The effect of threshold level on bone segmentation of cranial base structures from CT and CBCT images

Deep learning segmentation-based bone removal from computed tomography of the brain improves subdural hematoma detection

PURPOSE

Timely identification of intracranial blood products is clinically impactful, however the detection of subdural hematoma (SDH) on non-contrast CT scans of the head (NCCTH) is challenging given interference from the adjacent calvarium. This work explores the utility of a NCCTH bone removal algorithm for improving SDH detection.

METHODS

A deep learning segmentation algorithm was designed/trained for bone removal using 100 NCCTH. Segmentation accuracy was evaluated on 15 NCCTH from the same institution and 22 NCCTH from an independent external dataset using quantitative overlap analysis between automated and expert manual segmentations. The impact of bone removal on detecting SDH by junior radiology trainees was evaluated with a reader study comparing detection performance between matched cases with and without bone removal applied.

RESULTS

Average Dice overlap between automated and manual segmentations from the internal and external test datasets were 0.9999 and 0.9957, which was superior to other publicly available methods. Among trainee readers, SDH detection was statistically improved using NCCTH with and without bone removal applied compared to standard NCCTH alone (P value <0.001). Additionally, 12/14 (86 %) of participating trainees self-reported improved detection of extra axial blood products with bone removal, and 13/14 (93 %) indicated that they would like to have access to NCCTH bone removal in the on-call setting.

CONCLUSION

Deep learning segmentation-based NCCTH bone removal is rapid, accurate, and improves detection of SDH among trainee radiologists when used in combination with standard NCCTH. This study highlights the potential of bone removal for improving confidence and accuracy of SDH detection.

Evaluation of CBCT reconstructed tooth models at different thresholds and voxels and their accuracy in fusion with IOS data: an in vitro validation study

BACKGROUND

This study aims to evaluate the impact of different thresholds and voxel sizes on the accuracy of Cone-beam computed tomography (CBCT) tooth reconstruction and to assess the accuracy of fused CBCT and intraoral scanning (IOS) tooth models using curvature continuity algorithms under varying thresholds and voxel conditions.

METHODS

Thirty-two isolated teeth were digitized using IOS and CBCT at two voxel sizes and five threshold settings. Crown-root fusion was performed using a curvature continuity algorithm. Volume, surface area, and crown width of tooth models were compared to laser scanning models, and RMS error was measured. Data were analyzed using Wilcoxon signed-rank test, paired t-test, and one-way ANOVA.

RESULTS

Volume amplification errors of CBCT with 0.15 mm and 0.3 mm voxels ranged from 1.22 to 19.07%, surface area errors from 0.18 to 7.78%, crown linearity errors ranged from 2.47 to 7.69%, root linearity errors ranged from - 1.02 to 2.26% and RMS from 0.0691 mm to 0.2408 mm. Crown-root fusion of IOS and CBCT data reduced volume error to -0.90-5.10%, surface area error to -0.66-4.15%, and RMS to 0.0359 mm to 0.0945 mm.

CONCLUSIONS

Voxel size and threshold settings significantly affect the accuracy of CBCT reconstruction and crown-root fusion. Smaller voxel sizes yield higher reconstruction precision, and different voxel sizes and tooth regions correspond to distinct optimal segmentation thresholds. The validated semi-automated crown-root fusion algorithm significantly enhances overall model accuracy, offering new possibilities for clinical applications.

Insights into geometric deviations of medical 3d-printing: a phantom study utilizing error propagation analysis

BACKGROUND

The use of 3D-printing in medicine requires a context-specific quality assurance program to ensure patient safety. The process of medical 3D-printing involves several steps, each of which might be prone to its own set of errors. The segmentation error (SegE), the digital editing error (DEE) and the printing error (PrE) are the most important partial errors. Approaches to evaluate these have not yet been implemented in a joint concept. Consequently, information on the stability of the overall process is often lacking and possible process optimizations are difficult to implement. In this study, SegE, DEE, and PrE are evaluated individually, and error propagation is used to examine the cumulative effect of the partial errors.

METHODS

The partial errors were analyzed employing surface deviation analyses. The effects of slice thickness, kernel, threshold, software and printers were investigated. The total error was calculated as the sum of SegE, DEE and PrE.

RESULTS

The higher the threshold value was chosen, the smaller were the segmentation results. The deviation values varied more when the CT slices were thicker and when the threshold was more distant from a value of around -400 HU. Bone kernel-based segmentations were prone to artifact formation. The relative reduction in STL file size [as a proy for model complexity] was greater for higher levels of smoothing and thinner slice thickness of the DICOM datasets. The slice thickness had a minor effect on the surface deviation caused by smoothing, but it was affected by the level of smoothing. The PrE was mainly influenced by the adhesion of the printed part to the build plate. Based on the experiments, the total error was calculated for an optimal and a worst-case parameter configuration. Deviations of 0.0093 mm ± 0.2265 mm and 0.3494 mm ± 0.8001 mm were calculated for the total error.

