A relation-based framework for effective teeth recognition on dental periapical X-rays.

Dental periapical X-rays are used as a popular tool by dentists for diagnosis. To provide dentists with diagnostic support, in this paper, we achieve automated teeth recognition of dental periapical X-rays by using deep learning techniques, including teeth location and classification. Convolutional neural network(CNN) is a popular method and has made large improvements in medical image applications. However, in our specific task, the performance of CNN is limited by lack of data and too many teeth positions in X-rays. Addressing this problem, we consider to utilize the prior dental knowledge, and therefore we propose a relation-based framework to handle the teeth location and classification task. According to the relation in teeth labels, we apply a special label reconstruction technique to decompose the teeth classification task, and use a multi-task CNN to classify the teeth positions. Meanwhile, for teeth location task, we design a proposal correlation module to use the information in teeth positions, and insert it into the multi-task CNN. A teeth sequence refinement module is used for the post processing. Our experiment results show that our relation-based framework achieves high teeth classification and location performance, which is a big improvement compared to the direct use of famous detection structures. With reliable teeth information, our method can provide automated diagnostic support for the dentists.

A Novel Method for Adjusting the Taper and Adaption of Automatic Tooth Preparations with a High-Power Femtosecond Laser.

This study explored the effect of the light-off delay setting in a robotically controlled femtosecond laser on the taper and adaption of resin tooth preparations. Thirty resin teeth (divided into six equal groups) were studied under different light-off delay conditions. Tapers from six vertical sections of the teeth were measured and compared among the light-off delay groups. The mean taper decreased from 39.268° ± 4.530° to 25.393° ± 5.496° as the light-off delay increased (p < 0.05). The average distance between the occlusal surfaces of the scanned data and the predesigned preparation data decreased from 0.089 ± 0.005 to 0.013 ± 0.030 µm as the light-off delay increased (p < 0.05). The light-off delay of the femtosecond laser is correlated with the taper and adaption of automatic tooth preparations; this setting needs to be considered during automatic tooth preparation.

In vivo evaluation of bending strengths and degradation rates of different magnesium pin designs for oral stapler.

Magnesium alloys have been potential biodegradable implants in the areas of bone, cardiovascular system, gastrointestinal tract, and so on. The purpose of this study is to evaluate Mg-2Zn alloy degradation as a potential suture material. The study included Sprague-Dawley (SD) rats in vivo. In 24 male SD rats, tests in the leg muscle were conducted using traditional surgical incision and insertion of magnesium alloys of different designs into the tissue. The material degradation topography, elemental composition, and strength of the pins were analyzed. This paper explores magnesium pins with different cross-sectional shapes and diameters to establish a suitable pin diameter and shape for use as an oral stapler, which must have a good balance of degradation rate and strength. The results showed there were good bending strengths over different degradation periods in groups with diameters of 0.8 mm and 0.5 mm, and no significantly different bending strength between the groups of triangle and round cross-section shapes with same diameter of 0.3 mm, although the degradation rate still needs to be improved.

Method for Accurately Preparing Cavities on Cortical Bones Using Picosecond Laser.

Objective: This study was conducted to (1) evaluate a new method for accurately and automatically preparing dental implant cavities; (2) investigate the quantitative relationships between the number of focal-plane additive pulse layers (n) in two-dimensional ablation, the Z-axis feed rate, and the ablation depth (d) during cortical-bone ablation using a numerically controlled three-axis picosecond laser; and (3) establish appropriate methods for precise ablation control. Materials and methods: Two-dimensional ablation was performed on swine-rib blocks in the focal plane on a preset circular path using a picosecond laser device and an in-house-developed three-axis numerically controlled micro-laser galvanometer scanner. The maximum two-dimensional d and the quantitative relationship between n and d within the maximum d were consequently obtained. The measured and theoretical values of the ablated cavities were then compared to obtain n and d values corresponding to the minimum difference, and to evaluate the error in d, resulting in a higher-accuracy d value (i.e., single-step ablation depth) being obtained. Results: The diameter and deep errors between the measured and design data for 24 cavities were 2.76 ± 1.51 and 10.23 ± 4.82 µm, respectively. Thus, high-quality cortical-bone cavities preparation was achieved using a picosecond laser with the parameters employed in this study. Conclusions: Precise control of cortical-bone ablation using a picosecond laser can be attained by optimizing the single-step ablation parameters.

