Beam Computed Tomography Data Sets Research Articles

CRML-Net: Cross-Modal Reasoning and Multi-Task Learning Network for tooth image segmentation

Data from a single modality may suffer from noise, low contrast, or other imaging limitations that affect the model’s accuracy. Furthermore, due to the limited amount of data, most models trained on single-modality data tend to overfit the training set and perform poorly on out-of-domain data. Therefore, in this paper, we propose a network named Cross-Modal Reasoning and Multi-Task Learning Network (CRML-Net), which combines cross-modal reasoning and multi-task learning, aiming to leverage the complementary information between different modalities and tasks to enhance the model’s generalization ability and accuracy. Specifically, CRML-Net consists of two stages. In the first stage, our network extracts a new morphological information modality from the original image and then performs cross-modal fusion with the original modality image, aiming to leverage the morphological information to enhance the model’s robustness to out-of-domain datasets. In the second stage, based on the output of the previous stage, we introduce a multi-task learning mechanism, aiming to improve the model’s performance on unseen data by sharing surface detail information from auxiliary tasks. We validated our method on a publicly available tooth cone beam computed tomography dataset. Our evaluation demonstrates that our method outperforms state-of-the-art approaches.

Definition of New Three-Dimensional Cephalometric Analysis of Maxillomandibular Sagittal Relationship for Orthodontics and Orthognathic Surgery: Normative Data Based on 700 CBCT Scans.

The objective of the study was to define the norm of new 3-dimensional cephalometric analysis of maxillomandibular sagittal relationship with the patient in Natural Head Position. A cross-sectional study was performed using 700 consecutives cone beam computed tomography datasets of pre-orthodontic patients received for three-dimensional craniofacial analysis. To stablish the clinical norm of the new sagittal reference (linear distance A-B), the correlation with the gold standard (ANB angle) was estimated with the Pearson's correlation coefficient. Subsequently, the prognostic values of the linear distance A-B was calculated to define the clinical norm. The sample was composed by 463 women (66.1%) and 237 men (33.9%). The mean age was 30 ± 14,5 years old (range 6-71y old). According to the skeletal class classification (ANB), 46.1% (323) were class I, 42% (294) class II, and 11.9% (83) class III. The regression model found that each additional grade of the ANB angle imply a mean increase of 1.24mm of the distance A-B ( P <0.001). The normative value of the linear distance A-B was obtained through the prognostic values of the distance for the limits of the ANB norm 0 to 4. These values were on the range of 0.52 to 5.48mm. Therefore, the clinical norm for cephalometric maxillomandibular sagittal relationship using linear distance from point A-B is: 3±2.48mm. With this new approach, we can define the skeletal sagittal relationship of the patient in natural head position overcoming the limitations of using intracranial or occlusal plane references improving the diagnosis and orthognathic surgical planning process.

Semi-Automatic Segmentation of Cone Beam Computed Tomography Datasets for Volume Measurements of Equine Cheek Teeth.

The volumes of equine teeth may change considerably over time for several reasons including domestication, routine dental floating, and the hypsodont and anelodont nature of the teeth. Cone beam computed tomography (CBCT) of the head is routinely performed in standing horses and, in this proof of concept study, the feasibility of measuring tooth volume from CBCT datasets was determined. The CBCT images of 5 equine cadaver cheek teeth were segmented with a software 3-dimensional (3D) Slicer using a predefined protocol, corrected manually, and re-assembled into a 3D model. Individual tooth volume (VS) was calculated from the model. After extraction, the volumes were also measured using the "gold-standard" water displacement method (VW) for comparison. The VS of 77 teeth ranged from 7114 to 42,300 mm3 which strongly correlated with VW (r = 0.99), and on average VS was 6.1% less than VW. There was no significant difference in VS between the right and left arcades in individual animals. Maxillary cheek tooth volume was on average 40% larger than it was for mandibular counterparts. Semi-automatic image segmentation of equine cheek teeth from CBCT data is feasible and accurate but requires some manual intervention. This preliminary study provides initial data on the volume of equine cheek teeth and creates new possibilities for future in vivo studies.

