We propose FALCON, an unfolded variational network for 3D blind deconvolution in Cone Beam Computed Tomography (CBCT), derived from a convexified Mumford-Shah formulation. The method unrolls a Split Bregman solver into a fixed number of learnable stages, combining explicit modelbased regularization with data-driven parameter adaptation, while jointly refining a Gaussian PSF parameter. On dental CBCT volumes, FALCON improves structural sharpness and boundary definition relative to a classical blind variational solver, without requiring paired $\mu \text{CT}$ supervision at inference. We additionally report a simple clustering-based analysis on reconstructed volumes to qualitatively assess structural separability of canal and dentin regions.

Medical image restoration, in particular X-ray computed tomography, remains, to date, an open challenge. A trade-off between the X-ray doses and the image resolution is necessary in medical routine. Thus, many investigations focused on dedicated image processing techniques to allow high-resolution imaging at low doses. While the initial methods were model-based, currently, deep learning architecture-based approaches provide state-of-the-art results in this domain. Recently, physics-aware approaches combining quantum mechanics principles and convolutional neural networks have been proposed to tackle challenging inverse problems, with promising results. These approaches rely on supervised architectures and have been only tested on photographic images. This paper proposes an unsupervised version of a quantum mechanics-based convolutional neural network with a special focus on dental computed tomography image restoration. Comparisons with the initial supervised network and state-of-the-art image restoration methods confirm the promising performance of the proposed unsupervised technique.

In this work, we address the problem of cone beam computed tomography (CBCT) image resolution enhancement by exploiting a newly introduced deep unrolled quantum denoiser, based on quantum interaction theory adapted to computational imaging. Following recent advances in image restoration using the Plug-and-Play (PnP) framework, we impose this external deep learning denoiser as a regularizer within the super-resolution (SR) problem. The quantum-based deep unrolled denoiser combined with a computationally efficient way to deal with the degradation operators, and the PnP scheme, result in an original way of approaching the image resolution enhancement problem. Experiments conducted on dental CBCT images are presented to illustrate the efficiency of the proposed model for image resolution enhancement tasks. The numerical results suggest that the proposed method allows better restoration performances compared to existing state-of-the-art approaches.

Enhancing the spatial resolution of an image is an important field of research in number of applications including medical ones. In this paper, we address the super-resolution (SR) problem exploiting a newly introduced adaptive quantum denoiser which is based on quantum interaction theory applied in an imaging context. In particular, following recent developments, we impose this external denoiser as a prior function within the Plug-and-Play (PnP) and Regularization by Denoising (RED) approaches. This quantum denoiser combined with, on the one hand, a computationally efficient way of handing both decimation and blur operators, and on the other hand PnP and RED schemes, shows an original way of solving the SR problems. Dental computed tomography images are used to illustrate the potential of the proposed algorithms for high-resolution image retrieval. Numerical experiments show that the proposed methods provide comparable or slightly better results than existing methods.

This paper introduces a novel computationally efficient method of solving the 3D single image super-resolution (SR) problem, i.e., reconstruction of a high-resolution volume from its low-resolution counterpart. The main contribution lies in the original way of handling simultaneously the associated decimation and blurring operators, based on their underlying properties in the frequency domain. In particular, the proposed decomposition technique of the 3D decimation operator allows a straightforward implementation for Tikhonov regularization, and can be further used to take into consideration other regularization functions such as the total variation, enabling the computational cost of state-of-the-art algorithms to be considerably decreased. Numerical experiments carried out showed that the proposed approach outperforms existing 3D SR methods.

