Open Access
Add to BibSonomy
Abstract
There remains a clinical need for an accurate and non-invasive imaging tool for intraoral evaluation of dental conditions. Optical coherence tomography (OCT) is a potential candidate to meet this need, but the design of current OCT systems limits their utility in the intraoral examinations. The inclusion of light-induced autofluorescence (LIAF) can expedite the image collection process and provides a large field of view for viewing the condition of oral tissues. This study describes a novel LIAF-OCT system equipped with a handheld probe designed for intraoral examination of microstructural (via OCT) and microvascular information (via OCT angiography, OCTA). The handheld probe is optimized for use in clinical studies, maintaining the ability to detect and image changes in the condition of oral tissue (e.g., hard tissue damage, presence of dental restorations, plaque, and tooth stains). The real-time LIAF provides guidance for OCT imaging to achieve a field of view of approximately 6.9 mm × 7.8 mm, and a penetration depth of 1.5 mm to 3 mm depending on the scattering property of the target oral tissue. We demonstrate that the proposed system is successful in capturing reliable depth-resolved images from occlusal and palatal surfaces and offers added design features that can enhance its usability in clinical settings.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
The most common oral diseases such as dental caries and gingiva-related diseases remain a public health challenge according to a systematic analysis for the global burden of diseases [1]. Visual-tactile assessment is the most conventional method to detect and diagnose caries and gingival diseases; and these assessments rely on the expertise of the clinicians. Several imaging modalities were proposed for detecting dental caries [2,3], and gingiva [4–6], which can be classified into ionizing and non-ionizing methods. Ionizing methods include cone beam computed tomography (CBCT) and micro-CT, which are of limited spatial resolution [2] to be reliable for early detection of caries [3]. The non-ionizing imaging techniques operate either in the visible/UV spectrum (fluorescence and absorption properties) or near-infrared spectrum (scattering properties). Fluorescence technique performs well in detecting advanced caries but shows less sensitivity in detecting early caries [7,8]. Other non-ionizing methods mostly focus on the detection of dental caries, and rely on the distinct changes in optical properties as the caries progresses [9]. Few studies exist examining the sensitivity and specificity of near-infrared imaging technique, yet recent development of such devices show promising potential [10]. In this paper describing an intra-oral probe, we integrated both fluorescence (light-induced auto-fluorescence of the tooth – LIAF) and near-infrared technique (optical coherence tomography – OCT) for the detection of dental caries and gingival diseases.
OCT is a well-established depth-resolved imaging technique, generating images through the time delay and magnitude of light echoes in a similar way to how ultrasonography utilizes sound waves to image tissue structures [11,12]. Due to its non-invasive, cellular-level resolution and real-time capabilities, OCT has rapidly expanded and is currently used in ophthalmology [13–16], endoscopy [17–19], dermatology [20–27] and dentistry [28–35]. For dental applications, there is increased interest in utilizing OCT for the investigation of oral diseases, including mucosal abnormalities [36,37], periodontal diseases [29–32], and caries [38,39]. However, the design of current commercial and non-commercial OCT systems is not flexible to accommodate intraoral, full mouth inspections, resulting in the scan region typically being limited to the buccal side of anterior teeth. There continues to be a need for hand-held or endoscopic OCT probes applicable for intraoral use in the oral cavity.
Several groups have developed customized OCT systems for oral examination [39–43]. Won et al. [40], for example, developed a handheld spectral domain OCT (SD-OCT) capable of 2-D imaging of the molars and surrounding gingivae. They showed specific optical features related to gingivitis, for example, lower signal strength compared to healthy gingival tissue. Schneider et al. [39] modified a commercial SD-OCT to investigate proximal carious lesions, highlighting the potential of OCT to complement current diagnostic methods. However, SD-OCT is limited by a low signal-to-noise ratio at deeper than 1 mm ranging distance (typically 3dB drop per mm), reducing imaging quality in deeper layers such as dentin. This problem is most challenging when the surface topology of scanned tissue is curved, which is often the case for oral tissue. Recent advances in swept source OCT (SS-OCT) make it a more suitable tool for oral imaging because of the better signal-to-noise ratio at longer depth range, negligible sensitivity roll-off, and faster imaging speed [44]. Another extension of OCT, called functional angiography (OCTA), is a powerful tool for evaluating the condition of oral soft tissue since it can image the microvascular information inside the gingiva, without injecting invasive contrast agents. Upon reviewing recent developments, we have identified four major challenges in the adaptation of SS-OCT into a diagnostic tool for dental applications: accessibility, the field of view (FoV), real-time guidance, and OCTA (applicable to soft tissues only). For example, Schneider et al. [39] demonstrated an intraoral probe that could enhance the detection of proximal caries. Their OCT system had good accessibility and FoV; and whilst Schneider did not report any problem with the guidance of the OCT probe, the authors highlighted a need for ‘self-explanatory real-time OCT image stacks’ to simplify the procedure. This suggests that the OCT system might not be readily operable without prior expertise in OCT, and without real-time image guidance. More recently, Li et al. [43] have also reported a handheld 100 kHz SS-OCT, which was capable of imaging mucosa and teeth. The report did not discuss the need for imaging guidance and simply demonstrated the imaging capability with high-resolution OCT images of the soft and hard oral tissues. In addition, this study did not provide evidence that the intraoral OCT could access lingual, palatal, or posterior oral surfaces. All of the in vivo demonstrations were of the buccal surfaces of anterior teeth and the tongue. The imaging probe was also limited to a small FoV, and therefore low applicability in clinical settings.
Moreover, none of the above studies demonstrated the angiography capacity. Tsai et al. [36], Wei et al. [37] and Choi et al. [45], have demonstrated the potential of intraoral OCTA to image the vasculature of the oral mucosa and assess the vascular response to mouth ulcers. However, the authors did not report the ability to image all teeth surfaces (i.e., covering buccal, occlusal, palatal, and lingual sites). Additionally, the implementation of the OCTA intraoral probe in a typical dental clinic would still be challenging due to the lack of real-time imaging guidance and a limited FOV. One solution could be to directly use OCT/OCTA and compress the 3-dimensional information into an en face projection for viewing as a traditional photograph. This approach would be, however, computationally heavy, and would take several seconds to complete one projection image. Another solution could be to integrate a CCD image sensor into the OCT scanning probe to provide real-time guidance [30].
While intraoral OCT/OCTA is a promising tool for providing real time clinical dental information, the field-of-view is relatively small and the scanning probe often blocks the operators’ line of sight, hindering the operator from accurately locating abnormal regions during scanning. To meet these challenges, we combined OCT with light-induced autofluorescence (LIAF) into an intraoral probe. LIAF is also a non-ionizing imaging technique that focus on the autofluorescence and absorption properties of the teeth as caries are formed. In our setup, LIAF is excited by violet light (here 395 nm), and detected by a visible-spectrum CCD camera [30]. In addition to providing a large field-of-view for real-time guidance, LIAF is also useful to distinguish pathological tissues by detecting changes in fluorescence: (i) areas where dental carious lesions are present would appear darker than the healthy tooth due to the suppression of green autofluorescence [46]; (ii) mature dental plaque is also associated with red fluorescence of bacterial metabolites [47]. In this paper, we describe the development of a custom LIAF-OCT(A) system prototype equipped with a handheld probe to access posterior intraoral locations, with little effort and training. The buccal, occlusal, palatal, and lingual surfaces of all teeth can be reached with the handheld probe. Since the lingual and palatal sides, (i.e. the surfaces that are adjacent the tongue), are more challenging to access than the buccal surfaces due to their anatomical locations, we chose to highlight the advantage of this intraoral probe by reporting images taken from the top-right palatal quadrant. Finally, we show the LIAF-OCT system’s capability of capturing images of dental fillings, demineralized enamel, and plaque present in this region of the volunteer.