Automatic drawing of customized removable partial denture diagrams based on textual design for the clinical decision support system.

Qualified diagrams of removable partial denture (RPD) designs created by dentists provide technicians with clear and dynamic information. Generating RPD design in the clinical decision support system (CDSS) can be achieved by producing the RPD design in a textual format and then transferring the design onto diagrams. The drawing of RPD diagrams automatically and efficiently for the given textual designs is still under investigation. A new workflow consisting of three major steps is developed to produce and visualize two-dimensional RPD design diagrams. Annotations and orientations of teeth are established from the base maps in the first step, and built-in rules are then incorporated to describe the variations caused by the interactions of the RPD components. Finally, the software draws each component using a series of curve functions. To validate the performance of the software, 112 RPD clinical design plans are randomly selected as inputs for the software, and the outputs are independently verified by experienced clinicians. The proposed methods are proven to be efficient and accurate and thus can be used to improve clinical quality.

Potassium iodide enhances the photobactericidal effect of methylene blue on Enterococcus faecalis as planktonic cells and as biofilm infection in teeth.

OBJECTIVE: To explore the effectiveness, biosafety, photobleaching and mechanism of antimicrobial photodynamic therapy (aPDT) using methylene blue (MB) plus potassium iodide (KI), for root canal infections. METHODS: Different combinations and concentrations of MB, KI and 660 nm LED light were used against E. faecalis in planktonic and in biofilm states by colony-forming unit (CFU), confocal laser scanning microscopy (CLSM), scanning electron microscopy (SEM). Human gingival fibroblasts (HGF) were used for safety testing by Cell Counting Kit-8 (CCK8) and fluorescence microscopy (FLM). The photobleaching effect and mechanisms were analyzed. RESULTS: KI could not only enhance MB aPDT on E. faecalis in both planktonic and biofilm states even in a hypoxic environment, but also produced a long-lasting bactericidal effect after end of the illumination. KI could accelerate photobleaching to reduce tooth staining by MB, and the mixture was harmless for HGFs. Mechanistic studies showed the generation of hydrogen peroxide and free iodine, and iodine radicals may be formed in hypoxia. CONCLUSION: aPDT with MB plus KI could be used for root canal disinfection and clinical studies are worth pursuing.

Photobiomodulation therapy for management of inferior alveolar nerve injury post-extraction of impacted lower third molars.

PURPOSE: Inferior alveolar nerve (IAN) injury is one of the most serious complications after extraction of impacted lower third molars. Photobiomodulation (PBM) therapy has been noted to reduce pain and inflammation while promoting tissue healing. This study examined the efficacy of PBM therapy tested in a case series of patients with postoperative IAN injury. MATERIAL AND METHODS: 20 patients with post-extraction IAN injury were involved in this study and divided into two groups. In the study group, PBM therapy (808-nm laser, 16 mW, 3 J/cm2) was used every other day for 2 weeks solely on post-extraction sockets in 10 patients diagnosed with IAN injury. In the control group, mecobalamine was prescribed to 10 patients with IAN injury. Objective and subjective recovery of IAN paresthesia was evaluated using clinical neurosensory testing and visual analog score. RESULTS: All patients showed improvement in both objective and subjective examination. Notably, the visual analog score was significantly improved after PBM treatment compared to the mecobalamine treatment (p < 0.05). CONCLUSION: PBM therapy with 808-nm laser appears to be an effective approach to manage paresthesia post-IAN injury following impacted third molar surgery. Given the limited sample size in this study, large-scale, placebo-controlled, multi-center randomized controlled trials are needed for further validation of this innovative treatment.