CONCLUSIONS

Various parameters affecting geometric deviations in medical 3D-printing were analyzed. Especially, soft reconstruction kernels seem to be advantageous for segmentation. The concept of error propagation can contribute to a better understanding of the process specific errors and enable future analytical approaches to calculate the total error based on process parameters.

Image Quality Assessment of a Deep Learning-Based Automatic Bone Removal Algorithm for Cervical CTA

BACKGROUND

The present study aims to evaluate the postprocessing image quality of a deep-learning (DL)-based automatic bone removal algorithm in the real clinical practice for cervical computed tomography angiography (CTA).

MATERIALS AND METHODS

A total of 100 patients (31 females, 61.4 ± 12.4 years old) who had performed cervical CTA from January 2022 to July 2022 were included retrospectively. Three different types of scanners were used. Ipsilateral cervical artery was divided into 10 segments. The performance of the DL algorithm and conventional algorithm in terms of bone removal and vascular integrity was independently evaluated by two radiologists for each segment. The difference in the performance between the two algorithms was compared. Inter- and intrarater consistency were assessed, and the correlation between the degree of carotid artery stenosis and the rank of bone removal and vascular integrity was analyzed.

RESULTS

Significant differences were observed in the rankings of bone removal and vascular integrity between the two algorithms on most segments on both sides. Compared to DL algorithm, the conventional algorithm showed a higher correlation between the degree of carotid artery stenosis and vascular integrity ( r = -0.264 vs r = -0.180). The inter- and intrarater consistency of DL algorithm were found to be higher than or equal to those of conventional algorithm.

CONCLUSIONS

The DL algorithm for bone removal in cervical CTA demonstrated significantly better performance than conventional postprocessing method, particularly in the segments with complex anatomical structures and adjacent to bone.

The Reproducibility of Reference Landmarks in the External Acoustic Meatus (EAM) on Cone Beam Computed Tomography (CBCT) Images

Objective: The aim of the present study is to identify a more reliable reference point in three-dimensional cephalometric analysis to replace the Porion point used in two-dimensional analysis, enhancing the accuracy of assessments. Methods: The methodology assessed potential alternative landmarks for three-dimensional cephalometric analysis. Utilizing a segmenting technique, anatomical landmarks were accurately pinpointed from the external acoustic meatus of 26 Cone Beam Computed Tomography (CBCT) scans. These landmarks were chosen for their clear and unambiguous detectability. To assess reproducibility, each landmark was replicated twice with a one-week interval by a master's student. Reproducibility was quantitatively evaluated by analyzing the absolute difference per axis. Results: Five possible candidate landmarks were identified: the most anterior, posterior, superior, and inferior points of the external acoustic meatus (EAM) and a notch delineating the epitympanic recess. The reproducibility of pinpointing these landmarks ranged from 0.56 mm to 2.2 mm. The absolute mean differences between measurements were 0.46 mm (SD 0.75) for the most anterior point, 0.36 mm (SD 0.44) for the most posterior point, 0.25 mm (SD 0.26) for the most superior point, 1.11 mm (SD 1.03) for the most inferior point, and 0.78 mm (SD 0.57) for the epitympanic notch. Conclusions: The most superior point of the EAM might successfully replace the Porion as an anatomical reference.

Accuracy of facial skeletal surfaces segmented from CT and CBCT radiographs

The accuracy of three-dimensional (3D) facial skeletal surface models derived from radiographic volumes has not been extensively investigated yet. For this, ten human dry skulls were scanned with two Cone Beam Computed Tomography (CBCT) units, a CT unit, and a highly accurate optical surface scanner that provided the true reference models. Water-filled head shells were used for soft tissue simulation during radiographic imaging. The 3D surface models that were repeatedly segmented from the radiographic volumes through a single-threshold approach were used for reproducibility testing. Additionally, they were compared to the true reference model for trueness measurement. Comparisons were performed through 3D surface approximation techniques, using an iterative closest point algorithm. Differences between surface models were assessed through the calculation of mean absolute distances (MAD) between corresponding surfaces and through visual inspection of facial surface colour-coded distance maps. There was very high reproducibility (approximately 0.07 mm) and trueness (0.12 mm on average, with deviations extending locally to 0.5 mm), and no difference between radiographic scanners or settings. The present findings establish the validity of lower radiation CBCT imaging protocols at a similar level to the conventional CT images, when 3D surface models are required for the assessment of facial morphology.

The impact of teeth and dental restorations on gray value distribution in cone-beam computer tomography: a pilot study

PURPOSE

To investigate the influence of teeth and dental restorations on the facial skeleton's gray value distributions in cone-beam computed tomography (CBCT).