2. Materials and methods
2.1 System setup
The OCT module employed a 1310 nm swept-source laser with a Michelson interferometer setup (Fig. 1(a)). The swept-source 1310 nm laser was MEMS-VCSEL type (SL1310 V1-20048, Thorlabs Inc., NJ, USA), running at 200 kHz (line-scan rate), and 100 nm bandwidth. The laser source also had an integrated A-trigger and K-clock that triggered the start of acquisition and the sampling interval for a linear k-sampling, respectively. The integrated K-clock was generated by a 48 mm Mach-Zehnder interferometer, giving a typical depth range of approximately 12 mm. During the post-processing step, we cropped this depth range to 8 mm for faster OCT/OCTA processing. The digitizer in our system has a bandwidth of 100 MHz to 1 GHz (ATS-9371, AlazarTech, QC, Canada). Proprietary custom-designed software platform controlled the galvo-scanners by sending positioning signals to the galvo-controller, and the synchronization between the digitizer and the galvo-scanner.
Fig. 1. The schematics of the system design along with the intraoral probe design and its operation. (a) Schematics of the LIAF-OCT system. (b) The design of the intraoral probe comprising a lens tube (left) and a probe housing (right). The lens tube was designed for 10 µm and 50 µm OCT resolution, and the probe housing was designed for both LIAF and OCT. The probe housing had a 395 nm UV LED within the enclosure for LIAF illumination. LIAF: light-autofluorescence imaging; OCT: optical coherence tomography; UV: ultraviolet; LED: light emitting diode. (c) An example of occlusal (left) and palatal (right) scan with the intraoral probe in action.
All optical fibers in the OCT module were single-mode at 1310 nm (SMF-28). From the laser output port, the laser was coupled and traveled through a circulator (Thorlabs Inc., NJ, USA) and then through a 50/50 coupler, splitting into both the reference arm and sample arm with equal power. Typical power in the sample and reference arms was approximately 10 mW. The reference arm was attenuated to approximately 0.4 mW to avoid saturating the detector. In the sample arm, the laser was collimated (F220APC-1310, Thorlabs, NJ, USA) and launched into a pair of galvanometric mirrors (GV S002, Thorlabs, Inc.). The beam then passed through a dichroic mirror inside the probe housing and was focused by a customized lens tube. The power of the sample beam launched into in vivo tissues was 3.5 mW (mostly due to coupling losses). The system had a measured sensitivity roll-off of less than 6 dB full-range and a signal-to-noise ratio of 105 dB [30,45]. The scanning probe was a lightweight piece designed for hand-held configuration (Fig. 1(b)). In this study, the probe was mounted on a kinetic arm, permitting positioning of the probe with 3D translational and rotational motions as described in our previous publication [29]. The scanning of the OCT beam was driven by a pair of galvanometer mirrors (6210H, Cambridge Technology, Bedford, Massachusetts) located within the probe, sweeping the OCT beam in lateral directions of x and y (note z-direction is along the OCT beam).
We designed the probe to optimize repeatability and sensitivity during imaging. In the sections below, we describe the design in detail, including the design of the disposable reflector with LIAF parts and the scanning protocols in order to reach all possible surfaces (buccal, occlusal, lingual/palatal). Figure 1(c) demonstrates that our handheld intraoral probe can easily reach the occlusal and palatal surface of the molar teeth. The functional OCTA is discussed in post-processing section.
The LIAF portion of the setup used a 395 nm LED light source for illumination and an IMX CCD module (IMX322, Sony, USA) for detection. The optical setup integrating LIAF with the OCT beam has been described in our previous publication [30]. Here, we briefly describe our setup and configurations for the excitation light source in the intraoral probe.
The design of the intraoral cap had two important parts: a reflector and 395nm LIAF source. Both parts were designed to be tight-fit and held together by friction. Additional adhesive may be used to ensure a water-proof feature, especially when operated inside the mouth. The reflector was an aluminum-coated mirror positioned at a 45-degree angle and was coated with an anti-fog solution which prevented the mirror from fogging which is typical during intraoral imaging.
The 395 nm LED was driven with a constant current source of 45 mA (100 mW of electrical power). The detector was an IMX322 camera with 16 mm focal length lenses and a wide-band filter with 450 nm cut-on wavelength. When in operation, the camera settings were set with the exposure at 0.0015, the brightness at 100, and the white balance at 6500K, respectively. The raw LIAF was captured before the OCT scan automatically. The raw LIAFs were JPEG-compressed images with RGB channels.
2.2 OCT/OCTA post-processing
The acquired signal from the digitizer was an interferometric signal sampled in linear wavenumber. In the simplest form, the signal acquired from a single scattering event (from a mirror for example) is:
where k is the sweeping wave vector, associated with its spectral intensity $S(k )$ emitted from the laser source; $\zeta $ is the path difference of the mirror relative to the reference arm (where if the distance between reference arm and sample arm is equal, then ζ =0); ${k_0} = \frac{{2\pi }}{{{\lambda _0}}}$ is central wave vector (rad⋅nm-1). Here, we assumed that the DC part from the interference signal was perfectly canceled out by the differential balanced detector. In practice, the DC might be detected due to the un-balanced detection arm caused by the optical circulator (Fig. 1(a)). To obtain the complex OCT signal $I(z )$, the Fourier transform is applied to the interferometric signal and with the help of the convolution theorem, resulting:Fig. 2. Illustration of the interference signal and OCT intensity signal in in vivo enamel/dentin and gingiva. (a) Interference signal as a function of ζ distance (µm) and k wavenumber (nm-1) in a theoretical model with Gaussian-shape spectrum. (b) The real part of interference signal $I(k )$ as detected by the balanced detector in our system. (c) The OCT signal obtained from the Fourier transform of $I(k )$ showing an identical copy of $|{I(z )} |$. Only one copy was used to reconstruct the OCT A-line-scan as indicated by the yellow-box overlay. (d) The obtained OCT A-line -scan was acquired multiple times at different scanning positions, forming a complete B-frame of the in vivo tooth/gingiva, showing depth profiles such as dentin enamel junction (DEJ) and junctional epithelium (JE). (e) Every A-line in the B-frame was then dispersion-corrected to improve the axial-resolution, denoted by the sharpness on the surface reflection of the tooth.
