DetSegDiff: A joint periodontal landmark detection and segmentation in intraoral ultrasound using edge-enhanced diffusion-based network.

Individuals with malocclusion require an orthodontic diagnosis and treatment plan based on the severity of their condition. Assessing and monitoring changes in periodontal structures before, during, and after orthodontic procedures is crucial, and intraoral ultrasound (US) imaging has been shown a promising diagnostic tool in imaging periodontium. However, accurately delineating and analyzing periodontal structures in US videos is a challenging task for clinicians, as it is time-consuming and subject to interpretation errors. This paper introduces DetSegDiff, an edge-enhanced diffusion-based network developed to simultaneously detect the cementoenamel junction (CEJ) and segment alveolar bone structure in intraoral US videos. An edge feature encoder is designed to enhance edge and texture information for precise delineation of periodontal structures. Additionally, we employed the spatial squeeze-attention module (SSAM) to extract more representative features to perform both detection and segmentation tasks at global and local levels. This study used 169 videos from 17 orthodontic patients for training purposes and was subsequently tested on 41 videos from 4 additional patients. The proposed method achieved a mean distance difference of 0.17 ± 0.19 mm for the CEJ and an average Dice score of 90.1% for alveolar bone structure. As there is a lack of multi-task benchmark networks, thorough experiments were undertaken to assess and benchmark the proposed method against state-of-the-art (SOTA) detection and segmentation individual networks. The experimental results demonstrated that DetSegDiff outperformed SOTA approaches, confirming the feasibility of using automated diagnostic systems for orthodontists.

Intraoral Ultrasound Imaging Using a Rotational Transducer with Periodontal Feature Identification by Machine Learning.

Innovative intraoral ultrasound devices with smart artificial intelligence-based identification for dento-anatomy could provide crucial information for oral health diagnosis and treatment and shed light on real-time detection of developmental dentistry. However, the grand challenge is that the current ultrasound technologies are meant for external use due to their bulkiness and low frequency. We report a compact versatile ultrasound intraoral device that consists of a rotational probe head robustly pivoted around a hand-held and portable handle for real-time imaging of intraoral anatomy using high-frequency ultrasonography (up to 25 MHz). The intraoral ultrasound device that could be adjusted for various orientations of the imaging planes by rotating the head provides real-time, high-resolution ultrasonograms of intraoral structures, including dento-periodontium of most tooth types and maxillary palate. Machine learning-based algorithms are integrated to automate the identification of important structures, including alveolar bone and cementum-enamel junction. The intraoral ultrasound device smartened with artificial intelligence could innovate oral health diagnosis and treatment plans toward precision health and patient care.

Ultrasonic mapping of midpalatal suture - An ex-vivo study.

OBJECTIVE: Rapid maxillary expansion is a common orthodontic procedure to correct maxillary constriction. Assessing the midpalatal suture (MPS) expansion plays a crucial role in treatment planning to determine its effectiveness. The objectives of this preliminary investigation are to demonstrate a proof of concept that the palatal bone underlying the rugae can be clearly imaged by ultrasound (US) and the reconstructed axial view of the US image accurately maps the MPS patency. METHODS: An ex-vivo US scanning was conducted on the upper jawbones of two piglet's carcasses before and after the creation of bone defects, which simulated the suture opening. The planar images were processed to enhance bone intensity distribution before being orderly stacked to fuse into a volume. Graph-cut segmentation was applied to delineate the palatal bone to generate a bone volume. The accuracy of the reconstructed bone volume and the suture opening was validated by the micro-computed tomography (µCT) data used as the ground truth and compared with cone beam computed tomography (CBCT) data as the clinical standard. Also included in the comparison is the rugae thickness. Correlation and Bland-Altman plots were used to test the agreement between the two methods: US versus µCT/CBCT. RESULTS: The reconstruction of the US palatal bone volumes was accurate based on surface topography comparison with a mean error of 0.19 mm for pre-defect and 0.15 mm and 0.09 mm for post-defect models of the two samples, respectively when compared with µCT volumes. A strong correlation (R2 ≥ 0.99) in measuring MPS expansion was found between US and µCT/CBCT with MADs of less than 0.05 mm, 0.11 mm and 0.23 mm for US, µCT and CBCT, respectively. CONCLUSIONS: It was possible to axially image the MPS opening and rugae thickness accurately using high-frequency ultrasound. CLINICAL SIGNIFICANCE: This study introduces an ionizing radiation-free, low-cost, and portable technique to accurately image a difficult part of oral cavity anatomy. The advantages of conceivable visualization could promise a successful clinical examination of MPS to support the predictable treatment outcome of maxillary transverse deficiency.