METHODS

Gray value selection for the upper and lower jaw segmentation was performed in 40 patients. In total, CBCT data of 20 maxillae and 20 mandibles, ten partial edentulous and ten fully edentulous in each jaw, respectively, were evaluated using two different gray value selection procedures: manual lower threshold selection and automated lower threshold selection. Two sample t tests, linear regression models, linear mixed models, and Pearson's correlation coefficients were computed to evaluate the influence of teeth, dental restorations, and threshold selection procedures on gray value distributions.

RESULTS

Manual threshold selection resulted in significantly different gray values in the fully and partially edentulous mandible. (p = 0.015, difference 123). In automated threshold selection, only tendencies to different gray values in fully edentulous compared to partially edentulous jaws were observed (difference: 58-75). Significantly different gray values were evaluated for threshold selection approaches, independent of the dental situation of the analyzed jaw. No significant correlation between the number of teeth and gray values was assessed, but a trend towards higher gray values in patients with more teeth was noted.

CONCLUSIONS

Standard gray values derived from CT imaging do not apply for threshold-based bone segmentation in CBCT. Teeth influence gray values and segmentation results. Inaccurate bone segmentation may result in ill-fitting surgical guides produced on CBCT data and misinterpreting bone density, which is crucial for selecting surgical protocols. Created with BioRender.com.

Three-dimensional Reconstruction and Comparison of Temporomandibular Joint Space Volume Before and After Orthognathic Surgery in Patients with Skeletal Class III Malocclusion With Mandibular Deviation

OBJECTIVE

To look into the association between the degree of deviation and the changing trend in the temporomandibular joint (TMJ) space volume after orthognathic surgery in patients with skeletal Class III malocclusion.

METHODS

Twenty patients having combined orthodontic-orthognathic treatment for skeletal Class III malocclusions with mandibular deviation were chosen, and craniofacial spiral CT was performed before (T0), two weeks after (T1), and six months after (T2) surgery. Using 3D volume reconstruction, further partitioning, and analysis of each domain's volume changes over time, the TMJ space volume is to be obtained. The differences in changes between groups A (mild deviation group) and B (severe deviation group) were examined to examine the impact of the degree of deviation on the TMJ space volume.

RESULTS

A statistically significant difference ( P <0.05) existed between the postoperative TMJ space volume in group A and the preoperative overall, anterolateral, and anteroinferior space volume; the same difference also existed between the postoperative TMJ space volume in the NDS and the preoperative posterolateral, posteroinferior space volume. In group B, the postoperative TMJ space volume was statistically significant ( P <0.05) compared with the preoperative total and anteroinferior space volume in the DS; the difference between the total volume of the T1 stage on the NDS and the total volume of the T0 stage was statistically significant ( P <0.05). The two groups showed substantial differences in the space volume changes between the T1-T0 phase and the T2-T1 period.

CONCLUSION

Patients with skeletal Class III malocclusion and mandibular deviation after orthognathic surgery see a change in the TMJ space volume. All patient types experience a largely consistent space volume change trend two weeks after surgery, and the degree of mandibular deviation is correlated with the intensity and longevity of the alteration.

A Deep Learning Approach for Automated Bone Removal from Computed Tomography Angiography of the Brain

Advanced visualization techniques such as maximum intensity projection (MIP) and volume rendering (VR) are useful for evaluating neurovascular anatomy on CT angiography (CTA) of the brain; however, interference from surrounding osseous anatomy is common. Existing methods for removing bone from CTA images are limited in scope and/or accuracy, particularly at the skull base. We present a new brain CTA bone removal tool, which addresses many of these limitations. A deep convolutional neural network was designed and trained for bone removal using 72 brain CTAs. The model was tested on 15 CTAs from the same data source and 17 CTAs from an independent external dataset. Bone removal accuracy was assessed quantitatively, by comparing automated segmentation results to manual segmentations, and qualitatively by evaluating VR visualization of the carotid siphons compared to an existing method for automated bone removal. Average Dice overlap between automated and manual segmentations from the internal and external test datasets were 0.986 and 0.979 respectively. This was superior compared to a publicly available noncontrast head CT bone removal algorithm which had a Dice overlap of 0.947 (internal dataset) and 0.938 (external dataset). Our algorithm yielded better VR visualization of the carotid siphons than the publicly available bone removal tool in 14 out of 15 CTAs (93%, chi-square statistic of 22.5, p-value of < 0.00001) from the internal test dataset and 15 out of 17 CTAs (88%, chi-square statistic of 23.1, p-value of < 0.00001) from the external test dataset. Bone removal allowed subjectively superior MIP and VR visualization of vascular anatomy/pathology. The proposed brain CTA bone removal algorithm is rapid, accurate, and allows superior visualization of vascular anatomy and pathology compared to other available techniques and was validated on an independent external dataset.