The A-scans were acquired at consecutive positions to form a B-scan, showing a cross-section of the tooth and gingiva (Fig. 2(d)), where in this example the location was at buccal pre-molar (#12 UNS or #24 FDI) side. OCT scanning pattern was rasterized 3D-volumetric scan, where each depth-resolved horizontal scan (or B-frame) were repeated 4 times. We cropped the z-axis from 12 mm to 8 mm to reduce the OCT/OCTA processing time. The final pixel density in the z–x–y direction is 1088–500–500 (nZ–nX–nY) covering a volume of ∼ 8 × 6.9 × 7.8 mm3. To enable angiographic imaging [48], each B-scan was repeated 4 times at the same y-location, before moving to the adjacent (indexed) y+1 location. The 3D k-space data set was then saved and post-processed offline. Post-processing included dispersion correction up to second-order in k-space [49], Fourier-transformation into Cartesian-space, and functional angiogram extraction [50]. The dispersion compensation optimized the sharpness of the B-frame OCT along the z-axis (Fig. 2(e)). In brief, the OCT data was the average intensity signal $I^{\prime}$ and OCTA was a measure of variance in complex signal $I^{\prime}$. These B-frames were then appended iteratively along the y-direction to form 3D-OCT/OCTA. The 3D tomography scans of the oral tissues, with penetration of up to ∼1.5 to 3 mm from the tissue’s surface depending on the tissue’s scattering property. The noise floor is approximately -75 dBFS, and the sample signal is approximately -35 to -45 dBFS from the maximum full-scale signal. The average signal-to-noise ratio in oral tissues is therefore between 30 to 40dB. For visualization, the final volumetric OCTA data were also segmented into slabs with various thicknesses of interest, with a reference to the tissue surface [51]. Average/maximum intensity projection images were then generated from the segmented slabs by mapping the average/maximum value in each A-line. This collapsed 2D map is termed as an en face projection map.
2.3 Consideration of OCT resolution and field of view
In this section, we discuss the lens tube design that governs the limit of resolution and FoV in the intraoral probe. In LIAF-OCT, the OCTA function is most sensitive to the motion artifacts, which in turn is a function of resolution and field of view [52,53]. Typically, high-resolution and large FoV reduce the image quality of the OCTA. Since OCTA is the most challenging imaging mode compared to other functions, the repeatability of the LIAF-OCT can be judged by the quality of OCTA scans. The quality of OCTA scans can be observed directly from the sharpness of the vessels and the motion artifacts presented on the corresponding en face projection OCTA images [29,45]. We also quantified the vessel density index as measured by the different designs to aid the discussion.
The design of the intraoral tube started with the estimate of the required resolution. From previous literature, the average diameter of capillaries within gingival tissue is between 5 µm to 10 µm [54,55], and the mean separation of capillaries is 40 µm [56]. Therefore, a resolving power between 5 µm to 40 µm would be relevant for OCTA, where the lower limit favors accuracy and the upper limit favors repeatability. We employed two aspheric lenses with an equivalent focal length of f = 12 mm to f = 50 mm for the 2 mm-collimated scanning laser (collimator F220APC-1310). These focal lengths of f = 12 mm and f = 50 mm in our OCT system had an equivalent laser spot size of 10.0 µm and 41.7 µm, respectively.
We then designed the two-lens tube that accommodated the lenses and tested it with in-vivo subjects using disposable reflector housings. For the high resolution (HR) f = 12 mm probe, we repeated 6 OCTA scans, where 5 consecutive repeated volumes were performed in each OCTA scan. Here, ‘consecutive repeated volumes’ mean the patient does not move during scanning, and the probe is kept at the same position. For the general resolution (GR) probe f = 50 mm, we also repeated 6 OCTA scans, but each OCTA scan had 3 consecutive repeated volumes. We followed this protocol because the results from the HR probe presented significant challenges in OCTA due to its high sensitivity to patient motion. The HR probe resolved much smaller features inside the sample. As a result, small patient movement caused the appearance of bright stripes which could be incorrectly identified as blood vessels. The incorrect identification often leads to a lower blood vessel signal-to-noise ratio. In contrast, the results from the GR probe showed more resistance against interference from small vibrations and movement, resulting in better quality OCTA images. Therefore, the most accurate results were obtained by collecting 5 consecutive volume datasets using the HR probe and 3 consecutive volume datasets for the GR probe. We also modified the scanning protocols between the GR and HR design so that the pixel density within the airy-disc was approximately equal, i.e., 50 pixels/mm for the GR and 200 pixels/mm for the HR probe.
We then assessed the success rate, repeatability and relative accuracy of the blood vessel density from each probe in order to determine the optimal intraoral probe design. The success rate was calculated by the number of volume scans that had good OCTA quality divided by the total number of volume scans. The good OCTA quality volume was defined as: having no severe motion artifacts, and correctly targeting the ROI; examples of failures are shown with red crosses in Fig. 3. The blood vessel density was measured from the extracted z-projection en face image (Fig. 3) using our previous methodology [57]. The ROI was selected in the gingival margin, at an approximately similar region (Fig. 3) for both scenarios. The success rate from the GR probe was 77.8% (N = 18) and the HR probe was 40% (N = 30). The calculated vessel density index from the GR probe was 0.332 ± 0.009 and from the HR probe was 0.328 ± 0.011. The repeatability between the two probes can be compared based on the success rate, and the standard deviation of the measured vessel density. The comparison was based on the calculated values of the vessel density, with the assumption that the scanning probe with lower resolution (larger spot size) would cause blood vessels to appear larger than in reality, and thus would result in a higher measured density. We concluded that the GR probe was more suitable in practice since the GR probe had a much better success rate (77.8% vs. 40% with HR) and repeatability (error of 2.7% vs 3.4% with HR); however, the GR probe suffered from an overestimation in the measurements when compared to HR probe because of a relatively larger spot size (VDI 0.332 vs. 0.328). The overestimation is assumed, from an estimated point spread function and the fact that the average size of capillaries is smaller than both resolutions from the two probes.
Fig. 3. Comparison between GR f = 50 mm lens and HR f = 12 mm lens probes. GR: general resolution. HR: high resolution. Using the HR probe, the blood vessel diameter is smaller and presumably closer to their physical size; however, these images are severely defocused and/or contain motion artifacts, potentially produce erroneous measurement (indicated by the red cross). Using the GR probe, the blood vessels appear larger due to a larger spot size, however, OCTA is less sensitive to motion and defocusing. The tabulated measurements show that the repeatability using the GR probe is higher than for the HR probe, whilst vessel density in HR is slightly lower than GR as expected.
Another important aspect of the intraoral design is the maximum FoV. With OCTA measurements, the choice is usually a compromise between the FoV and resolution, as illustrated in the above GR and HR design. With the development of an intraoral probe, the high moisture environment requires the probe to have a smaller window to adequately protect the optics inside and to delay the onset of fogging. This issue further limits the FoV. We found that a maximum FoV of 12 mm x 12 mm was an acceptable choice for intra-oral application with open window design. With this 12 mm x 12 mm FoV, the time for water condensation to build up on the reflecting mirror was approximately 2 minutes, an acceptable timeframe for targeting the site and performing the scan. A thin anti-fog coating was also added to delay the onset of fogging. In this paper, a GR probe with a 6.9 × 7.8 mm2 FoV, was used as determined by the compromise between FoV, resolution and pixel density.
2.4 Calibration of the GR probe
Once the design was determined, we calibrated the LIAF-OCT system using a USAF 1951 x1 target (Edmunds Optics, NJ, USA). The calibration would confirm the resolution limit. With a beam diameter of D (Ø = 2 mm) (collimated using F220APC-1310, Thorlabs, NJ, USA) at the objective lens of a focal length f (= 50 mm), the theoretical resolution at focus is 41.7 um at the wavelength of $\lambda $ (=1310 nm), according to the Gaussian-beam estimation:
where ${w_0}$ is the focused spot diameter. Additionally, the beam quality parameter ${M^2}$ is assumed to be close to 1, but in fact may deviate from an ideal Gaussian beam in practice.The GR probe resolved up to group-4 element-3 when viewed with the en face projection images (Fig. 4). There are 22.63-line pairs in group 4 element 3 (highlighted in Fig. 4); therefore, the GR probe can resolve up to 44.2µm, which is very close to the theoretical resolution of 41.7 µm. For this reason, we had minimized the scanning step (or pixel density) such that the pixel width is less than ½ of the theoretical resolution, i.e., 16 µm pixel separation in our study.
