BackgroundTwo-dimensional projection radiographs have been traditionally considered the modality of choice for cephalometric analysis. To overcome the shortcomings of two-dimensional images, three-dimensional computed tomography (CT) has been used to evaluate craniofacial structures. However, manual landmark detection depends on medical expertise, and the process is time-consuming. The present study was designed to produce software capable of automated localization of craniofacial landmarks on cone beam (CB) CT images based on image registration and to evaluate its accuracy.MethodsThe software was designed using MATLAB programming language. The technique was a combination of feature-based (principal axes registration) and voxel similarity-based methods for image registration. A total of 8 CBCT images were selected as our reference images for creating a head atlas. Then, 20 CBCT images were randomly selected as the test images for evaluating the method. Three experts twice located 14 landmarks in all 28 CBCT images during two examinations set 6 weeks apart. The differences in the distances of coordinates of each landmark on each image between manual and automated detection methods were calculated and reported as mean errors.ResultsThe combined intraclass correlation coefficient for intraobserver reliability was 0.89 and for interobserver reliability 0.87 (95% confidence interval, 0.82 to 0.93). The mean errors of all 14 landmarks were <4 mm. Additionally, 63.57% of landmarks had a mean error of <3 mm compared with manual detection (gold standard method).ConclusionThe accuracy of our approach for automated localization of craniofacial landmarks, which was based on combining feature-based and voxel similarity-based methods for image registration, was acceptable. Nevertheless we recommend repetition of this study using other techniques, such as intensity-based methods.

No abstract

Background: Scoliosis is the most common type of spinal deformity. A universal and standard method for evaluating scoliosis is Cobb angle measurement, but several studies have shown that there is intra- and inter- observer variation in measuring cobb angle manually.Objective: Develop a computer- assisted system to decrease operator-dependent errors in Cobb angle measurement.Methods: The spinal cord in the given x-ray image of the spine is highlighted using contract-stretching technique. The overall structural curvature of the spine is determined by a semi-automatic algorithm aided by the operator. Once the morphologic curve of the spine is determined, in the last step the cobb-angle is estimated by calculating the angle between two normal lines to the spinal curve at the inflection points of the curve.Results: Evaluation results of the developed algorithms using 14 radiographs of patients ( 4 - 40 years old) with cobb angle ranges from 34 - 82 degrees, revealed that the developed algorithm accurately estimated cobb angle. Statistical analysis showed that average angle values estimated using the developed method and that provided by experts are statistically equal. The correlation coefficient between the angle values estimated using the developed algorithm and those provided by the expert is 0.81.Conclusion: Compared with previous algorithms, the developed system is easy to use, less operator-dependent, accurate, and reliable. The obtained results are promising and show that the developed computer-based system could be used to quantify scoliosis by measuring Cobb angle.

Two-dimensional (2D) arrays of thick segmented scintillators are of interest as X-ray detectors for both 2D and 3D image-guided radiotherapy (IGRT). Their detection process involves ionizing radiation energy deposition followed by production and transport of optical photons. Only a very limited number of optical Monte Carlo simulation models exist, which has limited the number of modeling studies that have considered both stages of the detection process. We present ScintSim1, an in-house optical Monte Carlo simulation code for 2D arrays of scintillation crystals, developed in the MATLAB programming environment. The code was rewritten and revised based on an existing program for single-element detectors, with the additional capability to model 2D arrays of elements with configurable dimensions, material, etc., The code generates and follows each optical photon history through the detector element (and, in case of cross-talk, the surrounding ones) until it reaches a configurable receptor, or is attenuated. The new model was verified by testing against relevant theoretically known behaviors or quantities and the results of a validated single-element model. For both sets of comparisons, the discrepancies in the calculated quantities were all <1%. The results validate the accuracy of the new code, which is a useful tool in scintillation detector optimization.

Objectives: The aim of this study was to design and evaluate a new method for automated localization of the inferior alveolar nerve canal on CBCT images. Methods: The proposed method is based on traversing both panoramic and cross-sectional slices. For the panoramic slices, morphological skeletonization is imposed, and a modified Hough transform is used while traversing the cross-sectional slices. A total of 40 CBCT images were randomly selected. Two experts twice located the inferior alveolar nerve canal during two examinations set 6 weeks apart. Agreement between experts was achieved, and the result of this manual technique was considered the gold standard for our study. The distances for the automated method and those determined using the gold standard method were calculated and recorded. The mean time required for the automated detection was also recorded. Results: The average mean distance error from the baseline was 0.75 ± 0.34 mm. In all, 86% of the detected points had a mean error of ,1 mm compared with those determined by the manual gold standard method. Conclusions: The proposed method is far more accurate and faster than previous methods. It also provides more accuracy than human annotation within a shorter time.