A deep learning approach to automatic teeth detection and numbering based on object detection in dental periapical films.

We propose using faster regions with convolutional neural network features (faster R-CNN) in the TensorFlow tool package to detect and number teeth in dental periapical films. To improve detection precisions, we propose three post-processing techniques to supplement the baseline faster R-CNN according to certain prior domain knowledge. First, a filtering algorithm is constructed to delete overlapping boxes detected by faster R-CNN associated with the same tooth. Next, a neural network model is implemented to detect missing teeth. Finally, a rule-base module based on a teeth numbering system is proposed to match labels of detected teeth boxes to modify detected results that violate certain intuitive rules. The intersection-over-union (IOU) value between detected and ground truth boxes are calculated to obtain precisions and recalls on a test dataset. Results demonstrate that both precisions and recalls exceed 90% and the mean value of the IOU between detected boxes and ground truths also reaches 91%. Moreover, three dentists are also invited to manually annotate the test dataset (independently), which are then compared to labels obtained by our proposed algorithms. The results indicate that machines already perform close to the level of a junior dentist.

Adaptation of removable partial denture frameworks fabricated by selective laser melting.

STATEMENT OF PROBLEM: Selective laser melting (SLM) is a novel 3-dimensional (3D) printing technology that can directly form the metal frameworks of removable partial dentures. The adaptation of SLM frameworks has not been thoroughly evaluated. PURPOSE: The purpose of this in vitro study was to evaluate the tissue surface adaptation of removable partial denture frameworks fabricated by an SLM technique. MATERIAL AND METHODS: Four types of maxillary partial edentulous resin models were custom made: bilateral second premolars and molars missing, bilateral premolars and first molars missing, all teeth missing except 2 canines, and 2 central incisors missing. According to these dentition-defect patterns, 4 types (I, II, III, and IV) of virtual removable partial denture frameworks were designed, and an SLM printer was used for 3D printing using cobalt-chromium (Co-Cr) alloys (repeated 3 times). As a control, refractory casts duplicated from the resin models were used to fabricate denture frameworks by the lost-wax casting technique. Average gaps and maximum gaps between frameworks and models were measured using the silicone impression material. Two-way ANOVA was used to determine the influence of production methods and design types on the gaps (α=.05). RESULTS: The 2-way ANOVA showed that average gaps were significantly influenced by the production methods and design types, as well as their interactions (P<.001). With design Types I and II, the average gaps of the SLM-printed frameworks were larger than those of the cast ones (P<.001). However, no such differences were found for design Type III, P=.325, or IV, P=.862. CONCLUSIONS: SLM-printed frameworks achieved an acceptable adaptation. However, among frameworks with a large span and relatively more retainers and clasps, the adaptation of those made by the precision casting technique was slightly better than that of those printed by the SLM technique.

Hydroxyapatite/Collagen Three-Dimensional Printed Scaffolds and Their Osteogenic Effects on Human Bone Marrow-Derived Mesenchymal Stem Cells.

IMPACT STATEMENT: Bone loss due to trauma, inflammation, and surgical processes has posed great difficulty in the aesthetic reconstruction of a functional alveolar bone. Tissue engineering and biomaterials, which can promote alveolar bone regeneration, have become a popular focus of current studies. Three-dimensional (3D) printing provides a novel approach to repair bone defects using customized biomimetic tissue scaffolds. Nano hydroxyapatite (nHA) and deproteinized bovine bone (DBB) are two materials mainly used in clinical practice, particularly DBB are widely used in dentistry and craniomaxillofacial orthosis because of the porosity characteristic. To make a bone substitute closest to natural bone structure and composition, nHA and DBB were dispersed into collagen (CoL) to prepare the bioink for 3D printing. The physicochemical and biological properties between the two 3D printing scaffolds were compared. Both nHA/CoL and DBB/CoL 3D printing scaffold would be promising candidate for the clinical applications in the future.

Temperature and depth evaluation of the in vitro effects of femtosecond laser on oral soft tissue, with or without air-cooling.