Measuring the alveolar bone level in adolescents: A comparison between ultrasound and cone beam computed tomography.

BACKGROUND: Cone beam computed tomography (CBCT) is an imaging modality, which is used routinely in orthodontic diagnosis and treatment planning but delivers much higher radiation than conventional dental radiographs. Ultrasound is a noninvasive imaging method that creates an image without ionizing radiation. AIM: To investigate the reliability of ultrasound and the agreement between ultrasound and CBCT in measuring the alveolar bone level (ABL) on the buccal/labial side of the incisors in adolescent orthodontic patients. DESIGN: One hundred and eighteen incisors from 30 orthodontic adolescent patients were scanned by CBCT with 0.3-mm voxel size and ultrasound at 20 MHz frequency. The ABL, distance from the cementoenamel junction (CEJ) to the alveolar bone crest (ABC), was measured twice to evaluate the agreement between ultrasound and CBCT. In addition, the intra- and inter-rater reliabilities in measuring the ABL by four raters were compared. RESULTS: The mean difference (MD) in the ABL between ultrasound and CBCT was -0.07 mm with 95% limit of agreement (LoA) from -0.47 to 0.32 mm for all teeth. For each jaw, the MDs between the ultrasound and CBCT were -0.18 mm (for mandible with 95% LoA from -0.53 to 0.18 mm) and 0.03 mm (for maxilla with 95% LoA from -0.28 to 0.35 mm). In comparison, ultrasound had higher intra-rater (ICC = 0.83-0.90) and inter-rater reliabilities (ICC = 0.97) in ABL measurement than CBCT (ICC = 0.56-0.78 for intra-rater and ICC = 0.69 for inter-rater reliabilities). CONCLUSION: CBCT parameters used in orthodontic diagnosis and treatment planning in adolescents may not be a reliable tool to assess the ABL for the mandibular incisors. On the contrary, ultrasound imaging, an ionizing radiation-free, inexpensive, and portable diagnostic tool, has potential to be a reliable diagnostic tool in assessing the ABL in adolescent patients.

Estimating Crestal Thickness of Alveolar Bones on Intra-oral Ultrasonograms.

OBJECTIVE: Alveolar crestal bone thickness and level provide important diagnostic and prognostic information for orthodontic treatment, periodontal disease management and dental implants. Ionizing radiation-free ultrasound has emerged as a promising clinical tool in imaging oral tissues. However, the ultrasound image is distorted when the wave speed of the tissue of interest is different from the mapping speed of the scanner and, therefore, the subsequent dimension measurements are not accurate. This study was aimed at deriving a correction factor that can be applied to the measurements to correct for discrepancy caused by speed variation. METHODS: The factor is a function of the speed ratio and the acute angle that the segment of interest makes with the beam axis perpendicular to the transducer. The phantom and cadaver experiments were designed to validate the method. DISCUSSION: The comparisons agree well with absolute errors not more than 4.9%. Dimension measurements on ultrasonographs can be properly corrected by applying the correction factor without recourse to the raw signals. CONCLUSION: The correction factor has reduced the measurement discrepancy on the acquired ultrasonographs for the tissue whose speed is different from the scanner's mapping speed.