Fig. 4. USAF 1951 resolution target x1, as observed under en face projection of 3D-OCT, showing resolution limit at group-4 element-3 (highlighted by the orange square). The resolution target shows that the experimental resolution, 44.2 µm, agrees with the theoretical resolution limit of the GR f50mm lens, which is 41.7um.
2.5 Volunteer, imaging and scanning protocols
Two volunteers were recruited, consented and scanned. The use of OCT laboratory equipment on human subjects was approved by the Institutional Review Board (IRB) of the University of Washington. Cleaning and disinfection were performed before each scan to reduce the risk of infection. An intraoral cap was attached to the LIAF-OCT scanning probe as an additional sanitization measure. The probe was disinfected by the operator using 70% isopropyl alcohol wipes, and the intraoral cap was designed to be disposable after each scan.
To demonstrate the capabilities of our intraoral system, we acquired LIAF and OCT/OCTA images on the occlusal and palatal surfaces of the volunteers. Due to the anatomical symmetry of the left and right quadrants in most patients, we only scanned the quadrant that had more visible tooth stains and evidence of previous caries experience, i.e., a restoration. Preliminary inspection also showed that the buccal sides are healthy and had no obvious condition. Therefore, we only acquired palatal (from #2 to #7 UNS) and occlusal areas (approximal #12 and #13 UNS) which had many conditions that could be detected using the OCT and LIAF system. We chose to omit the results from buccal scan since there were no interesting features that could have been contrasted by either OCT or LIAF, in addition to the palatal demonstration.
Photo examples of occlusal and palatal scans are shown in Fig. 1(c). LIAF images were acquired followed by OCT and functional OCTA measurements for each tooth and marginal gingiva. Image collection was repeated at least three times on each tooth with a small adjustment to ensure that all areas were in good focus. This was necessary since the palatal surfaces of the teeth were mostly curved which could make the imaging outside the range of the OCT depth-of-focus.
The scanning protocol consisted of raster scans for OCT with an array of 1088×500×500 pixels, and raster 4-repeated B-scans for OCTA with an array of 1088×500×4×500 pixels. Both scanning protocols had a calibrated volume of 8×6.9×7.8 mm3. For each 3D scan, the acquisition time was about 6.5 seconds. The subject sitting time was approximately 20 to 30 minutes for imaging each quadrant, which included the preparation and probe positioning time.
3. Results and discussion
3.1 Evaluation of dental filling and interproximal caries lesion
To demonstrate that our system is capable of imaging an interdental composite restoration, we intentionally imaged the filling area with the aid of the live-view, on-screen LIAF imaging. This restoration was a filling from a previous interdental caries. Figure 5 are the images from a dental filling (between tooth #12 and #13 UNS, or #23 and #24 FDI) observed under OCT/LIAF.
Fig. 5. Original structural images of the top-left second premolar (tooth #12) and first molar (tooth #13). (a) 3D rendering of OCT volumetric data with dental filling (yellow arrows) and suspected inter-proximal carious lesion (red arrows); since the inter-proximal carious lesion was below the enamel surface, side view (inset) better illustrates the strong intensity from the carious lesion. (b) en face structural image obtained from average intensity z-projection where the filling appears more homogeneous than the surrounding tissue structures. The yellow patch highlights the suspected inter-proximal carious lesion. (c) raw LIAF image where the filling appears more whitish color (yellow arrows). (d, e and f) representative cross-sectional B-frames of OCT volumetric structural image corresponding to the dashed lines in (b). Yellow arrows: dental filling. Red arrows: interproximal caries lesion. Green arrows: enamel/gingival sulcus boundary. Blue arrows: bubble. LIAF: Light-induced autofluorescence.
Figure 5(a) is a 3D-rendered OCT volume image with the dental filling showing higher intensity (regions marked by yellow arrows) as compared to the neighboring tooth enamel. Figure 5(b) is an average-intensity z-projection of the volumetric OCT to show the differences in total OCT intensity of the tooth and dental filling. The enamel has relatively low scattering in the near -infrared spectrum, allowing a much deeper OCT penetration depth. As a result, the enamel appears bright in the z-projection average-intensity image (Fig. 5(b)). In contrast, the dental filling has relatively high scattering properties, resulting in a shallower imaging depth and an overall lower z-projection average intensity. Figure 5(c) is the raw LIAF image showing the filling region with a slightly different color (more whitish) than the nearby teeth. However, the boundary between the teeth and composite restoration is superficial and cannot reveal depth information as shown in OCT B-scans (e.g. Figure 5(d)–5(f)). The filling appeared to have a defect that was successfully captured by the OCT (Fig. 5(f)), where a bubble is highlighted by the blue arrow in the composite restoration area. The bubble had an ovoid shape with a major axis and a minor axis of 337 µm and 201 µm. This defect would have been impossible to detect by the LIAF alone, and most likely so by visual examination or conventional X-ray, demonstrating the usefulness of the high-resolution volumetric OCT scan. In the best-case scenario, our OCT setup can detect a defect of up to 44 µm lateral resolution, and 7 µm axially, at a depth of up to 2 mm.
Figure 5(e) highlights a bright area with a clear border, marked by red arrows, that could be indicative of an incipient enamel caries lesion. This area is adjacent to a previously restored caries lesion and thus, over time, could have progressed to a secondary lesion. Detection of incipient caries lesions is very difficult and often requires subtraction radiography to quantitatively measure progression. Figure 5(d) shows the neighboring B-scan of healthy enamel. There is a very consistent drop in OCT intensity up to the enamel/gingival sulcus boundary indicated by green arrows. In contrast, the suspected caries area (Fig. 5(e)) had a stronger attenuation of light (which limited the penetration depth) and also brighter OCT intensity as indicated by red arrows. The possibility of an enamel crack was ruled out since enamel cracks are accompanied by either a shadow or a transparent enamel area depending on the severity of the crack [40], but not a strong gradual attenuation as presented in this case. We concluded that the site was likely a demineralized enamel, probably secondary caries lesion. Without an extraction, we could not reach a definite answer but simply refer to this demineralized enamel as caries lesion for simplicity. There is a possibility that this region has not progressed into a clinical caries and simply an initial break-down of the enamel structure.
Enamel caries lesions can be managed using non-surgical treatment strategies; and there are several existing fluorescence-based clinical tools recommended to help support visual-tactile diagnoses [3,8]. More accurate tools to directly detect early enamel lesions are needed. Several studies have demonstrated the potential of OCT to image enamel caries lesions and reported that the OCT signal in the demineralized area is higher vs sound enamel [39,58,59], consistent with our study findings. The increased OCT signal in demineralized regions may be due to the increased porosity caused by the mineral loss in the lesion [54]. The increased interfaces between crystals and water within dental pores may lead to higher reflectivity, thus resulting in a brighter appearance in the OCT image in the caries lesions [60]. Incipient interproximal carious lesions are easily missed in the clinic, using radiography technique [3]. However, our intraoral SS-OCT has shown the capability to detect interproximal caries through scanning the occlusal surface thanks to the high translucency of enamel to near-infrared light [38,61], particularly at 1310 nm. There was a good contrast between healthy and demineralized enamel under OCT, where conventional radiography failed to pick up during a regular dental check-up per volunteer’s report. Furthermore, our system provides depth-resolved 3D OCT data which is helpful to accurately localize the lesion location (Fig. 5(a) and 5(b)).