Femtosecond laser is an effective and safe tool in many surgeries, but the studies of its effect on oral soft tissue ablation are insufficient. This study aimed to investigate the effect of soft tissue ablation with a 1030-nm femtosecond laser on temperature and depth. Twenty Sprague-Dawley rat tongue specimens were obtained and flat-mounted. The 1030-nm femtosecond laser was controlled by a computer system, with a set distance of 4.7 mm between the laser aperture and soft tissue surfaces. Ten specimens were ablated for > 1 min with or without air-cooling for temperature measurement, while the other 10 specimens were ablated for depth measurements, using the following parameters: (i) 3 W, 2000 mm/s; (ii) 3 W, 4000 mm/s; (iii) 5 W, 2000 mm/s; (iv) 5 W, 4000 mm/s; (v) 8 W, 2000 mm/s; (vi) 8 W, 4000 mm/s. Temperature changes were measured using a type-K thermocouple. The depth attained using different power and scanning speed settings was measured by a three-dimensional morphology measurement laser microscope. Laser power, scanning speed, and air-cooling effects were determined. Higher energy and lower speed induced higher temperatures (p < 0.05), which were significantly decreased by air-cooling (p < 0.05). The lowest ablation depth was obtained at 3 W and 4000 mm/s (72.63 ± 6.47 µm) (p < 0.05). The greatest incision depth was achieved at 8 W and 2000 mm/s (696.19 ± 35.37 µm), or 4000 mm/s (681.16 ± 55.65 µm) (p < 0.05). The 1030-nm femtosecond laser application demonstrates clinically acceptable ablation efficiency, without marked temperature damage, in a controlled manner.

Author Correction: Detection of cervical lymph node metastasis from oral cavity cancer using a non-radiating, noninvasive digital infrared thermal imaging system.

A correction to this article has been published and is linked from the HTML and PDF versions of this paper. The error has been fixed in the paper.

An effective teeth recognition method using label tree with cascade network structure.

In this article, we apply the deep learning technique to medical field for the teeth detection and classification of dental periapical radiographs, which is important for the medical curing and postmortem identification. We detect teeth in an input X-ray image and distinguish them from different position. An adult usually has 32 teeth, and some of them are similar while others have very different shape. So there are 32 teeth position for us to recognize, which is a challenging task. Convolutional neural network is a popular method to do multi-class detection and classification, but it needs a lot of training data to get a good result if used directly. The lack of data is a common case in medical field due to patients' privacy. In this work, limited to the available data, we propose a new method using label tree to give each tooth several labels and decompose the task, which can deal with the lack of data. Then use cascade network structure to do automatic identification on 32 teeth position, which uses several convolutional neural network as its basic module. Meanwhile, several key strategies are utilized to improve the detection and classification performance. Our method can deal with many complex cases such as X-ray images with tooth loss, decayed tooth and filled tooth, which frequently appear on patients. The experiments on our dataset show: for small training dataset, compared to the precision and recall by training a 33-classes (32 teeth and background) state-of-the-art convolutional neural network directly, the proposed approach reaches a high precision and recall of 95.8% and 96.1% in total, which is a big improvement in such a complex task.

Detection of cervical lymph node metastasis from oral cavity cancer using a non-radiating, noninvasive digital infrared thermal imaging system.

This study aimed to evaluate the diagnostic performance of a non-radiating, noninvasive infrared (IR) thermal imaging system in the detection of cervical lymph node metastasis from oral cavity cancer. In this prospective clinical trial, a total of 90 oral cavity cancer patients suspected of having cervical lymph node metastasis underwent IR imaging of the neck prior to neck dissection. Analysis of the IR images was performed by two methods: manual qualitative analysis and automatic analysis by an entropy-gradient support vector machine (EGSVM). The efficacies of the EGSVM-based infrared thermal imaging system and contrast-enhanced computed tomography (CT) were compared by using the Noninferiority Testing. Compared with manual qualitative analysis, the EGSVM-based automatic analysis had a higher sensitivity (84.8% vs. 71.7%), specificity (77.3% vs. 72.7%), accuracy (81.1% vs. 72.2%), positive predictive value (79.6% vs. 73.3%) and negative predictive value (82.9% vs. 71.1%). The EGSVM-based infrared thermal imaging system was noninferior to contrast-enhanced CT (P < 0.05). The EGSVM-based infrared thermal imaging system showed a trend of higher sensitivity, whereas contrast-enhanced CT showed a trend of higher specificity. The EGSVM-based infrared thermal imaging system is a promising non-radiating, noninvasive tool for the detection of lymph node metastasis from oral cavity cancer.

