Search

Purpose. Although deep learning (DL) technique has been successfully used for computed tomography (CT) reconstruction, its implementation on cone-beam CT (CBCT) reconstruction is extremely challenging due to memory limitations. In this study, a novel DL technique is developed to resolve the memory issue, and its feasibility is demonstrated for CBCT reconstruction from sparsely sampled projection data. Methods. The novel geometry-guided deep learning (GDL) technique is composed of a GDL reconstruction module and a post-processing module. The GDL reconstruction module learns and performs projection-to-image domain transformation by replacing the traditional single fully connected layer with an array of small fully connected layers in the network architecture based on the projection geometry. The DL post-processing module further improves image quality after reconstruction. We demonstrated the feasibility and advantage of the model by comparing ground truth CBCT with CBCT images reconstructed using (1) GDL reconstruction module only, (2) GDL reconstruction module with DL post-processing module, (3) Feldkamp, Davis, and Kress (FDK) only, (4) FDK with DL post-processing module, (5) ray-tracing only, and (6) ray-tracing with DL post-processing module. The differences are quantified by peak-signal-to-noise ratio (PSNR), structural similarity index measure (SSIM), and root-mean-square error (RMSE). Results. CBCT images reconstructed with GDL show improvements in quantitative scores of PSNR, SSIM, and RMSE. Reconstruction time per image for all reconstruction methods are comparable. Compared to current DL methods using large fully connected layers, the estimated memory requirement using GDL is four orders of magnitude less, making DL CBCT reconstruction feasible. Conclusion. With much lower memory requirement compared to other existing networks, the GDL technique is demonstrated to be the first DL technique that can rapidly and accurately reconstruct CBCT images from sparsely sampled data. [ABSTRACT FROM AUTHOR]

Introduction: Transverse discrepancies of the maxilla is one of the most prevalent skeletal problems. Due to the drawbacks of conventional RPE and invasive surgical expansion, alternative methods were developed, such as Mini Implant-supported rapid maxillary expanders (MARPE). Many studies reported by Haas, Davis and Kronman, Akkaya et al. regarding the maxillary response after MARPE on the sagittal dimensions were inconclusive. Aim: The aim of this study was to analyze the sagittal effects on the maxilla after Mini Implant assisted rapid palatal expansion (MARPE) in adolescents. Materials and methods: Pre-treatment CBCT images (T1) were taken as a standard initial record for all patients and post-treatment CBCT (T2) was taken three months in retention after expansion with MARPE. The parameters SNA angle, Frankfort horizontal plane to NA angle, A-N perpendicular in mm, ANS - PNS (in mm) were measured and calculated for evaluation of the sagittal maxillary changes. Comparisons were then made between the two treatment groups at T1, T2 and T2 - T1. Results: A positive increase was seen from pretreatment to post-treatment in both angular and linear measurements showing the significant sagittal skeletal change. Conclusion: The study found a statistically significant forward displacement of maxilla after transverse skeletal expansion in adolescent patients with transverse problems achieved through MARPE. [ABSTRACT FROM AUTHOR]

Abstract A modern cone-beam computed tomography (CBCT) system with a C-arm gantry incorporating a large-area flat-panel detector is an important imaging tool widely used for diagnosis and image-guidance in spine surgery, orthopedic and interventional suite, and image-guided radiation therapy. In this study, the experimental prototype CBCT imaging platform consists of a benchtop system that is integrated with a cone-beam X-ray tube, a collimator, an anti-scatter grid, and a large-area TFT-based flat-panel detector. The different projection images in the C-arm CT system were usually acquired at short scanning angles with a constant interval for various exposure conditions. The performance of CT imaging quality was performed using the Feldkamp, Davis and Kress (FDK) reconstruction algorithm through acquired two-dimensional projection images at different scanning angles and projection numbers. Quantitative analysis of the image quality was performed by using the cone-beam CT phantom for spatial resolution, low-contrast resolution, noise, and two different phantoms, such as the head and pelvis. [ABSTRACT FROM AUTHOR]

The authors propose a dual-energy approach to reduce cone-beam artefacts in a circular orbit cone-beam computed tomography (CT) system. When there exist multiple dense structures within a large cone angle, two-pass algorithm is ineffective for reducing cone-beam artefacts because high-density materials are incorrectly segmented with simple thresholding. The proposed algorithm utilises projection data obtained with dual-energy X-ray spectra to estimate projections of the high-density materials from multi-density anatomical structures. Using bone and water as two of basis functions, the projection data of bone materials can be precisely estimated via dual-energy scans. Then, bone-only images are reconstructed by a constrained total-variation minimisation-based iterative algorithm. The reconstructed bone images are used to generate cone-beam artefacts and the final corrected images are acquired by subtracting the generated cone-beam artefacts from the original Feldkamp, Davis, and Kress images. The proposed method was validated using an extended cardiac-torso phantom with complex vertebral anatomy and compared with the two-pass algorithm. The results show that the proposed method restores distorted bony structures and intensity values, especially in regions with large cone angles. A qualitative evaluation of the mean-squared errors and structural similarities demonstrates the effectiveness of the proposed method compared to the two-pass algorithm. [ABSTRACT FROM AUTHOR]