Three-dimensional maxillary virtual patient creation by convolutional neural network-based segmentation on cone-beam computed tomography images

OBJECTIVE

To qualitatively and quantitatively assess integrated segmentation of three convolutional neural network (CNN) models for the creation of a maxillary virtual patient (MVP) from cone-beam computed tomography (CBCT) images.

MATERIALS AND METHODS

A dataset of 40 CBCT scans acquired with different scanning parameters was selected. Three previously validated individual CNN models were integrated to achieve a combined segmentation of maxillary complex, maxillary sinuses, and upper dentition. Two experts performed a qualitative assessment, scoring-integrated segmentations from 0 to 10 based on the number of required refinements. Furthermore, experts executed refinements, allowing performance comparison between integrated automated segmentation (AS) and refined segmentation (RS) models. Inter-observer consistency of the refinements and the time needed to create a full-resolution automatic segmentation were calculated.

RESULTS

From the dataset, 85% scored 7-10, and 15% were within 3-6. The average time required for automated segmentation was 1.7 min. Performance metrics indicated an excellent overlap between automatic and refined segmentation with a dice similarity coefficient (DSC) of 99.3%. High inter-observer consistency of refinements was observed, with a 95% Hausdorff distance (HD) of 0.045 mm.

CONCLUSION

The integrated CNN models proved to be fast, accurate, and consistent along with a strong interobserver consistency in creating the MVP.

CLINICAL RELEVANCE

The automated segmentation of these structures simultaneously could act as a valuable tool in clinical orthodontics, implant rehabilitation, and any oral or maxillofacial surgical procedures, where visualization of MVP and its relationship with surrounding structures is a necessity for reaching an accurate diagnosis and patient-specific treatment planning.

Effect of hydration on the anatomical form of human dry skulls

In radiology research soft tissues are often simulated on bone specimens using liquid materials such as water, or gel-like materials, such as ballistic gel. This study aimed to test the effect of hydration on the anatomical form of dry craniofacial bone specimens. Sixteen human dry skulls and 16 mandibles were scanned with an industrial scanner in dry conditions and after water embedding. Ten skulls were also embedded for different time periods (5 or 15 min). The subsequent 3D surface models were best-fit superimposed and compared by calculating mean absolute distances between them at various measurement areas. There was a significant, primarily enlargement effect of hydration on the anatomical form of dry skeletal specimens as detected after water embedding for a short time period. The effect was smaller in dry skulls (median 0.20 mm, IQR 0.17 mm) and larger in mandibles (median 0.56 mm, IQR 0.57 mm). The effect of different water embedding times was negligible. Based on the present findings, we suggest to shortly hydrate the skeletal specimens prior to reference model acquisition so that they are comparable to hydrated specimens when liquid materials are used as soft-tissue simulants for various radiologic research purposes.

Facial changes in patients with skeletal class III deformity after bimaxillary surgery: an evaluation based on three-dimensional photographs registered with computed tomography

The objective of this study was to evaluate facial soft and hard tissue changes, individually and relative to each other, in patients with skeletal class III deformity after bimaxillary surgery using three-dimensional (3D) photos obtained by white light scanning. Thirty patients with skeletal class III deformity who underwent bimaxillary surgery were selected. Each patient underwent white light scanning and spiral computed tomography (CT) within two weeks before (T0) and six months after surgery (T1). The 3D photos were registered with CT soft tissue models for T0 and T1, and the skeletal area unaffected by treatment (cranial base) was used to register T0 and T1. Then, the 3D colour-coded map was analysed to assess both skeletal and soft tissue changes between T0 and T1. Changes in the 3D coordinates of each anatomical landmark were analysed using the Student's t-test. Maxillary advancement by 2-3 mm and mandibular recession by 5-6 mm were observed; the mandible was shortened in the vertical direction. Compared with the preoperative values, the nasal columella was 0.51 mm shorter, the upper lip was 0.71 mm longer, the base of the alar cartilage was 1.38 mm wider, and the nasolabial angle became larger. The ratio of change in the position of soft tissue point Sn to hard tissue point A was 0.73:1, and that of soft tissue point Pg to hard tissue point Pog was 0.86:1. Images obtained by structured white light scanning registered with CT can be used as an alternative to study facial changes after orthognathic surgery.

Using the anterior cranial base to provide a reliable reference plane for patients with or without facial asymmetry

INTRODUCTION

This study aimed to investigate the midsagittal reference plane (MSP) reliability derived from the 3-dimensional characteristics of patients with or without facial asymmetry in the anterior cranial base (ACB).