Three-dimensional analysis of the velopharyngeal region in patients with cleft palate and healthy individuals

PurposeThis study aims to attain metric data of the velopharyngeal dimensions of healthy subjects as well as patients with velopharyngeal insufficiency using the example of cleft and lip palate (CLP) in order to determine possible differences in the volumes of both groups.MethodsVolumes and distances of velopharyngeal areas were analyzed retrospectively using cone beam computed tomography data sets (n = 60). Group 1 included healthy patients receiving dental implants (n = 31). Group 2 was represented by patients with surgically closed cleft lip and palate (n = 29).ResultsBiggest differences among mean values of both groups were found for: minimum axial area (p = 0.000), airway area caudal (p = 0.000), distance between posterior nasal spine and posterior pharyngeal wall (PPW) (p = 0.014), mean distance between velum and PPW (p = 0.000), length of PPW (p = 0.000) and length of anterior pharyngeal wall (p = 0.000).ConclusionDifferences in the shape and geometry of the velopharyngeal area in subjects with a regular velopharyngeal structure and function and patients with cleft palate do exist. The significant differences found here can be categorized into two groups: one reflects distances between the anterior and posterior pharynx, presenting longer distances for patients with CLP. The second significant difference regards values of length in cranio-caudal direction, which is longer in healthy subjects. With regards to these values, one could conclude, that even though total volumes of both groups did not differ in size, group 1 shows three-dimensional velopharyngeal shapes that are longer and narrower, whereas shapes of patients of group 2 tend to be wider and shorter in general.

Automatic Method for Bone Segmentation in Cone Beam Computed Tomography Data Set

Due to technical aspects of Cone Beam Computed Tomography (CBCT), the automatic methods for bone segmentation are not widely used in the clinical practice of endodontics, orthodontics, oral and maxillofacial surgery. The aim of this study was to evaluate method’s accuracy for bone segmentation in CBCT data sets. The sliding three dimensional (3D) window, histogram filter and Otsu’s method were used to implement the automatic segmentation. The results of automatic segmentation were compared with the results of segmentation performed by an experienced oral and maxillofacial surgeon. Twenty patients and their forty CBCT data sets were used in this study (20 preoperative and 20 postoperative). Intraclass Correlation Coefficients (ICC) were calculated to prove the reliability of surgeon segmentations. ICC was 0.958 with 95% confidence interval [0.896 ... 0.983] in preoperative data sets and 0.931 with 95% confidence interval [0.836 ... 0.972] in postoperative data sets. Three basic metrics were used in order to evaluate the accuracy of the automatic method—Dice Similarity Coefficient (DSC), Root Mean Square (RMS), Average Distance Error (ADE) of surfaces mismatch and additional metric in order to evaluate computation time of segmentation was used. The mean value of preoperative DSC was 0.921, postoperative—0.911, the mean value of preoperative RMS was 0.559 mm, postoperative—0.647 mm, the ADE value of preoperative cases was 0.043 mm, postoperative—0.057 mm, the mean computational time to perform the segmentation was 46 s. The automatic method showed clinically acceptable accuracy results and thus can be used as a new tool for automatic bone segmentation in CBCT data. It can be applied in oral and maxillofacial surgery for performance of 3D Virtual Surgical Plan (VSP) or for postoperative follow-up.

A Feasibility Study on the Role of Ultrasound Imaging of Bladder Volume as a Method to Improve Concordance of Bladder Filling Status on Treatment with Simulation