Temporomandibular joint segmentation in MRI images using deep learning.

OBJECTIVES: Temporomandibular joint (TMJ) internal derangements (ID) represent the most prevalent temporomandibular joint disorder (TMD) in the population and its diagnosis typically relies on magnetic resonance imaging (MRI). TMJ articular discs in MRIs usually suffer from low resolution and contrast, and it is difficult to identify them. In this study, we applied two convolutional neural networks (CNN) to delineate mandibular condyle, articular eminence, and TMJ disc in MRI images. METHODS: The models were trained on MRI images from 100 patients and validated on images from 40 patients using 2D slices and 3D volume as input, respectively. Data augmentation and five-fold cross-validation scheme were applied to further regularize the models. The accuracy of the models was then compared with four raters having different expertise in reading TMJ-MRI images to evaluate the performance of the models. RESULTS: Both models performed well in segmenting the three anatomical structures. A Dice coefficient of about 0.7 for the articular disc, more than 0.9 for the mandibular condyle, and Hausdorff distance of about 2mm for the articular eminence were achieved in both models. The models reached near-expert performance for the segmentation of TMJ articular disc and performed close to the expert in the segmentation of mandibular condyle and articular eminence. They also surpassed non-experts in segmenting the three anatomical structures. CONCLUSION: This study demonstrated that CNN-based segmentation models can be a reliable tool to assist clinicians identifying key anatomy on TMJ-MRIs. The approach also paves the way for automatic diagnosis of TMD. CLINICAL SIGNIFICANCE: Accurately locating the articular disc is the hardest and most crucial step in the interpretation of TMJ-MRIs and consequently in the diagnosis of TMJ-ID. Automated software that assists in locating the articular disc and its surrounding structures would improve the reliability of TMJ-MRI interpretation, save time and assist in reader training. It will also serve as a foundation for additional automated analysis of pathology in TMJ structures to aid in TMD diagnosis.

Localization of cementoenamel junction in intraoral ultrasonographs with machine learning.

OBJECTIVE: Our goal was to automatically identify the cementoenamel junction (CEJ) location in ultrasound images using deep convolution neural networks (CNNs). METHODS: Three CNNs were evaluated using 1400 images and data augmentation. The training and validation were performed by an experienced nonclinical rater with 1000 and 200 images, respectively. Four clinical raters with different levels of experience with ultrasound tested the networks using the other 200 images. In addition to the comparison of the best approach with each rater, we also employed the simultaneous truth and performance level estimation (STAPLE) algorithm to estimate a ground truth based on all labelings by four clinical raters. The final CEJ location estimate was obtained by taking the first moment of the posterior probability computed using the STAPLE algorithm. The study also computed the machine learning-measured CEJ-alveolar bone crest distance. RESULTS: Quantitative evaluations of the 200 images showed that the comparison of the best approach with the STAPLE-estimate yielded a mean difference (MD) of 0.26 mm, which is close to the comparison with the most experienced nonclinical rater (MD=0.25 mm) but far better than the comparison with clinical raters (MD=0.27-0.33 mm). The machine learning-measured CEJ-alveolar bone crest distances correlated strongly (R = 0.933, p < 0.001) with the manual clinical labeling and the measurements were in good agreement with the 95% Bland-Altman's lines of agreement between -0.68 and 0.57 mm. CONCLUSIONS: The study demonstrated the feasible use of machine learning methodology to localize CEJ in ultrasound images with clinically acceptable accuracy and reliability. Likelihood-weighted ground truth by combining multiple labels by the clinical experts compared favorably with the predictions by the best deep CNN approach. CLINICAL SIGNIFICANCE: Identification of CEJ and its distance from the alveolar bone crest play an important role in the evaluation of periodontal status. Machine learning algorithms can learn from complex features in ultrasound images and have potential to provide a reliable and accurate identification in subsecond. This will greatly assist dental practitioners to provide better point-of-care to patients and enhance the throughput of dental care.