METHODS

We divided the cone-beam computed tomography (CBCT) images of 60 adult patients into maxillofacial symmetry and asymmetry groups. The ACB models were 3-dimensionally constructed, and then symmetrical characteristics were evaluated with surface asymmetry for each group. The reliability of the MSP derived from the symmetry of the anterior cranial base (MSPACB) was assessed in comparison with the true craniofacial symmetry plane determined using the morphometric method.

RESULTS

The ACB was symmetrical, as demonstrated by slight surface asymmetry. The MSPACB was reliable for maxillofacial asymmetrical analysis as the intraobserver and interobserver measurements using the MSPACB were of excellent agreement, and there was no significant difference between MSPACB and morphometric method in asymmetrical measurements in both groups. The MSPACB remained stable (maximum deviation <0.32 mm) when cranial landmark identification errors (1 mm and 4 mm) were simulated.

CONCLUSIONS

MSPACB is reliable for patients with or without facial asymmetry in maxillofacial asymmetry analysis, which is beneficial to patients with severe midfacial asymmetry or trauma when conventional landmarks are displaced or disappear. When using MSPACB for patients with cranial malformations or those whose ACBs differ from normal dimensions, caution should be taken.

Precision of a Hand-Held 3D Surface Scanner in Dry and Wet Skeletal Surfaces: An Ex Vivo Study

Three-dimensional surface scans of skeletal structures have various clinical and research applications in medicine, anthropology, and other relevant fields. The aim of this study was to test the precision of a widely used hand-held surface scanner and the associated software's 3D model generation-error in both dry and wet skeletal surfaces. Ten human dry skulls and ten mandibles (dry and wet conditions) were scanned twice with an industrial scanner (Artec Space Spider) by one operator. Following a best-fit superimposition of corresponding surface model pairs, the mean absolute distance (MAD) between them was calculated on ten anatomical regions on the skulls and six on the mandibles. The software's 3D model generation process was repeated for the same scan of four dry skulls and four mandibles (wet and dry conditions), and the results were compared in a similar manner. The median scanner precision was 31 μm for the skulls and 25 μm for the mandibles in dry conditions, whereas in wet conditions it was slightly lower at 40 μm for the mandibles. The 3D model generation-error was negligible (range: 5-10 μm). The Artec Space Spider scanner exhibits very high precision in the scanning of dry and wet skeletal surfaces.

Novel Anterior Cranial Base Area for Voxel-Based Superimposition of Craniofacial CBCTs

A standard method to assess changes in craniofacial morphology over time is through the superimposition of serial patient images. This study evaluated the reliability of a novel anterior cranial base reference area, principally including stable midline structures (EMACB) after an early age, and compared it to the total anterior cranial base (TACB) and an area including only midline structures (MACB). Fifteen pairs of pre-existing serial CBCT images acquired from growing patients were superimposed with all techniques by applying a best-fit registration algorithm of corresponding voxel intensities (Dolphin 3D software). The research outcomes were the reproducibility of each technique and the agreement between them in skeletal change detection, as well as their validity. The TACB and EMACB methods were valid, since the superimposed midline ACB structures consistently showed adequate overlap. They also presented perfect overall reproducibility (median error < 0.01 mm) and agreement (median difference < 0.01 mm). MACB showed reduced validity, higher errors, and a moderate agreement to the TACB. Thus, the EMACB method performed efficiently and mainly included the stable midline ACB structures during growth. Based on the technical, anatomical, and biological principles applied when superimposing serial 3D data to assess craniofacial changes, we recommend the EMACB method as the method of choice to fulfil this purpose.

Diagnosis of Heart Failure Complicated with Sleep Apnea Syndrome by Thoracic Computerized Tomography under Artificial Intelligence Algorithm

The aim of this study was to explore the application effect of thoracic computerized tomography (CT) under single threshold segmentation algorithm in the diagnosis of heart failure (HF) complicated with sleep apnea syndrome. 30 patients diagnosed with HF complicated with sleep apnea syndrome were chosen for the research. Another 30 patients without sleep apnea syndrome were selected as the control group, whose age, height, and weight were similar to those of the experimental group. Then, a model for thoracic CT image segmentation was proposed under the single threshold segmentation algorithm, and the faster region convolutional neural network (Faster RCNN) was applied to label the thoracic respiratory lesions. All the patients underwent thoracic CT examination, and the obtained images were processed using the algorithm model above. After that, the morphology of the patient's respiratory tract after treatment was observed. The results suggested that the improved single threshold segmentation algorithm was effective for the image segmentation of patient lesions, and the Faster RCNN could effectively finish the labeling of the lesion area in the CT image. The classification accuracy of the Faster RCNN was about 0.966, and the loss value was about 0.092. With CT scanning under the algorithm, it was found that the airway collapse of the posterior palatal area, retrolingual area, and laryngopharyngeal area of the sleep apnea syndrome patients was significantly greater than that of the control group (P < 0.05). But there was no significant difference of the collapse of the nasopharyngeal area between the two groups (P > 0.05). The single threshold segmentation algorithm had a better segmentation accuracy for thoracic CT images in patients with HF and sleep apnea syndrome, so it had a highly promising application prospect in the diagnosis of the disease.