3.2 Detection of dental plaque (including soft plaque and calculus) and stain
Detecting and removing mature dental plaque is of critical importance in preventing caries and periodontal diseases [62]. Therefore, we further tested our system’s ability to image dental plaque. Figure 6 captures images of the palatal side of the second molar (tooth #2 UNS, or #17 ISO) which is covered with plaque biofilm, calculus (also known as tartar or mineralized plaque), and stain, which were confirmed by an orthodontist.
Fig. 6. Images of the top-right second molar showing dental plaque and stain. (a) The raw LIAF image. (b) En face maximum projected vascular image. Color bar represents the vessel depth. (c) En face maximum projected structural image. (d) Representative cross-sectional B-frame of OCT vasculature corresponding to the dashed line in (c). (e) Representative cross-sectional B-frame of OCT structure corresponding to the dashed line in (b). Blue arrow: a mixed plaque including soft plaque and calculus. Red arrow: calculus.
Figure 6(a) is the raw LIAF image showing variable colorization of this tooth from white to red to dark. Figure 6(b) is an en face maximum intensity projection of the vascular image from a slab with a thickness of ∼700 µm (measured from the surface) showing high dynamic intensity along the cervical margins highlighted by the blue arrow where dental plaque is often deposited. Figure 6(c) is an en face maximum intensity projection of the structural image from the same slab showing high intensity along the cervical margins and the middle area corresponding to the red coloration in the raw LIAF image. Figure 6(d) and 6(e) are the representative B-frames of structure and blood flow, respectively, corresponding to the locations marked by dashed lines in Fig. 6(b).
The dark areas in Fig. 6(a) are believed to be extrinsic tooth stains. The red coloration (i.e., higher red fluorescence) along the cervical margin and the middle area highlighted by the arrows in Fig. 6(a) is thought to be dental plaque or calculus. Mature dental plaque and calculus contain bacterial metabolites, namely porphyrin, that emit red fluorescence [47]. The red-outlined region highlights the area with high intensity in the OCT structural image, Fig. 6(c) and 6(d). This region is more indicative of dental calculus given the hyper-reflection properties of the OCT light [63]. Additionally, this region also elicits strong OCTA signals (vascular map in Fig. 6(b) and 6(e)), suggesting this might be soft plaque. This is because dental soft plaque is a biofilm of microorganisms (mostly bacteria, but also fungi) forming on the surfaces within the mouth [61], and self-mobility of bacteria and fungi in the soft plaque would lead to the high dynamic intensity in the blood flow map. In addition to high intensity in the blood flow map, we also observed that this region had high intensity in the structural image (Fig. 6(d)), hence, it is likely a mixed plaque with both soft plaque and calculus. To conclude, our LIAF-OCT system demonstrates potential for the detection dental soft plaque, calculus, and stains.
3.3 LIAF-OCT images from palatal side
In this section, the teeth and mucosae of the top-right palatal quadrant were imaged to demonstrate the capability of our LIAF-OCT system. The sites were chosen based on preliminary screening and selection, where only areas with existing oral conditions were scanned. Previous studies have been successful in demonstrating the intraoral OCT ability of imaging teeth and gingiva [32,40,42,45]. However, none have demonstrated the 3D lingual or palatal view of posterior teeth. To the best of our knowledge, this is the first intraoral OCT to show structural and vascular images of the posterior, palatal surfaces.
A total of 6 in vivo human teeth from the top right lateral incisor to the second molar (tooth #7 to #2 UNS, or #12 to #17 ISO) were chosen for the scan. For each tooth, three different positions: medial, zenith and distal were acquired. Raw LIAF images were stitched (Fig. 7(a)) to show a photomontage of the top right palatal surfaces of the aforementioned teeth. These teeth had heavy stains and appeared black in the photo. Several teeth also had a slight pink color on the surfaces when observed under LIAF, indicating the presence of plaque.
Fig. 7. The teeth and gingiva of the top-right palatal quadrant is imaged by the described imaging probe. Shown are the montaged images of (a) raw LIAF, (b) OCT en face structures, and (c) OCTA en face angiography of the teeth and gingiva in the top-right palatal quadrant. All en face images are projected by maximum intensity. Red arrow: gingival scar. Black arrow: enamel crack.
The OCT and OCTA were acquired concurrently with the guidance of LIAF. The presented images from OCT and OCTA (Fig. 7(b) and 7(c)) are en face maximum intensity projections of a slab with a thickness of ∼140 µm (measured from 280 to 420 µm below the surface). The OCT structural images provided several useful information capturing enamel cracks (black arrows at lateral incisor) and a gingival abnormality (red arrow at the first molar). We suspected this gingival abnormality was a scar since it has significantly less capillaries in the corresponding OCTA, and a white blemish in the corresponding LIAF [64]. The characteristics of abnormal tissues in OCT/OCTA images have been described in detail above (sections 3.1 and 3.2).
The results demonstrated the capability of the intraoral probe design to image the posterior and anterior oral tissues at a capillary scale. The photomontage of LIAF, OCT and OCTA provides structural and functional information of an area of interest across multiple surfaces. For example, the montage of OCTA (Fig. 7(c)) shows that blood vessels on the first molar are lower in vessel density than what’s observed for the surrounding teeth. Through a montage scan, the prospect of whole-mouth scan is implied, since buccal and occlusal in different quadrant can be achieved by rotation of the reflector cap. There were concerns with different level of motion artifacts in different quadrant, but we did not observe any differences. We recognize that this hypothesis, however, is yet to be proven. The prospective advantages of the intra-oral probe are in the potential of vessel quantification, objective measurements, longitudinal monitoring, automated assessment and efficiencies in diagnoses.
The current total time for acquiring the montage scan was 20 to 30 minutes. The variation in the total duration mostly depends on the operator’s training. The duration for imaging the whole quadrant is not yet practical in a dental clinic, and may impede its adoption as routine imaging method for dental care. The difficulties are often related to the orientation of the image guidance with respect to the clinician. For example, buccal and lingual/palatal direction may be reversed when the probe is positioned from left to right quadrant, as imaged and displayed by the LIAF. The post-processing is currently done offline after the imaging and may also delay the diagnosis in practice. One OCTA scan takes approximately 3-4 minutes during post-processing and may take even further for any additional objective measurements.
