardless of imaging modality and that ΔFmax measurements demonstrated a systematic relationship between QLF-D and the microscope-QLF hybrid technique.

The crack lines exhibited characteristic fluorescence changes when evaluated using fluorescence-based imaging. Fluorescence loss along the crack line has been consistently observed and is thought to result from altered light scattering caused by microstructural disruption within the mineralized tissues19). Red fluorescence was also detected in a substantial proportion of the specimens, reflecting the presence of bacterial metabolites within the crack lines. These observations support the concept that fluorescence imaging can capture both structural and biological changes associated with cracked teeth23).

QLF-D is a commercially available fluorescence imaging system widely used for the detection of caries and structural defects13). However, its clinical application in endodontic practice remains limited due to its high cost and restricted accessibility. To address these limitations, the present study evaluated a microscope-QLF hybrid technique that integrates fluorescence imaging into an operating microscope. The observed correlations between the corresponding fluorescence variables obtained from the two modalities indicated that the fluorescence-derived information acquired using the hybrid technique was generally consistent with that obtained using QLF-D. Based on these findings, ΔFmax emerged as the most relevant fluorescence-derived variable, demonstrating the strongest correlation with fracture depth for both QLF-D and the microscope-QLF hybrid technique18). ΔFmax represents the maximum fluorescence loss within the ROI and is therefore less influenced by local variations along the crack line17,24). The consistency of this relationship across imaging modalities supports the use of ΔFmax as a primary quantitative indicator of crack severity.

While correlation analysis demonstrated strong associations between corresponding measurements obtained from QLF-D and the microscope-QLF hybrid technique, correlation alone does not indicate whether the two methods can be used interchangeably. Therefore, Bland-Altman analysis was performed to evaluate the agreement between the ΔFmax values obtained using the two modalities. This analysis revealed a small systematic bias, with the microscope-QLF hybrid technique slightly overestimating ΔFmax compared with QLF-D. This systematic difference is likely attributable to technical factors inherent to the hybrid imaging setup, including shorter camera exposure time and non-uniform illumination associated with the operating microscope. These factors may result in darker images and, consequently, larger apparent fluorescence loss. Importantly, however, all ΔFmax measurements fell within the 95% limits of agreement, indicating that the observed bias was consistent and limited in magnitude.

From a methodological perspective, the Bland-Altman results suggest that, despite minor systematic differences, ΔFmax measurements obtained using the microscope-QLF hybrid technique are sufficiently comparable to those obtained using QLF-D. This level of agreement indicates that the two methods demonstrate a comparable measurement trend for assessing relative crack severity, particularly in contexts where QLF-D is unavailable. The Bland-Altman analysis was limited to ΔFmax because this variable demonstrated the strongest and most consistent association with fracture depth and showed stable behavior across imaging modalities. Although the Bland-Altman plot revealed a correlation, the relatively wide limits of agreement suggest that the two methods should be considered complementary rather than fully interchangeable. Red fluorescence-related variables such as ΔR and ΔRmax are more sensitive to imaging conditions, including exposure settings and illumination homogeneity, which may reduce the reliability of agreement analysis for these measures. From a clinical perspective, integrating fluorescence imaging into an operating microscope enables simultaneous high-magnification visualization and quantitative assessment of crack severity without the need for additional staining agents. Although the microscope-QLF hybrid technique tended to slightly overestimate ΔFmax, this bias is unlikely to be clinically significant when ΔFmax is used to assess relative crack severity rather than absolute fracture depth. Therefore, the hybrid technique may provide a practical and accessible alternative for fluorescence-based evaluation of cracked teeth in clinical practice.

Recent advances in operating microscope technology have incorporated fluorescence-based visualization as an integrated diagnostic feature. For example, the ZEISS EXTARO 300 microscope provides a dedicated fluorescence mode that facilitates real-time detection of mineral loss and structural alterations in dental hard tissues25,26). However, this fluorescence mode is primarily designed for visual enhancement and does not offer quantitative metrics for objective assessment. In contrast, the present study employed fluorescence imaging to derive numerical parameters reflecting fluorescence loss, enabling quantitative analysis of crack severity. By demonstrating a significant association between fluorescence loss values and fracture depth, as well as a consistent level of agreement across different imaging configurations, these findings highlight the added value of quantitative fluorescence analysis beyond visual detection alone. Overall, while microscope-integrated fluorescence systems enhance clinical detection, quantitative fluorescence assessment provides complementary information that supports objective interpretation and research-level validation of crack extent.

This study has several limitations. First, fluorescence analysis was performed on extracted teeth under controlled experimental conditions, which may not fully replicate the optical environment encountered in vivo. Second, the Bland-Altman analysis was limited to ΔFmax, as this variable showed the most robust association with fracture depth and stable behavior across imaging modalities; therefore, agreement for red fluorescence-related variables was not evaluated. Third, fracture depth was assessed using micro-CT in a standardized orientation, whereas clinical cracks often exhibit complex and non-linear geometries. Further studies are warranted to validate the applicability of fluorescence-based crack assessment under clinical conditions and to determine its impact on treatment decision-making. Additionally, in vivo conditions may introduce optical variability due to saliva, blood contamination, soft tissue interference, and restricted access angles. These factors may attenuate fluorescence intensity and affect quantitative measurements. Although complete dark-room conditions are not always feasible in clinical practice, localized light control in dedicated treatment rooms may allow practical implementation of microscope-integrated fluorescence imaging27).

Within the limitations of this study, fluorescence-based imaging revealed characteristic changes along tooth crack lines, and ΔFmax showed the strongest association with fracture depth regardless of imaging modality. Measurements of ΔFmax obtained using the microscope-QLF hybrid technique demonstrated a systematic relationship with those obtained using QLF-D. These findings suggest that ΔFmax is a reliable quantitative indicator of crack severity and that the microscope-QLF hybrid technique may serve as a practical alternative for fluorescence-based assessment of cracked teeth. Integration of fluorescence imaging into an operating microscope may facilitate objective evaluation of crack depth and support clinical decision-making in endodontic practice.

Acknowledgements

None.

Funding Statement

Funding None.

Notes

Conflict of interest

No potential conflict of interest relevant to this article was reported.

Ethical approval

The study protocol was reviewed and approved by the Institutional Review Board of Yonsei University Dental Hospital (IRB No. 2-2017-0038).

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

References

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.


Articles from Journal of Dental Hygiene Science are provided here courtesy of The Korean Society of Dental Hygiene Science

RESOURCES