Behavior of human periodontal ligament cells on dentin surfaces ablated with an ultra-short pulsed laser.

This study aimed to evaluate the effects of an ultrashort pulsed laser (USPL) (1064 nm, 20 ps, 100 kHz) with different laser fluences (F, 4, 6, 8 J/cm2) and pulse overlaps (PO, 0, 50%) on human periodontal ligament cells (hPDLs) behavior. Dentin samples were ablated with USPL with different combinations of fluences and pulse overlaps; some samples were ablated with an Er:YAG laser (2940 nm, 150 µs, 100 mJ/pulse, 5 J/cm2) and some samples were ground with a carbide bur. Then hPDLs were grown on the samples after different treatments. Dentin morphology and cell adhesion were observed with SEM and gene expressions were measured by RT-PCR. The results showed dentin surfaces ablated with USPL when F = 4 J/cm2, PO = 0, and F = 6 J/cm2, PO = 0 were partially intact with obvious ridges and valleys and cells on these surfaces grew mostly along the valleys. USPL ablated surfaces in other groups were entirely ablated and cell cluster formation was observed. The RT-PCR results showed an upregulation of osteocalcin of cells grown on the dentin after some laser treatment. It can be concluded that USPL could improve the attachment and differentiation of hPDLs and thus potentially promote periodontal tissue regeneration.

Red (660 nm) or near-infrared (810 nm) photobiomodulation stimulates, while blue (415 nm), green (540 nm) light inhibits proliferation in human adipose-derived stem cells.

We previously showed that blue (415 nm) and green (540 nm) wavelengths were more effective in stimulating osteoblast differentiation of human adipose-derived stem cells (hASC), compared to red (660 nm) and near-infrared (NIR, 810 nm). Intracellular calcium was higher after blue/green, and could be inhibited by the ion channel blocker, capsazepine. In the present study we asked what was the effect of these four wavelengths on proliferation of the hASC? When cultured in proliferation medium there was a clear difference between blue/green which inhibited proliferation and red/NIR which stimulated proliferation, all at 3 J/cm2. Blue/green reduced cellular ATP, while red/NIR increased ATP in a biphasic manner. Blue/green produced a bigger increase in intracellular calcium and reactive oxygen species (ROS). Blue/green reduced mitochondrial membrane potential (MMP) and lowered intracellular pH, while red/NIR had the opposite effect. Transient receptor potential vanilloid 1 (TRPV1) ion channel was expressed in hADSC, and the TRPV1 ligand capsaicin (5uM) stimulated proliferation, which could be abrogated by capsazepine. The inhibition of proliferation caused by blue/green could also be abrogated by capsazepine, and by the antioxidant, N-acetylcysteine. The data suggest that blue/green light inhibits proliferation by activating TRPV1, and increasing calcium and ROS.

Regulation and Measurement of the Heat Generated by Automatic Tooth Preparation in a Confined Space.