Reliability of Different Anterior Cranial Base Reference Areas for Voxel-Based Superimposition

The study aimed to evaluate the reliability and reproducibility and compare the outcomes of two 3D voxel-based superimposition techniques for craniofacial CBCT images, using anterior cranial base areas of different extent as references. Fifteen preexisting pairs of serial CBCTs (initial age: 11.7 ± 0.6 years; interval: 1.7 ± 0.4 years) were superimposed on total anterior cranial base (TACB) or middle anterior cranial base (MACB) structures through the Dolphin 3D software. The overlap of the reference structures was assessed visually to indicate reliability. All superimpositions were repeated by the same investigator. Outcomes were compared to assess the agreement between the two methods. Reliability was perfect for the TACB and moderate for the MACB method (p = 0.044). Both areas showed good overall reproducibility, though in individual cases there were notable differences for MACB superimpositions, ranging from −1.84 to 1.64 mm (TACB range: −0.48 to 0.31 mm). The overall agreement in the detected T0/T1 changes was also good, though it was significantly reduced for individual measurements (median < 0.01 mm, IQR: 0.46 mm, range: −2.81 to 0.73 mm). In conclusion, the voxel-based superimposition on TACB was more reliable and showed higher reproducibility than the superimposition on MACB. Thus, the extended anterior cranial base area is recommended for the assessment of craniofacial changes.

Superimposition of serial 3-dimensional facial photographs to assess changes over time: A systematic review

INTRODUCTION

Superimpositions of 3-dimensional photographs enable a thorough and risk-free assessment of facial changes over time. However, the available methods and the evidence supporting them have not been assessed systematically. The paper summarizes and assesses the current evidence on superimposition methods of serial 3-dimensional facial photographs available in the literature.

METHODS

The following databases were searched without time restriction (last updated December 2020): MEDLINE via PubMed, EMBASE, Cochrane Library, and Google Scholar. Unpublished literature was searched on Open Grey and Grey Literature Report. Authors were contacted if necessary, and reference lists of relevant papers were screened. All studies with sample size ≥6 that tested the accuracy or precision of a superimposition technique, or agreement between different techniques regarding facial surface changes, were considered. The 2 authors performed data extraction independently using predefined forms. The risk of bias was assessed through the Quality Assessment and Diagnostic Accuracy Tool 2 tool.

RESULTS

Eight studies fulfilled the inclusion criteria. The total risk of bias of 7 studies was high and of 1 low. Seven studies had high total applicability concerns, and 1 was unclear. There was high heterogeneity among studies, which tested constructed planes through manually selected landmarks, a configuration of 9 landmarks, various surface areas, and the entire facial surface as superimposition references. A small rectangular area on the forehead combined with one on the middle part of the nose and the lower wall of the orbital foramen showed promising results.

CONCLUSIONS

The limited available evidence suggests that surface-based registration is superior to landmark-based registration. Further research in the field is mandatory.

Marginal fit of 3-unit CAD-CAM zirconia frameworks fabricated using cone beam computed tomography scans: an experimental study

Whether cone beam computed tomography (CBCT) scans can be used for the fabrication of computer-aided design and computer-aided manufacturing (CAD-CAM) fixed dental prostheses (FDPs) is not known. The purpose of the present study was to compare the marginal fit of 3-unit zirconia FDPs fabricated by using CBCT or 3-dimensional (3D) laboratory scanning. Extracted second premolar and molar teeth in a maxillary typodont model were prepared. The first molar was removed and the typodont model was scanned with a laboratory or a CBCT scanner to generate two virtual 3D cast groups (3DL and CBCT). Forty four 3-unit zirconia FDPs were designed on virtual casts and milled. The vertical marginal discrepancy (VMD) was measured by ×100-magnification microscopy at seven locations on each abutment. A total of 616 measurements were made at 14 fixed locations in two groups of 22 specimens. The VMD data for 3DL and CBCT groups were statistically analyzed using the Mann-Whitney U test (α = 0.05). The mean VMDs on premolar ranged between 44 and 55 µm (median: 43-55 µm) in 3DL, and 74 and 100 µm (median: 72-93 µm) in CBCT; and on the molar, between 47 and 114 µm (median: 46-114 µm) in 3DL, and 91 and 162 µm (median: 93-156 µm) in CBCT. There was a significant difference between the gaps in 3DL and CBCT groups (p < 0.001). FDPs fabricated using 3D laboratory scanner had significantly smaller VMDs. Nevertheless, the 3-unit zirconia FDPs fabricated using CBCT scans presented promising marginal integrity.

