Automatic detection of tooth cracks in optical coherence tomography images

Human Tooth Crack Image Analysis with Multiple Deep Learning Approaches

Tooth cracks, one of the most common dental diseases, can result in the tooth falling apart without prompt treatment; dentists also have difficulty locating cracks, even with X-ray imaging. Indocyanine green (ICG) assisted near-infrared fluorescence (NIRF) dental imaging technique can solve this problem due to the deep penetration of NIR light and the excellent fluorescence characteristics of ICG. This study extracted 593 human cracked tooth images and 601 non-cracked tooth images from NIR imaging videos. Multiple imaging analysis methods such as classification, object detection, and super-resolution were applied to the dataset for cracked image analysis. Our results showed that machine learning methods could help analyze tooth crack efficiently: the tooth images with cracks and without cracks could be well classified with the pre-trained residual network and squeezenet1_1 models, with a classification accuracy of 88.2% and 94.25%, respectively; the single shot multi-box detector (SSD) was able to recognize cracks, even if the input image was at a different size from the original cracked image; the super-resolution (SR) model, SR-generative adversarial network demonstrated enhanced resolution of crack images using high-resolution concrete crack images as the training dataset. Overall, deep learning model-assisted human crack analysis improves crack identification; the combination of our NIR dental imaging system and deep learning models has the potential to assist dentists in crack diagnosis.

Three Visual-Diagnostic Methods for the Detection of Enamel Cracks: An In Vitro Study

Tooth fractures are a common cause of tooth loss, frequently starting as enamel cracks. However, methods for the detection of enamel cracks are poorly investigated. The aim of the study was the validation of three clinical methods for the detection of enamel cracks: dental operating microscope (DOM), near-infrared transillumination (NIR), and fiber-optic transillumination (FOTI), with hard-tissue slices serving as controls. A total of 89 extracted teeth, set up as diagnostic models, were investigated, and the maximum crack depth was scored by two examiners. The actual crack depth was determined microscopically (25×) using horizontal sections. The accuracy of each method was analyzed using receiver operating characteristic (ROC) curves. Across all tooth surfaces, the area under the curve (AUC) amounted to 0.57 (DOM), 0.70 (FOTI), and 0.67 (NIR). For crack detection on vestibular/oral surfaces, the AUC was 0.61 (DOM), 0.78 (FOTI), and 0.74 (NIR); for proximal surfaces, it was 0.59 (DOM), 0.65 (FOTI), and 0.67 (NIR). However, the actual crack depth was underestimated with each method (p < 0.001). Under in vitro conditions, FOTI and NIR are suitable for detection of enamel cracks, especially on vestibular and oral tooth surfaces. However, an exact estimation of crack depth is not possible. Therefore, FOTI and NIR seem to be helpful for the clinical detection of enamel cracks.

Diagnosis of cracked tooth: Clinical status and research progress

Cracked tooth is a common dental hard tissue disease.The involvement of cracks directly affects the selection of treatment and restoration of the affected teeth.It is helpful to choose more appropriate treatment options and evaluate the prognosis of the affected tooth accurately to determine the actual involvement of the crack.However, it is often difficult to accurately and quantitatively assess the scope of cracks at present.So it is necessary to find a real method of early quantitative and non-destructive crack detection.This article reviews the current clinical detection methods and research progress of cracked tooth in order to provide a reference for finding a clinical detection method for cracked tooth.

Recent Advances in the Diagnosis of Enamel Cracks: A Narrative Review

Cracked teeth can pose a diagnostic dilemma for a clinician as they can mimic several other conditions. The constant physiological stress along with any pathological strain like trauma or iatrogenic causes can lead to the development of microcracks in the teeth. Constant exposure to immense stress can cause the progression of these often-undiagnosed tooth cracks to cause tooth fractures. This review aims to outline the etiology of tooth cracks, their classification, and recent advances in the diagnosis of enamel cracks. Diagnosing a cracked tooth can be an arduous task as symptoms differ according to the location and extension of the incomplete fracture. Early detection is critical because restorative treatment can prevent fracture propagation, microleakage, pulpal or periodontal tissue involvement, and catastrophic cusp failure. Older methods of crack detection are not sensitive or specific. They include clinical examination, visual inspection, exploratory excavation, and percussion test. The dye test used blue or gentian violet stains to highlight fracture lines. Modern methods include transillumination, optical coherence tomography Swept-Source Optical Coherence Tomography (SSOCT), near-infrared imaging, ultrasonic system, infrared thermography, and near-infrared laser. These methods appear to be more efficacious than traditional clinical dental imaging techniques in detecting longitudinal tooth cracks. Clinically distinguishing between the various types of cracks can be difficult with patient-reported signs and symptoms varying according to the location and extension of the incomplete fracture. Cracks are more common in restored teeth. Technological advances such as transillumination allow for early detection and enhanced prognosis.