Three-dimensional periodontal investigations using a prototype handheld ultrasound scanner with spatial positioning reading sensor.

AIM: To demonstrate the feasibility of the 3D ultrasound periodontal tissue reconstruction of the lateral area of a porcine mandible using standard 2D ultrasound equipment and spatial positioning reading sensors. MATERIAL AND METHOD: Periodontal 3D reconstructions were performed using a free-hand prototype based on a 2D US scanner and a spatial positioning reading sensor. For automated data processing, deep learning algorithms were implemented and trained using semi-automatically seg-mented images by highly specialized imaging professionals. RESULTS: US probe movement analysis showed that non-parallel 2D frames were acquired during the scanning procedure. Comparing 3 different 3D periodontal reconstructions of the same porcine mandible, the accuracy ranged between 0.179 mm and 0.235 mm. CONCLUSION: The present study demonstrated the diagnostic potential of 3D reconstruction using a free-hand 2D US scanner with spatial positioning readings. The use of auto-mated data processing with deep learning algorithms makes the process practical in the clinical environment for assessment of periodontal soft tissues.

Polyacrylamide/Alginate double-network tough hydrogels for intraoral ultrasound imaging.

HYPOTHESIS: Intraoral ultrasound is a safer and economical approach to image dento-periodontal tissues and diagnose periodontal diseases compared with X-ray. A gel pad is often used as a couplant between the transducer and oral tissue to delay the ultrasound signals for better identification. However, the current commercial couplant, such as Aquaflex gel pad (AF), face many challenges, including low stability in water, poor mechanical properties, low coefficient of friction, and potential cytotoxicity issues. Polyacrylamide/sodium alginate (PAM/Alginate) double-network (DN) tough hydrogel could address these issues as the potential couplant for intraoral ultrasound imaging. EXPERIMENTS: Different critical properties required for intraoral ultrasound imaging, including stability in water, mechanical properties, frictional properties, ultrasound properties and biocompatibility of PAM/Alginate DN tough hydrogels were evaluated and compared with those of AF. FINDINGS: The PAM/Alginate DN hydrogel not only possesses better stability in water as well as improved mechanical properties and higher coefficients of friction than AF but also can provide similar ultrasound image quality as AF does. Moreover, the PAM/Alginate DN hydrogel shows lower cytotoxicity to both cancer (Hela) and fibroblast cells (MRC-5). With all these significant features, such tough hydrogels serve as a proof-of-concept ultrasound couplant with great potential in intraoral ultrasound imaging.

Mussel-Inspired Adhesive Double-Network Hydrogel for Intraoral Ultrasound Imaging.

Periodontal diseases could be diagnosed through intraoral ultrasound imaging with the advantages of simple operation procedures, low cost, and low safety risks. A couplant is normally placed between transducers and tissues for better ultrasound image quality. If applied intraorally, the couplants should possess good stability in water and robust mechanical properties, as well as strong adhesiveness to transducers and tissues. However, commercial couplants, such as Aquaflex (AF) cannot fulfill these requirements. In this work, inspired by the mussel adhesion mechanism, we reported a poly(vinyl alcohol)-polyacrylamide-polydopamine (PVA-PAM-PDA) hydrogel synthesized by incorporating PDA into the PAM-PVA double-network for intraoral ultrasound imaging. The hydrogel maintains good stability in water as well as exceptional mechanical properties and can adhere to different substrates (i.e., metal, glass, and porcine skin) without losing the original adhesion strength after multiple adhesion-strip cycles. Besides, when applied to porcine mandibular incisor imaging, the PVA-PAM-PDA hydrogel possesses good image quality for diagnosis as AF does. This work provides practical insights into the fabrication of multifunctional hydrogel-based interfaces between human tissues and medical devices for disease diagnosis applications.

Fully Automated Segmentation of Alveolar Bone Using Deep Convolutional Neural Networks from Intraoral Ultrasound Images.