PurposeAccurate positioning of the prostate is of paramount importance to ensure optimal target coverage and normal tissue sparing in stereotactic ablative body radiation when large doses per fraction are delivered with tight margins around the prostate. Bladder and rectal filling play an important part in controlling the accuracy of a patient's setup and therefore the overall toxicities and outcomes. The aim of this study was to establish the value of characterizing patients' bladder filling kinetics at the time of simulation with ultrasound scans so that a predictive model can be used to ensure that a bladder volume at treatment would match at simulation. MethodsA prospective trial was conducted in unfavorable risk prostate cancer patients to evaluate the utility of ultrasound bladder monitoring. Thirty patients (n = 30) were enrolled in this study. Patients were required to void before simulation and then were given 500 mL of fluids to drink. Ultrasound measurements of the bladder were documented at 15-minute intervals for up to four measurements before simulation. On treatment, bladder volumes were measured at a single time point; typically, half an hour after the patient voided and consumed 500 mL of fluids. The kinetic model was then used to predict the optimal time to set up the patient for treatment such that the bladder volume at treatment would match the volume at simulation. Every patient had a cone beam computed tomography scan before each fraction to ensure accurate patient positioning before dose delivery. Bladder volumes at treatment were measured and compared with those at simulation on the cone beam computed tomography data sets using MIMVISTA software. ResultsOf 30 patients, 26 were analyzed. The comparison of the bladder contours at treatment compared to simulation yielded a DICE coefficient (similarity) of 0.76 ± 0.11. The largest variation in bladder size was seen in the anterior-posterior direction. ConclusionsThis study demonstrated that ultrasound monitoring of the bladder status was a valuable tool in ensuring reproducible bladder filling on treatment. The bladder kinetic model indicated the general time required to achieve optimal bladder filling was 60 minutes after voiding and drinking 500 mL of water.

A longitudinal four-dimensional computed tomography and cone beam computed tomography dataset for image-guided radiation therapy research in lung cancer.

To describe in detail a dataset consisting of serial four-dimensional computed tomography (4DCT) and 4D cone beam CT (4DCBCT) images acquired during chemoradiotherapy of 20 locally advanced, nonsmall cell lung cancer patients we have collected at our institution and shared publicly with the research community. As part of an NCI-sponsored research study 82 4DCT and 507 4DCBCT images were acquired in a population of 20 locally advanced nonsmall cell lung cancer patients undergoing radiation therapy. All subjects underwent concurrent radiochemotherapy to a total dose of 59.4-70.2 Gy using daily 1.8 or 2 Gy fractions. Audio-visual biofeedback was used to minimize breathing irregularity during all fractions, including acquisition of all 4DCT and 4DCBCT acquisitions in all subjects. Target, organs at risk, and implanted fiducial markers were delineated by a physician in the 4DCT images. Image coordinate system origins between 4DCT and 4DCBCT were manipulated in such a way that the images can be used to simulate initial patient setup in the treatment position. 4DCT images were acquired on a 16-slice helical CT simulator with 10 breathing phases and 3 mm slice thickness during simulation. In 13 of the 20 subjects, 4DCTs were also acquired on the same scanner weekly during therapy. Every day, 4DCBCT images were acquired on a commercial onboard CBCT scanner. An optically tracked external surrogate was synchronized with CBCT acquisition so that each CBCT projection was time stamped with the surrogate respiratory signal through in-house software and hardware tools. Approximately 2500 projections were acquired over a period of 8-10 minutes in half-fan mode with the half bow-tie filter. Using the external surrogate, the CBCT projections were sorted into 10 breathing phases and reconstructed with an in-house FDK reconstruction algorithm. Errors in respiration sorting, reconstruction, and acquisition were carefully identified and corrected. 4DCT and 4DCBCT images are available in DICOM format and structures through DICOM-RT RTSTRUCT format. All data are stored in the Cancer Imaging Archive (TCIA, http://www.cancerimagingarchive.net/) as collection 4D-Lung and are publicly available. Due to high temporal frequency sampling, redundant (4DCT and 4DCBCT) data at similar timepoints, oversampled 4DCBCT, and fiducial markers, this dataset can support studies in image-guided and image-guided adaptive radiotherapy, assessment of 4D voxel trajectory variability, and development and validation of new tools for image registration and motion management.

Use of cone beam computed tomography in implant dentistry: current concepts, indications and limitations for clinical practice and research.