A perspective on the diagnosis of cracked tooth: imaging modalities evolve to AI-based analysis

Despite numerous clinical trials and pre-clinical developments, the diagnosis of cracked tooth, especially in the early stages, remains a challenge. Cracked tooth syndrome is often accompanied by dramatic painful responses from occlusion and temperature stimulation, which has become one of the leading causes for tooth loss in adults. Current clinical diagnostical approaches for cracked tooth have been widely investigated based on X-rays, optical light, ultrasound wave, etc. Advances in artificial intelligence (AI) development have unlocked the possibility of detecting the crack in a more intellectual and automotive way. This may lead to the possibility of further enhancement of the diagnostic accuracy for cracked tooth disease. In this review, various medical imaging technologies for diagnosing cracked tooth are overviewed. In particular, the imaging modality, effect and the advantages of each diagnostic technique are discussed. What's more, AI-based crack detection and classification methods, especially the convolutional neural network (CNN)-based algorithms, including image classification (AlexNet), object detection (YOLO, Faster-RCNN), semantic segmentation (U-Net, Segnet) are comprehensively reviewed. Finally, the future perspectives and challenges in the diagnosis of the cracked tooth are lighted.

Review of Cracked Tooth Syndrome: Etiology, Diagnosis, Management, and Prevention

Cracked tooth syndrome refers to a series of symptoms caused by cracked teeth. This article reviews the current literature on cracked tooth syndrome from four aspects, etiology, diagnosis, management, and prevention, to provide readers integrated information about this. The article begins with an introduction to the odontiatrogenic factors and then covers the noniatrogenic factors that induce cracked tooth syndrome. While the former discusses inappropriate root canal therapy and improper restorative procedures, the latter covers the topics such as the developmental and functional status of cracked tooth syndrome. This is then followed by the description of common clinical diagnosis methods, the prospects of new technologies, and summaries of current clinical management methods, including immediate management and direct and indirect restoration. In the final section, preventive methods and their importance are proposed, with the aim of educating the common population.

A method of crack detection based on digital image correlation for simulated cracked tooth

Background

Early clinical cracked tooth can be a perplexing disorder to diagnose and manage. One of the key problems for the diagnosis of the cracked tooth is the detection of the location of the surface crack.

Methods

This paper proposes an image-based method for the detection of the micro-crack in the simulated cracked tooth. A homemade three-axis motion platform mounted with a telecentric lens was built as an image acquisition system to observe the surface of the simulated cracked tooth, which was under compression with a magnitude of the masticatory force. By using digital image correlation (DIC), the deformation map for the crown surface of the cracked tooth was calculated. Through image analysis, the micro-crack was quantitatively visualized and characterized.

Results

The skeleton of the crack path was successfully extracted from the image of the principal strain field, which was further verified by the image from micro-CT. Based on crack kinematics, the crack opening displacement was quantitatively calculated to be 2–10 µm under the normal mastication stress, which was in good agreement with the value reported in the literature.

Conclusions

The crack on the surface of the simulated cracked tooth could be detected based on the proposed DIC-based method. The proposed method may provide a new solution for the rapid clinical diagnosis of cracked teeth and the calculated crack information would be helpful for the subsequent clinical treatment of cracked teeth.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12903-021-01897-2.

Comparison of diagnosis of cracked tooth using contrast-enhanced CBCT and micro-CT

OBJECTIVES

To evaluate the diagnostic accuracy using sodium iodide (NaI) and dimethyl sulfoxide (DMSO) as contrast agent in cone beam computed tomography (CBCT) scanning, and compare this with micro-CT.

METHODS

18 teeth were cracked artificially by soaking them cyclically in liquid nitrogen and hot water. After pre-treatment with artificial saliva, the teeth were scanned in four modes: CBCT routine scanning without contrast agent (RS); CBCT with meglumine diatrizoate (MD) as contrast agent (ES1); CBCT with NaI + DMSO as contrast agent (ES2); and micro-CT (mCT). The number of crack lines was evaluated in all four modes. Depth of crack lines and number of cracks presented from the occlusal surface to the pulp cavity (Np) in ES2 and micro-CT images were evaluated.

RESULTS

There were 63 crack lines in all 18 teeth. 45 crack lines were visible on ES2 images as against four on the RS and ES1 images (p<0.05) and 37 on micro-CT images (p>0.05). Further, 34 crack lines could be observed on both ES2 and micro-CT images, and the average depth presented on ES2 images was 4.56 ± 0.88 mm and 3.89 ± 1.08 mm on micro-CT images (p<0.05). More crack lines could be detected from the occlusal surface to the pulp cavity on ES2 images than on micro-CT images (22 vs 11).