Delineation of alveolar bone aids the diagnosis and treatment of periodontal diseases. In current practice, conventional 2D radiography and 3D cone-beam computed tomography (CBCT) imaging are used as the non-invasive approaches to image and delineate alveolar bone structures. Recently, high-frequency ultrasound imaging is proposed as an alternative to conventional imaging methods to prevent the harmful effects of ionizing radiation. However, the manual delineation of alveolar bone from ultrasound imaging is time-consuming and subject to inter and intraobserver variability. This study proposes to use a convolutional neural network-based machine learning framework to automatically segment the alveolar bone from ultrasound images. The proposed method consists of a homomorphic filtering based noise reduction and a u-net machine learning framework for automated delineation. The proposed method was evaluated over 15 ultrasound images of tooth acquired from procine specimens. The comparisons against manual ground truth delineations performed by three experts in terms of mean Dice score and Hausdorff distance values demonstrate that the proposed method yielded an improved performance over a recent state of the art graph cuts based method.

Comparison of ultrasound imaging and cone-beam computed tomography for examination of the alveolar bone level: A systematic review.

BACKGROUND AND OBJECTIVE: The current methods to image alveolar bone in humans include intraoral 2D radiography and cone-beam computed tomography (CBCT). However, these methods expose the subject to ionizing radiation. Therefore, ultrasound imaging has been investigated as an alternative technique, as it is both non-invasive and free from ionizing radiation. In order to assess the validity and reliability of ultrasonography in visualizing alveolar bone, a systematic review was conducted comparing ultrasound imaging to CBCT for examination of the alveolar bone level. STUDY DESIGN: Seven databases were searched. Studies addressing examination of alveolar bone level via CBCT and ultrasound were selected. Risk of bias under Cochrane guidelines was used as a methodological quality assessment tool. RESULTS: All the four included studies were ex vivo studies that used porcine or human cadaver samples. The alveolar bone level was measured by the distance from the alveolar bone crest to certain landmarks such as cemento-enamel junction or gingival margin. The risk of bias was found as low. The mean difference between ultrasound and CBCT measurements ranged from 0.07 mm to 0.68 mm, equivalent to 1.6% - 8.8%. CONCLUSIONS: There is currently preliminary evidence to support the use of ultrasonography as compared to CBCT for the examination of alveolar bone level. Further studies comparing ultrasound to gold standard methods would be necessary to help validate the accuracy of ultrasonography as a diagnostic technique in periodontal imaging.

High-Resolution Ultrasonic Imaging of Dento-Periodontal Tissues Using a Multi-Element Phased Array System.

Intraoral ultrasonography uses high-frequency mechanical waves to study dento-periodontium. Besides the advantages of portability and cost-effectiveness, ultrasound technique has no ionizing radiation. Previous studies employed a single transducer or an array of transducer elements, and focused on enamel thickness and distance measurement. This study used a phased array system with a 128-element array transducer to image dento-periodontal tissues. We studied two porcine lower incisors from a 6-month-old piglet using 20-MHz ultrasound. The high-resolution ultrasonographs clearly showed the cross-sectional morphological images of the hard and soft tissues. The investigation used an integration of waveform analysis, travel-time calculation, and wavefield simulation to reveal the nature of the ultrasound data, which makes the study novel. With the assistance of time-distance radio-frequency records, we robustly justified the enamel-dentin interface, dentin-pulp interface, and the cemento-enamel junction. The alveolar crest level, the location of cemento-enamel junction, and the thickness of alveolar crest were measured from the images and compared favorably with those from the cone beam computed tomography with less than 10% difference. This preliminary and fundamental study has reinforced the conclusions from previous studies, that ultrasonography has great potential to become a non-invasive diagnostic imaging tool for quantitative assessment of periodontal structures and better delivery of oral care.

Imaging the Cemento-Enamel Junction Using a 20-MHz Ultrasonic Transducer.