Valid 3D surface superimposition references to assess facial changes during growth

Currently, the primary techniques applied for the assessment of facial changes over time utilize 2D images. However, this approach has important limitations related to the dimensional reduction and the accuracy of the used data. 3D facial photography has been recently introduced as a risk-free alternative that overcomes these limitations. However, the proper reference areas that should be used to superimpose serial 3D facial images of growing individuals are not yet known. Here, we tested various 3D facial photo superimposition reference areas and compared their outcomes to those of a standard anterior cranial base superimposition technique. We found that a small rectangular area on the forehead plus an area including the middle part of the nose and the lower wall of the orbital foramen provided comparable results to the standard technique and showed adequate reproducibility. Other reference areas that have been used so far in the literature were less reliable. Within the limitations of the study, a valid superimposition reference area for serial 3D facial images of growing individuals is suggested. The method has potential to greatly expand the possibilities of this highly informative, risk free, and easily obtained 3D tool for the assessment of facial changes in growing individuals.

The clinical study of CBCT imaging technology in the restoration of upper anterior teeth of the elderly

OBJECTIVE

To evaluate and analyze the clinical effect of CBCT imaging technology on the restoration of upper anterior teeth of the elderly.

METHODS

36 elderly patients with upper anterior teeth loss in our hospital from January 2018 to January 2020 were selected for implant restoration. Patients were equally randomized into a curved tomographic restoration group (TR group) and a CBCT restoration group (CR group). Patients in the two groups underwent traditional implant restoration. Then we compared and analyzed the implant migration, the adjustment time of first wearing, and the success rate of axial gingival recession and restoration satisfaction of patients in the two groups.

RESULTS

The neck offset and the root offset of the implants in the CR group was (0.77±0.15) mm and (0.83±0.17) mm, respectively, which were significantly lower than (1.25±0.27) mm and (1.73±0.29) mm in the TR group (t=6.593, t=11.359, all P<0.01). The initial wearing adjustment time of patients in the CR group [(8.73±1.94) min] was significantly less than (18.79±4.85) min in the TR group (t=8.171, P<0.01); the CR group had a significantly higher success rate of axial gingival recession as compared to the CR group (94.44% vs 61.11%, χ²=6.0857, P<0.05); The restoration satisfaction rate of patients in the CR group was 100%, which was significantly higher than 77.78% of the TR group (χ²=8.7429, P<0.05).

CONCLUSION

The CBCT imaging technology has a significant clinical effect on the restoration of the upper anterior teeth of the elderly, which effectively reduces the deviation of implant placement, shorten the adjustment time of their initial wearing, and greatly improves the success rate of axial gingival recession, effectively guarantees the long-term stability and aesthetics of dental implant restoration, and significantly enhances the satisfaction of patients.

3D Occlusal Tooth Wear Assessment in Presence of Limited Changes in Non-Occlusal Surfaces

The study aimed to develop an accurate and convenient 3D occlusal tooth wear assessment technique, applicable when surfaces other than the occlusal undergo changes during the observation period. Various degrees of occlusal tooth wear were simulated in vitro on 18 molar and 18 premolar plaster teeth. Additionally, their buccal and lingual surfaces were gently grinded to induce superficial changes and digital dental models were generated. The grinded and the original tooth crowns were superimposed using six different 3D techniques (two reference areas with varying settings; gold standard: GS). Superimposition on intact structures provided the GS measurements. Tooth wear volume comprised the primary outcome measure. All techniques differed significantly to each other in their accuracy (p < 0.001). The technique of choice (CCD: complete crown with 30% estimated overlap of meshes) showed excellent agreement with the GS technique (median difference: 0.045, max: 0.219 mm³), no systematic error and sufficient reproducibility (max difference < 0.040 mm³). Tooth type, tooth alignment in the dental arches, and amount of tooth wear did not significantly affect the results of the CCD technique (p > 0.01). The suggested occlusal tooth wear assessment technique is straightforward and offers accurate outcomes when limited morphological changes occur on surfaces other than the occlusal.

Artificial intelligence-enabled automatic segmentation of skull CT facilitates computer-assisted craniomaxillofacial surgery

BACKGROUND

The image segmentation of skull CT is the cornerstone for the computer-assisted craniomaxillofacial surgery in multiple aspects. This study aims to introduce an AI-enabled automatic segmentation and propose its prospect in facilitating the computer-assisted surgery.

METHODS

Three patients enrolled in a clinical trial of computer-assisted craniomaxillofacial surgery were randomly selected for this study. The preoperative helical CT scans of the head and neck region were subjected to the AI-enabled automatic segmentation in Mimics Viewer. The performance of AI segmentation was evaluated based on the requirements of computer-assisted surgery.