The aim of this study was to compare shaping abilities of Protaper Gold® (PTG) and 2Shape® (TS) by using a new automatic process and micro‐computed tomography (Micro‐CT). 32 first mandibular molars with two separate mesial canals were selected. Only mesial roots were prepared with PTG and TS. Pre‐ and post‐operative scans were performed using Micro‐CT to provide volumes with a voxel size of 20 μm. Volumes, non‐instrumented area, amount of transportation and centering ability in coronal, middle and apical third shaping time and procedural errors were recorded. TS and PTG increased the endodontic volume of 2.98 mm3 (±1.56) and 3.21 mm3 (±1.78) respectively with no statistical difference (p = .168) and no procedural errors. No significant difference was found concerning canal transportation among groups but only within the same group PTG (p value < .001) and TS (p value < .001). The mean centering ratio was significantly different only between the section levels for PTG (p value < .001) and TS (p value = .01); it was significantly reduced in the cervical third. The percentage of untouched canal walls ranged between 29.78% (±15.145) and 36.60% (±11.968) respectively for PTG and TS with no statistical difference among groups (p value = .168). TS and PTG with post machining heat treatment were able to produce centered preparations with no significant difference or procedural errors. TS system provided a shorter preparation time than PTG files.

The resolution of dental computed tomography (CT) images is limited by detector geometry, sensitivity, patient movement, the reconstruction technique and the need to minimize radiation dose. Recently, the use of convolutional neural network (CNN) architectures has shown promise as a resolution enhancement method. In the current work, two CNN architectures—a subpixel network and the so called U-net—have been considered for the resolution enhancement of 2-D cone-beam CT image slices of ex vivo teeth. To do so, a training set of 5680 cross-sectional slices of 13 teeth and a test set of 1824 slices of 4 structurally different teeth were used. Two existing reconstruction-based super-resolution methods using $\boldsymbol {\ell _{2}}$ -norm and total variation regularization were used for comparison. The results were evaluated with different metrics (peak signal-to-noise ratio, structure similarity index, and other objective measures estimating human perception) and subsequent image-segmentation-based analysis. In the evaluation, micro-CT images were used as ground truth. The results suggest the superiority of the proposed CNN-based approaches over reconstruction-based methods in the case of dental CT images, allowing better detection of medically salient features, such as the size, shape, or curvature of the root canal.

We assessed the efficiency of two shaping file systems and two passive ultrasonic irrigation (PUI) devices for removing filling material during retreatment. The mesial canals from 44 extracted mandibular molars were prepared and obturated. The teeth were randomly divided into two groups, and then one group was retreated with Reciproc R25 (VDW, Munich, Germany) (n = 44) and the other group was retreated with 2Shape (TS, Micro Mega, Besançon, France) (n = 44). A micro-computed tomography (CT) scan was taken before and after the retreatment to assess the volume of the filling material remnants. The teeth were then randomly divided into four groups to test two different PUI devices: Irrisafe (Satelec Acteon Group, Merignac, France) and Endo Ultra (Vista Dental Products, Racine, WI, USA). The teeth in Group A were retreated with 2Shape to test the Endo Ultra (n = 22) device, the teeth in Group B were retreated with 2Shape in order to test the Irrisafe (n = 22) device, the teeth in Group C were retreated with Reciproc to test the Endo Ultra (n = 22) device, and Group D was retreated with Reciproc to test the Irrisafe (n = 22) device. A third micro-CT scan was taken after the retreatment to test the PUIs. The percentage of Gutta-Percha (GP) and sealer removed was 94.75% for TS2 (p < 0.001) and 89.3% for R25 (p < 0.001). The PUI significantly enhanced the removal of the filling material by 0.76% for Group A (p < 0.001), 1.47% for Group B (p < 0.001), 2.61% for Group C (p < 0.001), and by 1.66% for Group D (p < 0.001). 2Shape was more effective at removing the GP and sealer during retreatment (p = 0.018). The supplementary approach with PUI significantly improved filling material removal, with no statistical difference between the four groups (p = 0.106).

A volumetric non-blind single image super-resolution technique using tensor factorization has been recently introduced by our group. That method allowed a 2-order-of-magnitude faster high-resolution image reconstruction with equivalent image quality compared to state-of-the-art algorithms. In this work a joint alternating recovery of the high-resolution image and of the unknown point spread function parameters is proposed. The method is evaluated on dental computed to-mography images. The algorithm was compared to an existing 3D super-resolution method using low-rank and total variation regularization, combined with the same alternating PSF-optimization. The two algorithms have shown similar improvement in PSNR, but our method converged roughly 40 times faster, under 6 minutes both in simulation and on experimental dental computed tomography data.