The cemento-enamel junction (CEJ), which is the intersection between enamel and cementum, is an important landmark in the diagnosis of periodontal disease. Pulse-echo ultrasound was used to image the CEJs of six porcine lower central incisors with a single 20-MHz transducer. A notch was longitudinally created on the enamel as a stable marker, from which the CEJ was measured. Data were acquired along the tooth's axis at 0.4-mm intervals. Time-distance data were bandpass-filtered to enhance signal-to-noise ratio and record density was increased fourfold to 0.1-mm spacing by a frequency-distance interpolation scheme. Reflections from the CEJ were unambiguously identified along with those from enamel, dentin and cementum. The notch-CEJ distances measured by the ultrasound and micro-computed tomography methods correlated strongly (r = 0.996, p < 0.05) and were in good agreement with the 95% lines of agreement between -0.49 and 0.17 mm, as statistically determined by Bland-Altman analysis. The results indicate the potential of ultrasound to be a reliable and non-ionizing technique to image the CEJ.

Imaging ultrasonic dispersive guided wave energy in long bones using linear radon transform.

Multichannel analysis of dispersive ultrasonic energy requires a reliable mapping of the data from the time-distance (t-x) domain to the frequency-wavenumber (f-k) or frequency-phase velocity (f-c) domain. The mapping is usually performed with the classic 2-D Fourier transform (FT) with a subsequent substitution and interpolation via c = 2πf/k. The extracted dispersion trajectories of the guided modes lack the resolution in the transformed plane to discriminate wave modes. The resolving power associated with the FT is closely linked to the aperture of the recorded data. Here, we present a linear Radon transform (RT) to image the dispersive energies of the recorded ultrasound wave fields. The RT is posed as an inverse problem, which allows implementation of the regularization strategy to enhance the focusing power. We choose a Cauchy regularization for the high-resolution RT. Three forms of Radon transform: adjoint, damped least-squares, and high-resolution are described, and are compared with respect to robustness using simulated and cervine bone data. The RT also depends on the data aperture, but not as severely as does the FT. With the RT, the resolution of the dispersion panel could be improved up to around 300% over that of the FT. Among the Radon solutions, the high-resolution RT delineated the guided wave energy with much better imaging resolution (at least 110%) than the other two forms. The Radon operator can also accommodate unevenly spaced records. The results of the study suggest that the high-resolution RT is a valuable imaging tool to extract dispersive guided wave energies under limited aperture.

Excitation of ultrasonic Lamb waves using a phased array system with two array probes: phantom and in vitro bone studies.

Long bones are good waveguides to support the propagation of ultrasonic guided waves. The low-order guided waves have been consistently observed in quantitative ultrasound bone studies. Selective excitation of these low-order guided modes requires oblique incidence of the ultrasound beam using a transducer-wedge system. It is generally assumed that an angle of incidence, θi, generates a specific phase velocity of interest, co, via Snell's law, θi=sin(-1)(vw/co) where vw is the velocity of the coupling medium. In this study, we investigated the excitation of guided waves within a 6.3-mm thick brass plate and a 6.5-mm thick bovine bone plate using an ultrasound phased array system with two 0.75-mm-pitch array probes. Arranging five elements as a group, the first group of a 16-element probe was used as a transmitter and a 64-element probe was a receiver array. The beam was steered for six angles (0°, 20°, 30°, 40°, 50°, and 60°) with a 1.6-MHz source signal. An adjoint Radon transform algorithm mapped the time-offset matrix into the frequency-phase velocity dispersion panels. The imaged Lamb plate modes were identified by the theoretical dispersion curves. The results show that the 0° excitation generated many modes with no modal discrimination and the oblique beam excited a spectrum of phase velocities spread asymmetrically about co. The width of the excitation region decreased as the steering angle increased, rendering modal selectivity at large angles. The phenomena were well predicted by the excitation function of the source influence theory. The low-order modes were better imaged at steering angle ⩾30° for both plates. The study has also demonstrated the feasibility of using the two-probe phased array system for future in vivo study.