RESULTS

All three patients were successfully segmented by the AI-enabled automatic segmentation. The performance of AI segmentation was excellent regarding key anatomical structures. The overall quality of bone surface was satisfying. The median DICE coefficient was 92.4% for the maxilla, and 94.9% for the mandible, which fulfilled the requirements of computer-assisted craniomaxillofacial surgery.

CONCLUSIONS

The AI-enabled automatic segmentation could facilitate the preoperative virtual planning and postoperative outcome verification, which formed a feedback loop to enhance the current workflow of computer-assisted surgery. More studies are warranted to confirm the robustness of AI segmentation with more cases.

From pixel to image analysis

Over the past 50 years, computer technology has evolved enormously. This has made it possible to carry out radiography in a completely new way, allowing to process X-ray images in an advanced manner and to extract the information from the image data. This article gives an overview of some of the most important developments in dental radiology. These include capturing the images, analysing and interpreting the image information and using the images for 3D reconstruction.

3D Method for Occlusal Tooth Wear Assessment in Presence of Substantial Changes on Other Tooth Surfaces

Early diagnosis and timely management of tooth or dental material wear is imperative to avoid extensive restorations. Previous studies suggested different methods for tooth wear assessment, but no study has developed a three-dimensional (3D) superimposition technique applicable in cases where tooth surfaces, other than the occlusal, undergo extensive morphological changes. Here, we manually grinded plaster incisors and canines to simulate occlusal tooth wear of varying severity in teeth that received a wire retainer bonded on their lingual surfaces, during the assessment period. The corresponding dental casts were scanned using a surface scanner. The modified tooth crowns were best-fit approximated to the original crowns using seven 3D superimposition techniques (two reference areas with varying settings) and the gold standard technique (GS: intact adjacent teeth and alveolar processes as superimposition reference), which provided the true value. Only a specific technique (complete crown with 20% estimated overlap of meshes), which is applicable in actual clinical data, showed perfect agreement with the GS technique in all cases (median difference: −0.002, max absolute difference: 0.178 mm³). The outcomes of the suggested and the GS technique were highly reproducible (max difference < 0.040 mm³). The presented technique offers low cost, convenient, accurate, and risk-free tooth wear assessment.

Palatal rugae positional changes during orthodontic treatment of growing patients

OBJECTIVES

To investigate the anteroposterior and vertical changes of the median rugae area, which is commonly used as dental model superimposition reference, relevant to its underlying skeletal structures.

SETTINGS AND SAMPLE POPULATION

Retrospectively collected pre- and post-treatment cephalometric radiographs and 3D digital dental models of 24 orthodontic patients (age at treatment start: 12.26 ± 0.83 years; assessment period: 2.13 ± 0.68 years) were analysed. All had mild to moderate malocclusions that were treated non-extraction with full fixed appliances.

MATERIAL AND METHODS

The incisive papilla and rugae points were placed on the dental models that were then registered to the cephalometric radiographs. Afterwards, the radiographs were superimposed on Sella, ANS-PNS, and through a maxillary structural method. The vertical and horizontal movements of the papilla and the rugae points, as well as of a central incisor, were measured (Viewbox 4 software).

RESULTS

The incisive papilla and the three rugae points remained stable anteroposteriorly, but moved downwards in the vertical dimension (approximately 1-2 mm), in a similar manner (P > .05). However, the anteroposterior position of the papilla and the first rugae points were affected by changes in anterior tooth position and inclination (P < .05).

CONCLUSION

Both the second and third rugae can be used as superimposition references for tooth movement assessment. The use of the papilla and the first rugae area is not recommended, because they are affected by tooth movement. The outcomes of a palatal superimposition are comparable to those of a maxillary skeletal superimposition in the anteroposterior, but not in the vertical dimension.

Application of additive manufacturing in customized titanium mandibular implants for patients with oral tumors

The application of additive manufacturing (AM) technology has been widely used in various medical fields, including craniomaxillofacial surgery. The aim of the present study was to examine the surgical efficiency and post-operative outcomes of patient-specific titanium mandibular reconstruction using AM. Major steps in directly designing and manufacturing 3D customized titanium implants are discussed. Furthermore, pre-operative preparations, surgical procedures and post-operative treatment outcomes were compared among patients who received mandibular reconstruction using a customized 3D titanium implant, titanium reconstruction plates or vascularized autologous fibular grafting. Use of a customized titanium implant significantly improved surgical efficiency and precision. When compared with mandibular reconstruction using the two conventional approaches, patients who received the customized implant were significantly more satisfied with their facial appearance, and exhibited minimal post-operative complications in the 12-month follow-up period. Patients who underwent mandibular reconstruction using a customized titanium implant displayed improved mandibular contour symmetry, restored occlusal function, normal range of mouth opening and no temporomandibular joint related pain; all complications frequently experienced by patients who undergo conventional approaches of mandibular reconstruction.