CONCLUSION

CBCT with NaI +DMSO as the contrast agent was equivalent to micro-CT for number of crack lines and better for depth of crack lines. NaI + DMSO could be a potential CBCT contrast agent to improve diagnostic accuracy for cracked tooth.

Evaluation Through the Optical Coherence Tomography Analysis of the Influence of Non-Alcoholic Fatty Liver Disease on the Gingival Inflammation in Periodontal Patients

PURPOSE

The purpose of this ex vivo study is to exhibit the inflammatory changes that occur within the gingival tissue by using optical coherence tomography (OCT) in periodontal patients with non-alcoholic fatty liver disease (NAFLD) and if NAFLD could influence the local periodontal inflammation.

PATIENTS AND METHODS

Gingival tissue samples obtained from patients were divided into three groups - P (periodontitis), NAFLD+P (NAFLD+periodontitis) and H (healthy) groups - and were scanned using an OCT light beam, in order to perform a qualitative and quantitative analysis of images. The value of average pixel density has been associated with the degree of inflammation.

RESULTS

The highest average pixel density was found in patients from the H group, while the lowest value of average pixel density was recorded in gingival tissue samples collected from patients with NAFLD+P. The image assessments from NAFLD+P group delivered lower values of average pixel density than those of P group, suggesting a possible influence of this disease on the inflammatory tissular changes produced by periodontal disease.

CONCLUSION

After comparing the OCT analysis results obtained for the three groups of patients, we can consider that NAFLD may be an aggravating factor for the inflammation of periodontal disease.

Detection and analysis of enamel cracks by ICG-NIR fluorescence dental imaging

Cracked teeth are the third most common cause of tooth loss, but there is no reliable imaging tool for the diagnosis of cracks. Here, we demonstrate the feasibility of indocyanine green near-infrared fluorescence (ICG-NIRF) dental imaging for the detection of enamel cracks and enamel-dentin cracks in vitro in the first (ICG-NIRF-I, 700-950 nm) and second (ICG-NIRF-II, 950-1700 nm) imaging windows with transmission excitation light, and compared ICG-NIRF with conventional NIR illumination-II (NIRi-II) and X-ray imaging. Dentin cracks were detected by CT scan, while most enamel cracks, undetectable under X-ray imaging, were clearly visible in NIR images. We found that ICG-NIRF-II detected cracks more effectively than NIRi-II, and that light orientation is an important factor for crack detection: an angled exposure obtained better image contrast of cracks than parallel exposure, as it created a shadow under the crack. Crack depth could be evaluated from the crack shadow in ICG-NIRF and NIRi-II images; from this shadow we could determine crack depth and discriminate enamel-dentin cracks from craze lines. Cracks could be observed clearly from ICG-NIRF images with 1-min ICG tooth immersion, although longer ICG immersion produced images with greater contrast. Overall, our data show that ICG-NIRF dental imaging is a useful tool for diagnosing cracked teeth at an early stage.

An Experimental Review of Optical Coherence Tomography Systems for Noninvasive Assessment of Hard Dental Tissues

Optical coherence tomography (OCT) is a noninvasive, high-resolution, cross-sectional imaging technique. To date, OCT has been demonstrated in several areas of dentistry, primarily using wavelengths around 1,300 nm, low numerical aperture (NA) imaging lenses, and detectors insensitive to the polarization of light. The objective of this study is to compare the performance of three commercially available OCT systems operating with alternative wavelengths, imaging lenses, and detectors for OCT imaging of dental enamel. Spectral-domain (SD) OCT systems with (i) 840 nm (Lumedica, OQ LabScope 1.0), (ii) 1,300 nm (Thorlabs, Tel320) center wavelengths, and (iii) a swept-source (SS) OCT system (Thorlabs OCS1300SS) centered at 1,325 nm with optional polarization-sensitive detection were used. Low NA (0.04) and high NA (0.15) imaging lenses were used with system (iii). Healthy in vivo and in vitrohuman enamel and eroded in vitro bovine enamel specimens were imaged. The Tel320 system achieved greater imaging depth than the OQ LabScope 1.0, on average imaging 2.6 times deeper into the tooth (n = 10). The low NA lens provided a larger field of view and depth of focus, while the high NA lens provided higher lateral resolution and greater contrast. Polarization-sensitive imaging eliminated birefringent banding artifacts that can appear in conventional OCT scans. In summary, this study illustrates the performance of three commercially available OCT systems, objective lenses, and imaging modes and how these can affect imaging depth, resolution, field of view, and contrast in enamel. Users investigating OCT for dental applications should consider these factors when selecting an OCT system for clinical or basic science studies.