OBJECTIVE:Tooth 3D automatic segmentation (AS) is being actively developed in research and clinical fields. Here, we assess the effect of automatic segmentation using a watershed-based method on the accuracy and reproducibility of 3D reconstructions in volumetric measurements by comparing it with a semi-automatic segmentation(SAS) method that has already been validated.METHODS:The study sample comprised 52 teeth, scanned with micro-CT (41 µm voxel size) and CBCT (76; 200 and 300 µm voxel size). Each tooth was segmented by AS based on a watershed method and by SAS. For all surface reconstructions, volumetric measurements were obtained and analysed statistically. Surfaces were then aligned using the SAS surfaces as the reference. The topography of the geometric discrepancies was displayed by using a colour map allowing the maximum differences to be located.RESULTS:AS reconstructions showed similar tooth volumes when compared with SAS for the 41 µm voxel size. A difference in volumes was observed, and increased with the voxel size for CBCT data. The maximum differences were mainly found at the cervical margins and incisal edges but the general form was preserved.CONCLUSION:Micro-CT, a modality used in dental research, provides data that can be segmented automatically, which is timesaving. AS with CBCT data enables the general form of the region of interest to be displayed. However, our AS method can still be used for metrically reliable measurements in the field of clinical dentistry if some manual refinements are applied.

This paper aims at evaluating the potential of superresolution (SR) image processing to enhance the resolution of Cone Beam Computed Tomography (CBCT) images and to further improve the root canal segmentation in endodontics. First we perform SR based on a linear model, then, we apply an automated segmentation procedure to native and super-resolved CBCT volumes in order to extract the root canal structure. Seven intact extracted teeth have been used to evaluate the potential of SR CBCT in detecting the dental root canal. For all the considered teeth, the SR CBCT volumes provided a smaller error compared to the native CBCT data.

Root canal segmentation on cone beam computed tomography (CBCT) images is difficult because of the noise level, resolution limitations, beam hardening and dental morphological variations. An image processing framework, based on an adaptive local threshold method, was evaluated on CBCT images acquired on extracted teeth. A comparison with high quality segmented endodontic images on micro computed tomography (µCT) images acquired from the same teeth was carried out using a dedicated registration process. Each segmented tooth was evaluated according to volume and root canal sections through the area and the Feret's diameter. The proposed method is shown to overcome the limitations of CBCT and to provide an automated and adaptive complete endodontic segmentation. Despite a slight underestimation (−4, 08%), the local threshold segmentation method based on edge-detection was shown to be fast and accurate. Strong correlations between CBCT and µCT segmentations were found both for the root canal area and diameter (respectively 0.98 and 0.88). Our findings suggest that combining CBCT imaging with this image processing framework may benefit experimental endodontology, teaching and could represent a first development step towards the clinical use of endodontic CBCT segmentation during pulp cavity treatment.

Aim: The aim of this study was to assess whether an obturation, combining a custom guttapercha cone with the BIOROOTTM-RCS sealer, displays similar sealing quality to the orthograde apical plugs of MTA CAPS1 in immature teeth with irregular wide apices. Methodology: Thirty-four immature permanent premolars with apical diameter varying between (1-3 mm) were chosen for this study and were divided into two groups. They were imbedded in wet sponge, which simulated the periapex. In the first group; 5 mm orthograde plugs of MTAwere placed using an appropriate plugger. In the second group; a custom gutta-percha cone was fabricated and used for root canal filling with the BIOROOTTM-RCS sealer. The specimens were stored at 37 8C and 100% humidity during five weeks to allow the complete set of the filling materials. The apical leakage was evaluated using a dye penetration test with 50%-weight silver-nitrate. The teeth were then embedded in a transparent resin and sectioned transversally at 1 and 3 mm from the apex. The slices were examined under optical microscope and were given scores from (0) to (4). When scoring a slice was difficult, spectroscopy for energy dispersion using a scanning electron-microscope was used to confirm the score. The results were compared using the Fisher test with p < 0.05. Results: Silver-nitrate was found in both groups in all slices at 1 mm. At 3 mm, the difference of micro-leakage was not significant. Conclusions: The custom gutta-percha cone combined with BIOROOTtm-RCS sealer displays similar leakage resistance to the orthograde MTA plugs.