4. Conclusion
This paper demonstrates the feasibility of using the intraoral LIAF-OCT system for imaging posterior and anterior teeth as well as for detecting dental disease in vivo. The design of our intraoral probe addressed the limitations from previous studies and increased the capacity of OCT to measure all oral surfaces regardless of location, paving the way for whole mouth clinical application. We demonstrated at least one quadrant inside the mouth can be acquired within 30 minutes, and other scanning surfaces and locations are implied to be similar. The solutions to each prior limitation, such as limited FoV, probe accessibility, lack of real-time guidance and OCTA, were discussed and demonstrated. The LIAF provided real-time guidance to the operator for OCT imaging, allowing for improved accessibility and rapid positioning of regions of interest. The SS-OCT provided high-resolution structural and OCTA information associated with dental conditions, such as dental caries and the presence of dental plaque. The handheld probe along with the intraoral mouthpiece was designed for a relatively large FoV (i.e, 6.9 × 7.8 mm in this study) and the ease of use (i.e. high operating success rate by using the GR lens) of the system for the dental assessment. The resolution of the GR probe was shown to resolve individual capillaries, albeit having less resolving power. Future work should focus on identifying more robust de-fogging solutions such as employing an active-heating element behind the mirror or direct air spray. Increasing the FoV would also bring significant practical implications on how effective clinicians can target and diagnose the region of interest. Nevertheless, the current design of the intraoral LIAF-OCT system showed the potential to be a useful diagnostic and monitoring tool for the oral conditions.
Funding
Colgate-Palmolive Company; Carl Zeiss Meditec AG; Facebook.
Acknowledgement
This study was supported in part by the Colgate-Palmolive Company, Washington Research Foundation, and WRF David and Nancy Auth Innovation Award. Generous support from the Department of Bioengineering, University of Washington, is also acknowledged. The funding organizations had no role in the design or conduct of this research. Ethics statement. The imaging of subjects reported in this study using laboratory-built investigational device was conducted in accordance with a protocol approved by the Institutional Review Board of the University of Washington and informed consent was obtained from all subjects. The study followed the tenets of the Declaration of Helsinki and was conducted in compliance with the Health Insurance Portability and Accountability Act.
Disclosures
Dr. Wang discloses intellectual property owned by the Oregon Health and Science University and the University of Washington. Dr. Wang also receives research support from Colgate Palmolive Company, Carle Zeiss Meditec Inc, and Facebook Technologies LLC. Dr. Kilpatrick-Liverman and Dr. Subhash are employees of Colgate Palmolive Company. All other authors have no disclosures.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. E. Bernabe, W. Marcenes, C. R. Hernandez, J. Bailey, L. G. Abreu, V. Alipour, S. Amini, J. Arabloo, Z. Arefi, A. Arora, M. A. Ayanore, T. W. Bärnighausen, A. Bijani, D. Y. Cho, D. T. Chu, C. S. Crowe, G. T. Demoz, D. G. Demsie, Z. S. Dibaji Forooshani, M. Du, M. El Tantawi, F. Fischer, M. O. Folayan, N. D. Futran, Y. C. D. Geramo, A. Haj-Mirzaian, N. Hariyani, A. Hasanzadeh, S. Hassanipour, S. I. Hay, M. K. Hole, S. Hostiuc, M. D. Ilic, S. L. James, R. Kalhor, L. Kemmer, M. Keramati, Y. S. Khader, S. Kisa, A. Kisa, A. Koyanagi, R. Lalloo, Q. Le Nguyen, S. D. London, N. D. Manohar, B. B. Massenburg, M. R. Mathur, H. G. Meles, T. Mestrovic, A. Mohammadian-Hafshejani, R. Mohammadpourhodki, A. H. Mokdad, S. D. Morrison, J. Nazari, T. H. Nguyen, C. T. Nguyen, M. R. Nixon, T. O. Olagunju, K. Pakshir, M. Pathak, N. Rabiee, A. Rafiei, K. Ramezanzadeh, M. J. Rios-Blancas, E. M. Roro, S. Sabour, A. M. Samy, M. Sawhney, F. Schwendicke, F. Shaahmadi, M. A. Shaikh, C. Stein, M. R. Tovani-Palone, B. X. Tran, B. Unnikrishnan, G. T. Vu, A. Vukovic, T. S. S. Warouw, Z. Zaidi, Z. J. Zhang, and N. J. Kassebaum, “Global, regional, and national levels and trends in burden of oral conditions from 1990 to 2017: a systematic analysis for the global burden of disease 2017 study,” J. Dent. Res. 99(8), 898–906 (2020). [CrossRef]
2. L. B. Oliveira, C. Massignan, A. C. Oenning, K. Rovaris, M. Bolan, A. L. Porporatti, and G. De Luca Canto, “Validity of micro-CT for in vitro caries detection: a systematic review and meta-analysis,” Dentomaxillofacial Radiology 49(7), 20190347 (2020). [CrossRef]
3. T. Walsh, R. Macey, P. Riley, A. M. Glenny, F. Schwendicke, H. V. Worthington, J. E. Clarkson, D. Ricketts, T. L. Su, and A. Sengupta, “Imaging modalities to inform the detection and diagnosis of early caries,” Cochrane Database Syst. Rev. 2021(12), CD014545 (2021). [CrossRef]
4. P. Bastos, G. Carpentier, V. Patel, D. Papy-Garcia, T. Watson, and R. Cook, “Real-time optical vascular imaging, a new method for the diagnosis and monitoring of oral diseases,” Journal of Microscopy n/a (2020).
5. J. A. Lindeboom, K. R. Mathura, S. Harkisoen, H. P. van den Akker, and C. Ince, “Effect of smoking on the gingival capillary density: assessment of gingival capillary density with orthogonal polarization spectral imaging,” J. Clin. Periodontol. 32(12), 1208–1212 (2005). [CrossRef]
6. A. L. JanuÁRio, M. Barriviera, and W. R. Duarte, “Soft tissue cone-beam computed tomography: a novel method for the measurement of gingival tissue and the dimensions of the dentogingival unit,” J. Esthetic and Restorative Dentistry 20(6), 366–373 (2008). [CrossRef]
7. R. Macey, T. Walsh, P. Riley, A. M. Glenny, H. V. Worthington, P. A. Fee, J. E. Clarkson, and D. Ricketts, “Fluorescence devices for the detection of dental caries,” Cochrane Database of Systematic Reviews (2020).