Evaluation of Crack Formation and Propagation with Ultrasonic Root-End Preparation and Obturation Using a Digital Microscope and Optical Coherence Tomography

OBJECTIVE

This study is aimed at determining (1) the effect of root-end resection, ultrasonic root-end preparation, and root-end filling on the incidence of crack formation and propagation by using a digital microscope (DM) and optical coherence tomography (OCT) and (2) the performance of OCT on the detection of cracks by comparing with microcomputed tomography (micro-CT) as a reference standard.

METHODOLOGY

Thirty extracted lower incisors were endodontically treated and subjected to root-end resection and ultrasonic root-end cavity preparation. Then, the teeth were divided into three groups (n = 10, each), and the root-end cavity was either left unfilled or filled with mineral trioxide aggregate (MTA) or super-EBA. The resected surface was observed with OCT and DM after the root-end resection, ultrasonic root-end preparation, and root-end filling, and the frequency of incomplete and complete cracks were recorded. The observation was repeated after two weeks, one month, and two months, and micro-CT scans after two months were taken as the gold standard.

RESULTS

The DM results show dentinal crack formation in 47% of the samples following root-end resection and in 87% following ultrasonic preparation. After the ultrasonic preparation, no existing crack propagated to a complete crack, but new cracks were formed. MTA and super-EBA had no effect on crack formation. The Spearman correlation coefficient between OCT and DM was 0.186 (very weak correlation; p = 0.015). Sensitivity and specificity in comparison to micro-CT were 0.50 and 0.55 in OCT and 1.00 and 0.35 in DM, respectively. McNemar's test showed a significant difference between OCT and DM (p < 0.05).

CONCLUSION

Apical resection and ultrasonic preparation could form dentinal cracks. OCT and DM showed different detection frequencies of cracks with very weak correlation. DM showed superior sensitivity compared with OCT.

Optical coherence tomography assessment of gingival epithelium inflammatory status in periodontal - Systemic affected patients

INTRODUCTION

Optical coherence tomography (OCT) is an imaging tool used in various medical fields (ophthalmology, dermatology), which allows the observation of morphological particularities on the surface of tissues or internal constructive details of about 2-3mms in depth. In periodontology, it has been used as an experimental tool for periodontal pocket analysis (depth, calculus deposits) but not for the assessment of periodontal inflammation in the gingival tissues, which has been the subject of our in-vitro study.

MATERIAL AND METHOD

Gingival samples were collected from three types of patients: patients with periodontal disease; patients with periodontal disease and a systemic comorbidity; periodontal and systemic healthy patients. The samples were scanned with an OCT light beam, resulting two-dimensional images of the gingival tissue (full thickness epithelium and partial connective tissue). The images were assessed using dedicated software, which allowed the quantification of pixels on a given segment in the epithelium. The average pixel densities were then calculated for each patient group and statistically analyzed.

RESULTS

The resulted pixel densities were highest for the control group samples, while the lowest pixel densities were found in samples originating from periodontal patients with diabetes mellitus. For the other possible periodontal comorbidity, chronic hepatitis C, image assessment also exhibited lower pixel densities than those of the periodontal group, suggesting that this condition could also have an added effect on the tissular changes induced by periodontal disease.

CONCLUSION

OCT has proven that in an in-vitro environment it can be a useful tool for the assessment of periodontal inflammation in gingival samples of periodontal patients. In terms of inflammatory tissular changes observed by OCT analysis, chronic hepatitis C could be regarded as possible periodontal disease's comorbidity.

The Use of Optical Coherence Tomography in Dental Diagnostics: A State-of-the-Art Review

Optical coherence tomography provides sections of tissues in a noncontact and noninvasive manner. The device measures the time delay and intensity of the light scattered or reflected from biological tissues, which results in tomographic imaging of their internal structure. This is achieved by scanning tissues at a resolution ranging from 1 to 15 μm. OCT enables real-time in situ imaging of tissues without the need for biopsy, histological procedures, or the use of X-rays, so it can be used in many fields of medicine. Its properties are not only particularly used in ophthalmology, in the diagnosis of all layers of the retina, but also increasingly in cardiology, gastroenterology, pulmonology, oncology, and dermatology. The basic properties of OCT, that is, noninvasiveness and low wattage of the used light, have also been appreciated in analytical technology by conservators, who use it to identify the quality and age of paintings, ceramics, or glass. Recently, the OCT technique of visualization is being tested in different fields of dentistry, which is depicted in the article.