Background The aim was to compare the efficacy of the passive ultrasonic irrigation PUI and the Xp-endo Finisher (FKG-Dentaire, La-Chaux-de-Fonds, Switzerland) in removing the calcium hydroxide paste from root canals and from the apical third. Material and Methods Sixty-eight root canals of single-rooted teeth were shaped using the BT-Race files (FKG-Dentaire, La-Chaux-de-Fonds, Switzerland). Ca(OH)2 was placed in all samples except for the negative control group (n=4). Remaining teeth were randomly divided into three groups: G1-Xp (n=30), G2-PUI (n=30) and the positive control group (n=4). Removal procedure consisted of three repeated one-minute-cycles. Samples were split longitudinally, photos of halves were taken at X6.4 magnification and were analyzed using the ImageJ-Software (The National Institutes of Health NIH, Bethesda, Maryland, USA) to calculate the percentage of surfaces with residual Ca(OH)2; the results were compared using the Wilcoxon-Mann Whitney test. Photos of the apical thirds were taken at X16 and X40 magnifications and were scored by two examiners from (0) to (4). Scores of the apical third were compared using the Fisher test. Results The Xp-endo Finisher removed completely the Ca(OH)2 dressing from four teeth (13.33%) whereas the PUI in one tooth (3.33%). The mean values of the remaining Ca(OH)2 were (2.1%, 3.6%) respectively and the difference was not significant (p= 0.195). Both examiners found the Xp-endo Finisher more efficient in the apical third and the difference was significant; p= (0.025, 0.047) respectively. Conclusions The Xp-endo Finisher showed a superiority over the PUI in removing the Ca(OH)2 from the apical third after 3 minutes of activation. Key words:Calcium hydroxide removal, Passive ultrasonic irrigation, Xp-endo Finisher.

Validation of image processing techniques such as endodontic segmentations in cone-beam computed tomography (CBCT) is a challenging issue because of the lack of ground truth in in vivo experiments. The purpose of our study was to design an artificial surrounding tissues phantom able to provide CBCT image quality of real extracted teeth, similar to in vivo conditions. Note that these extracted teeth could be previously scanned using micro computed tomography (μCT) to access true quantitative measurements of the root canal anatomy. Different design settings are assessed in our study by comparison to in vivo images, in terms of the contrast-to-noise ratio (CNR) obtained between different anatomical structures. Concerning the root canal and the dentine, the best design setup allowed our phantom to provide a CNR difference of only 3% compared to clinical cases.

OBJECTIVES:To determine the optimal CBCT settings for an automatic edge-detection-based endodontic segmentation procedure by assessing the accuracy of the root canal measurements.METHODS:12 intact teeth with closed apexes were cut perpendicular to the root axis, at pre-determined levels to the reference plane (the first section made before acquisition). Acquisitions of each specimen were performed with Kodak 9000(®) 3D (76 µm, 14 bits; Kodak Carestream Health, Trophy, France) by using different combinations of milliamperes and kilovolts. Three-dimensional images were displayed and root canals were segmented with the MeVisLab software (edge-detection-based method; MeVis Research, Bremen, Germany). Histological root canal sections were then digitized with a 0.5- to 1.0-µm resolution and compared with equivalent two-dimensional cone-beam reconstructions for each pair of settings using the Pearson correlation coefficient, regression analysis and Bland-Altman method for the canal area and Feret's diameter. After a ranking process, a Wilcoxon paired test was carried out to compare the pair of settings.RESULTS:The best pair of acquisition settings was 3.2 mA/60 kV. Significant differences were found between 3.2 mA/60 kV and other settings (p < 0.05) for the root canal area and for Feret's diameter.CONCLUSIONS:The quantitative analyses of the root canal system with the edge-detection-based method could depend on acquisition parameters. Improvements in segmentation still need to be carried out to ensure the quality of the reconstructions when we have to deal with closer outlines and because of the low spatial resolution.