OBJECTIVE: The aim of this study was to assess and regulate heat generation in the dental pulp cavity and circumambient temperature around a tooth during laser ablation with a femtosecond laser in a confined space. BACKGROUND DATA: The automatic tooth preparing technique is one of the traditional oral clinical technology innovations. In this technique, a robot controlled an ultrashort pulse laser to automatically complete the three-dimensional teeth preparing in a confined space. The temperature control is the main measure for protecting the tooth nerve. METHODS: Ten tooth specimens were irradiated with a femtosecond laser controlled by a robot in a confined space to generate 10 teeth preparation. During the process, four thermocouple sensors were used to record the pulp cavity and circumambient environment temperatures with or without air cooling. A statistical analysis of the temperatures was performed between the conditions with and without air cooling (p < 0.05). RESULTS: The recordings showed that the temperature with air cooling was lower than that without air cooling and that the heat generated in the pulp cavity was lower than the threshold for dental pulp damage. CONCLUSIONS: These results indicate that femtosecond laser ablation with air cooling might be an appropriate method for automatic tooth preparing.

Influence of object translucency on the scanning accuracy of a powder-free intraoral scanner: A laboratory study.

STATEMENT OF PROBLEM: Limited information is available regarding the influence of object translucency on the scanning accuracy of a powder-free intraoral scanner. PURPOSE: The purpose of this in vitro study was to evaluate the scanning accuracy of a confocal microscopy principle powder-free intraoral scanner on ceramic copings and to analyze the relationship between scanning accuracy and object translucency. METHODS: Six slice specimens (12×10 mm) and 6 offset copings (1.00-mm thickness) were made from different translucent homogeneous ceramic blocks (CEREC Blocs, S0-M to S5-M, highest to lowest translucency). The primary sintered zirconia offset coping was produced in the same way as the control. Optical parameters related to the translucency of each slice were measured with a spectrophotometer. Three-dimensional (3D) datasets of the surface morphology of offset copings were obtained by using the intraoral scanner. The same white wax resin bases were used for registration. Quantitative parameters of scanning trueness and precision were measured. One-way ANOVA was used to analyze the values of each parameter among the 6 ceramic blocks. Bivariate correlation was used to analyze the relationships between each parameter of scanning accuracy and translucency (α=.05). RESULTS: Translucent copings showed a positive 3D bias (S0-M to S5-M: 0.149 ±0.038 mm to 0.068 ±0.020 mm), a narrower collar diameter (Dd=-0.067 mm), larger convergence angle (ΔΦ=2.79 degrees), and larger curvature radius of the internal gingivoaxial corner (Δρ=0.236 mm). The smaller the percentage sum of scattering and absorption, the greater was the occurrence of scanning bias (r=-0.918) and curvature (r=-0.935) decrease. CONCLUSIONS: Use of the tested powder-free intraoral scanner, higher translucency objects (greater translucency than S1-M/A1C) resulted in lower scanning accuracy and morphological changes. Therefore, more suitable methods of measurement are still required.

Photobiomodulation of human adipose-derived stem cells using 810nm and 980nm lasers operates via different mechanisms of action.

Photobiomodulation (PBM) using red or near-infrared (NIR) light has been used to stimulate the proliferation and differentiation of adipose-derived stem cells. The use of NIR wavelengths such as 810nm is reasonably well accepted to stimulate mitochondrial activity and ATP production via absorption of photons by cytochrome c oxidase. However, the mechanism of action of 980nm is less well understood. Here we study the effects of both wavelengths (810nm and 980nm) on adipose-derived stem cells in vitro. Both wavelengths showed a biphasic dose response, but 810nm had a peak dose response at 3J/cm2 for stimulation of proliferation at 24h, while the peak dose for 980nm was 10-100 times lower at 0.03 or 0.3J/cm2. Moreover, 980nm (but not 810nm) increased cytosolic calcium while decreasing mitochondrial calcium. The effects of 980nm could be blocked by calcium channel blockers (capsazepine for TRPV1 and SKF96365 for TRPC channels), which had no effect on 810nm. To test the hypothesis that the chromophore for 980nm was intracellular water, which could possibly form a microscopic temperature gradient upon laser irradiation, we added cold medium (4°C) during the light exposure, or pre-incubated the cells at 42°C, both of which abrogated the effect of 980nm but not 810nm. We conclude that 980nm affects temperature-gated calcium ion channels, while 810nm largely affects mitochondrial cytochrome c oxidase.