Diagnostic radiology is an essential component of treatment planning in the field of implant dentistry. This narrative review will present current concepts for the use of cone beam computed tomography imaging, before and after implant placement, in daily clinical practice and research. Guidelines for the selection of three-dimensional imaging will be discussed, and limitations will be highlighted. Current concepts of radiation dose optimization, including novel imaging modalities using low-dose protocols, will be presented. For preoperative cross-sectional imaging, data are still not available which demonstrate that cone beam computed tomography results in fewer intraoperative complications such as nerve damage or bleeding incidents, or that implants inserted using preoperative cone beam computed tomography data sets for planning purposes will exhibit higher survival or success rates. The use of cone beam computed tomography following the insertion of dental implants should be restricted to specific postoperative complications, such as damage of neurovascular structures or postoperative infections in relation to the maxillary sinus. Regarding peri-implantitis, the diagnosis and severity of the disease should be evaluated primarily based on clinical parameters and on radiological findings based on periapical radiographs (two dimensional). The use of cone beam computed tomography scans in clinical research might not yield any evident beneficial effect for the patient included. As many of the cone beam computed tomography scans performed for research have no direct therapeutic consequence, dose optimization measures should be implemented by using appropriate exposure parameters and by reducing the field of view to the actual region of interest.

Evaluation of two dental registration-splint techniques for surgical navigation in cranio-maxillofacial surgery

BackgroundSurgical navigation requires precise registration of the pre-operative image dataset to the patient in the operation theatre. Different marker-based and marker-free registration techniques are available, each of them with advantages and disadvantages regarding precision and clinical handling. In this model study, the precision of two dental splint techniques for marker-based registration is analyzed. Materials and methodsA synthetic full-size human skull was registered with its cone beam computed tomography dataset using (a) a dentally-mounted “rapid” occlusal splint with five titanium screws directly attached to the splint, (b) an “extender”, a dentally-mounted occlusal splint with similar fiducials fixed to an extension of the splint. The target registration error was measured for 170 landmarks distributed over the viscero- and neurocranium in 10 repeats per splint type using the Vector Vision2 (BrainLAB AG, Heimstetten, Germany) navigation system. Statistical and graphical evaluations were performed per anatomical region. ResultsIn the periorbital region, the rapid splint, with an average deviation of 1.50 mm (SD = 0.439) showed greater accuracy than the extender with 1.76 mm (SD = 0.525). The viscerocranial results for both splints were similar (extender 1.84 mm, SD = 0.559, rapid occlusal splint 1.86 mm, SD = 0.686). In the cranial vault region, registration with the extender (2.33 mm, SD = 0.685) proved to be more precise than with the rapid splint (2.86 mm, SD = 0.929). ConclusionsDue to the more compact dimension of the rapid occlusal splint, errors close to the splint were smaller compared to the extender technique. The advantage of greater distances between the registration fiducials on the extender is particularly important in areas such as the orbital roof, the cranial vault, and the lateral skull base.

An Evaluation of Face-Bow Transfer for the Planning of Orthognathic Surgery

The purpose of this study was to evaluate the error magnitude in the clinical application of face-bow devices. Technical and methodologic inaccuracies, as well as deviations from reference planes, were determined. The presented method is part of a 3-dimensional virtual planning procedure for orthognathic surgery and included 15 patients with dentoskeletal deformities. Cone beam computed tomography datasets obtained from patients with a referenced face-bow plane and a centric registration splint were matched with cone beam computed tomography datasets of the registered plaster model of the maxilla mounted in an articulator. To assess potential sources of methodologic errors, angulations were measured between the virtual face-bow plane and the horizontal cross bar of the virtual articulator. To evaluate the reproducibility of the anatomic reference plane, angulations between the Frankfort plane and the horizontal cross bar of the articulator were measured. Statistical significance was set at P < .05 and tested by univariate analysis of variance. Technical and methodologic errors showed a mean deviation of 3.5°, with a median of 3.6° and SD of 2.7°. The values did not reach statistical significance (P = .1). However, there was a significant error (P < .05) in determining the position of the anatomic reference plane by face-bow transfer. The mean deviation was 7.7° (values ranged between 1.2° and 18.9°), with a median of 6.7° and SD of 5.3°. In this study the traditional use of face-bow devices showed inaccuracies in model mounting as well as in assignment of anatomic reference planes. Three-dimensional virtual computer-assisted planning seems to be more accurate than conventional methods.