8. T. Gimenez, M. M. Braga, D. P. Raggio, C. Deery, D. N. Ricketts, and F. M. Mendes, “Fluorescence-based methods for detecting caries lesions: systematic review, meta-analysis and sources of heterogeneity,” PLoS One 8(4), e60421 (2013). [CrossRef]
9. L. Karlsson, “Caries detection methods based on changes in optical properties between healthy and carious tissue,” Int. J. Dentistry 2010, 1–9 (2010). [CrossRef]
10. A. D. Bounds and J. M. Girkin, “Early stage dental caries detection using near infrared spatial frequency domain imaging,” Sci. Rep. 11(1), 2433 (2021). [CrossRef]
11. D. Huang, E. A. Swanson, C. P. Lin, J. S. Schuman, W. G. Stinson, W. Chang, M. R. Hee, T. Flotte, K. Gregory, C. A. Puliafito, and J. G. Fujimoto, “Optical Coherence tomography,” Science 254(5035), 1178–1181 (1991). [CrossRef]
12. P. H. Tomlins and R. K. Wang, “Theory, developments and applications of optical coherence tomography,” J. Phys. D: Appl. Phys. 38(15), 2519–2535 (2005). [CrossRef]
13. A. H. Kashani, C. L. Chen, J. K. Gahm, F. Zheng, G. M. Richter, P. J. Rosenfeld, Y. Shi, and R. K. Wang, “Optical coherence tomography angiography: A comprehensive review of current methods and clinical applications,” Prog. Retin. Eye Res. 60, 66–100 (2017). [CrossRef]
14. M. Adhi and J. S. Duker, “Optical coherence tomography–current and future applications,” Curr. Opin. Ophthalmol. 24(3), 213–221 (2013). [CrossRef]
15. Z. Chu, L. Wang, X. Zhou, Y. Shi, Y. Cheng, R. Laiginhas, H. Zhou, M. Shen, Q. Zhang, L. de Sisternes, A. Y. Lee, G. Gregori, P. J. Rosenfeld, and R. K. Wang, “Automatic geographic atrophy segmentation using optical attenuation in OCT scans with deep learning,” Biomed. Opt. Express 13(3), 1328–1343 (2022). [CrossRef]
16. A. R. Motschi, P. K. Roberts, S. Desissaire, M. Schranz, F. Schwarzhans, H. Bogunović, M. Pircher, and C. K. Hitzenberger, “Identification and quantification of fibrotic areas in the human retina using polarization-sensitive OCT,” Biomed. Opt. Express 12(7), 4380–4400 (2021). [CrossRef]
17. M. J. Gora, M. J. Suter, G. J. Tearney, and X. Li, “Endoscopic optical coherence tomography: technologies and clinical applications [Invited],” Biomed. Opt. Express 8(5), 2405–2444 (2017). [CrossRef]
18. M. J. Marques, M. R. Hughes, A. F. Uceda, G. Gelikonov, A. Bradu, and A. Podoleanu, “Endoscopic en-face optical coherence tomography and fluorescence imaging using correlation-based probe tracking,” Biomed. Opt. Express 13(2), 761–776 (2022). [CrossRef]
19. D. C. Adams, A. Majid, and M. J. Suter, “Polarization mode dispersion correction in endoscopic polarization-sensitive optical coherence tomography with incoherent polarization input states,” Biomed. Opt. Express 13(6), 3446–3460 (2022). [CrossRef]
20. J. Welzel, “Optical coherence tomography in dermatology: a review,” Skin Res Technol 7(1), 1–9 (2001). [CrossRef]
21. A. J. Deegan, F. Talebi-Liasi, S. Song, Y. Li, J. Xu, S. Men, M. M. Shinohara, M. E. Flowers, S. J. Lee, and R. K. Wang, “Optical coherence tomography angiography of normal skin and inflammatory dermatologic conditions,” Lasers Surg. Med. 50(3), 183–193 (2018). [CrossRef]
22. A. J. Deegan, W. Wang, S. Men, Y. Li, S. Song, J. Xu, and R. K. Wang, “Optical coherence tomography angiography monitors human cutaneous wound healing over time,” Quant Imaging Med Surg 8(2), 135–150 (2018). [CrossRef]
23. A. J. Deegan, J. Lu, R. Sharma, S. P. Mandell, and R. K. Wang, “Imaging human skin autograft integration with optical coherence tomography,” Quant Imaging Med Surg 11(2), 784–796 (2021). [CrossRef]
24. J. Lu, A. J. Deegan, Y. Cheng, S. P. Mandell, and R. K. Wang, “OCT-based angiography and surface topography in burn-damaged skin,” Lasers Surg. Med. 53(6), 849–860 (2021). [CrossRef]
25. W. C. Lin, R. A. Byers, W. Li, S. G. Danby, M. J. Cork, and S. J. Matcher, “Imaging striae distensae: a comparison between PS-OCT and digital dermoscopy,” Biomed. Opt. Express 12(6), 3296–3311 (2021). [CrossRef]
26. X. Zhou, S. Maloufi, D. C. Louie, N. Zhang, Q. Liu, T. K. Lee, and S. Tang, “Investigating the depolarization property of skin tissue by degree of polarization uniformity contrast using polarization-sensitive optical coherence tomography,” Biomed. Opt. Express 12(8), 5073–5088 (2021). [CrossRef]
27. L. Waszczuk, J. Ogien, J.-L. Perrot, and A. Dubois, “Co-localized line-field confocal optical coherence tomography and confocal Raman microspectroscopy for three-dimensional high-resolution morphological and molecular characterization of skin tissues ex vivo,” Biomed. Opt. Express 13(4), 2467–2487 (2022). [CrossRef]
28. M. Machoy, J. Seeliger, L. Szyszka-Sommerfeld, R. Koprowski, T. Gedrange, and K. Woźniak, “The use of optical coherence tomography in dental diagnostics: a state-of-the-art review,” J. Healthcare Eng. 2017, 1–31 (2017). [CrossRef]
29. N. M. Le, S. Song, H. Zhou, J. Xu, Y. Li, C.-E. Sung, A. Sadr, K.-H. Chung, H. M. Subhash, L. Kilpatrick, and R. K. Wang, “A noninvasive imaging and measurement using optical coherence tomography angiography for the assessment of gingiva: An in vivo study,” J. Biophotonics 11(12), e201800242 (2018). [CrossRef]
30. N. Le, H. M. Subhash, L. Kilpatrick-Liverman, and R. K. Wang, “Noninvasive multimodal imaging by integrating optical coherence tomography with autofluorescence imaging for dental applications,” J Biophotonics 13(7), e202000026 (2020). [CrossRef]
31. G. Wang, N. M. Le, X. Hu, Y. Cheng, S. L. Jacques, H. Subhash, and R. K. Wang, “Semi-automated registration and segmentation for gingival tissue volume measurement on 3D OCT images,” Biomed. Opt. Express 11(8), 4536–4547 (2020). [CrossRef]
32. L. O. Fernandes, C. C. B. O. Mota, L. S. A. de Melo, M. U. S. da Costa Soares, D. da Silva Feitosa, and A. S. L. Gomes, “In vivo assessment of periodontal structures and measurement of gingival sulcus with Optical Coherence Tomography: a pilot study,” J. Biophotonics 10(6-7), 862–869 (2017). [CrossRef]
33. J. Walther, Q. Li, M. Villiger, C. S. Farah, E. Koch, K. Karnowski, and D. D. Sampson, “Depth-resolved birefringence imaging of collagen fiber organization in the human oral mucosa in vivo,” Biomed. Opt. Express 10(4), 1942–1956 (2019). [CrossRef]
34. P.-H. Chen, Y.-J. Chen, Y.-F. Chen, Y.-C. Yeh, K.-W. Chang, M.-C. Hou, and W.-C. Kuo, “Quantification of structural and microvascular changes for diagnosing early-stage oral cancer,” Biomed. Opt. Express 11(3), 1244–1256 (2020). [CrossRef]
35. E. Baradaran Shokouhi, R. Welch, K. Sivagurunathan, and A. Mandelis, “Multispectral truncated-correlation photothermal coherence tomography imaging modality for detection of early stage dental caries,” Biomed. Opt. Express 13(5), 2772–2781 (2022). [CrossRef]
36. M.-T. Tsai, Y. Chen, C.-Y. Lee, B.-H. Huang, N. H. Trung, Y.-J. Lee, and Y.-L. Wang, “Noninvasive structural and microvascular anatomy of oral mucosae using handheld optical coherence tomography,” Biomed. Opt. Express 8(11), 5001–5012 (2017). [CrossRef]
37. W. Wei, W. J. Choi, and R. K. Wang, “Microvascular imaging and monitoring of human oral cavity lesions in vivo by swept-source OCT-based angiography,” Lasers Med Sci 33(1), 123–134 (2018). [CrossRef]
38. Y. Shimada, H. Nakagawa, A. Sadr, I. Wada, M. Nakajima, T. Nikaido, M. Otsuki, J. Tagami, and Y. Sumi, “Noninvasive cross-sectional imaging of proximal caries using swept-source optical coherence tomography (SS-OCT) in vivo,” J Biophotonics 7(7), 506–513 (2014). [CrossRef]
39. H. Schneider, M. Ahrens, M. Strumpski, C. Rüger, M. Häfer, G. Hüttmann, D. Theisen-Kunde, H. Schulz-Hildebrandt, and R. Haak, “An Intraoral OCT Probe to Enhanced Detection of Approximal Carious Lesions and Assessment of Restorations,” J. Clin. Med. 9(10), 3257 (2020). [CrossRef]
40. J. Won, P. C. Huang, D. R. Spillman, E. J. Chaney, R. Adam, M. Klukowska, R. Barkalifa, and S. A. Boppart, “Handheld optical coherence tomography for clinical assessment of dental plaque and gingiva,” J. Biomed. Opt. 25(11), 116011 (2020). [CrossRef]
41. J. Walther, C. Schnabel, F. Tetschke, T. Rosenauer, J. Golde, N. Ebert, M. Baumann, C. Hannig, and E. Koch, “In vivo imaging in the oral cavity by endoscopic optical coherence tomography,” J. Biomed. Opt. 23(7), 1–13 (2018). [CrossRef]
42. V.-F. Duma, D. Demian, C. Sinescu, R. Cernat, G. Dobre, M. L. Negrutiu, F. Topala, G. Hutiu, A. Bradu, and A. Podoleanu, “Handheld scanning probes for optical coherence tomography: developments, applications, and perspectives,” Sixth International Conference on Lasers in Medicine (SPIE, 2016), Vol. 9670.
43. K. Li, Z. Yang, W. Liang, J. Shang, Y. Liang, and S. Wan, “Low-cost, ultracompact handheld optical coherence tomography probe for in vivo oral maxillofacial tissue imaging,” J. Biomed. Opt. 25(4), 1–13 (2020). [CrossRef]
44. A. Y. Alibhai, C. Or, and A. Witkin, “Swept source optical coherence tomography: a review,” Curr. Ophthalmol. Rep. 6(1), 7–16 (2018). [CrossRef]
45. W. J. Choi and R. K. Wang, “In vivo imaging of functional microvasculature within tissue beds of oral and nasal cavities by swept-source optical coherence tomography with a forward/side-viewing probe,” Biomed. Opt. Express 5(8), 2620–2634 (2014). [CrossRef]
46. E. de Josselin de Jong, S. M. Higham, P. W. Smith, C. J. van Daelen, and M. H. van der Veen, “Quantified light-induced fluorescence, review of a diagnostic tool in prevention of oral disease,” J. Appl. Phys. 105(10), 102031 (2009). [CrossRef]
47. E.-S. Lee, E. de Josselin de Jong, and B.-I. Kim, “Detection of dental plaque and its potential pathogenicity using quantitative light-induced fluorescence,” J. Biophotonics 12(7), e201800414 (2019). [CrossRef]
48. R. K. Wang, L. An, P. Francis, and D. J. Wilson, “Depth-resolved imaging of capillary networks in retina and choroid using ultrahigh sensitive optical microangiography,” Opt. Lett. 35(9), 1467–1469 (2010). [CrossRef]
49. M. Wojtkowski, V. J. Srinivasan, T. H. Ko, J. G. Fujimoto, A. Kowalczyk, and J. S. Duker, “Ultrahigh-resolution, high-speed, Fourier domain optical coherence tomography and methods for dispersion compensation,” Opt. Express 12(11), 2404–2422 (2004). [CrossRef]
50. N. Le, S. Song, Q. Zhang, and R. K. Wang, “Robust principal component analysis in optical micro-angiography,” Quant. Imaging Med. Surg. 7(6), 654–667 (2017). [CrossRef]
51. X. Yin, J. Chao, and R. Wang, “User-guided segmentation for volumetric retinal optical coherence tomography images,” J. Biomed. Opt. 19(8), 086020 (2014). [CrossRef]
52. E. C. Greig, J. S. Duker, and N. K. Waheed, “A practical guide to optical coherence tomography angiography interpretation,” Int. J. Retin. Vitr. 6(1), 55 (2020). [CrossRef]
53. T. E. de Carlo, A. Romano, N. K. Waheed, and J. S. Duker, “A review of optical coherence tomography angiography (OCTA),” Int. J. Retin. Vitr. 1(1), 5 (2015). [CrossRef]
54. S. Yoshida, K. Noguchi, K. Imura, Y. Miwa, M. Sunohara, and I. Sato, “A morphological study of the blood vessels associated with periodontal probing depth in human gingival tissue,” Okajimas Folia Anat Jpn 88(3), 103–109 (2011). [CrossRef]
55. H. Zoellner and N. Hunter, “Perivascular hyaline deposits in inflamed gingival tissues,” J. Oral Pathol. Med. 18(6), 333–338 (1989). [CrossRef]
56. R. J. A. Freitas, Nanomedicine, Volume I: Basic Capabilities (Landes Bioscience, 1999).
57. Z. Chu, J. Lin, C. Gao, C. Xin, Q. Zhang, C.-L. Chen, L. Roisman, G. Gregori, P. J. Rosenfeld, and R. K. Wang, “Quantitative assessment of the retinal microvasculature using optical coherence tomography angiography,” J. Biomed. Opt. 21(6), 066008 (2016). [CrossRef]
58. Y. Shimada, M. Yoshiyama, J. Tagami, and Y. Sumi, “Evaluation of dental caries, tooth crack, and age-related changes in tooth structure using optical coherence tomography,” Jpn Dent Sci Rev 56(1), 109–118 (2020). [CrossRef]
59. Y. Shimada, A. Sadr, Y. Sumi, and J. Tagami, “Application of optical coherence tomography (OCT) for diagnosis of caries, cracks, and defects of restorations,” Curr Oral Health Rep 2(2), 73–80 (2015). [CrossRef]
60. I. Hariri, A. Sadr, S. Nakashima, Y. Shimada, J. Tagami, and Y. Sumi, “Estimation of the enamel and dentin mineral content from the refractive index,” Caries Res 47(1), 18–26 (2013). [CrossRef]
61. N. Patara, L. D. Cynthia, F. Daniel, B. Jeff, and B. Steve, “PS-OCT of occlusal and interproximal caries lesions viewed from occlusal surfaces,” Proc.SPIE 6137, 61370L (2006). [CrossRef]
62. P. D. Marsh, “Dental plaque as a biofilm and a microbial community – implications for health and disease,” BMC Oral Health 6(S1), S14 (2006). [CrossRef]
63. M. Tsubokawa, A. Aoki, S. Kakizaki, Y. Taniguchi, K. Ejiri, K. Mizutani, G. Koshy, T. Akizuki, S. Oda, Y. Sumi, and Y. Izumi, “In vitro and clinical evaluation of optical coherence tomography for the detection of subgingival calculus and root cementum,” J Oral Sci 60(3), 418–427 (2018). [CrossRef]
64. J. Kleinheinz, A. Büchter, T. Fillies, and U. Joos, “Vascular basis of mucosal color,” Head Face Med. 1(1), 4 (2005). [CrossRef]
