Beam hardening in x-ray computed tomography (CT) is inevitable because of the polychromatic x-ray spectrum and energy-dependent attenuation coefficients of materials, leading to the underestimation of artifacts arising from projection data, especially on metal regions. State-of-the-art research on beam-hardening artifacts is based on a numerical method that recursively performs CT reconstruction, which leads to a heavy computational burden. To address this computational issue, we propose a constrained beam-hardening estimator that provides an efficient numerical solution via a linear combination of two images reconstructed only once during the entire process. The proposed estimator reflects the geometry of metal objects and physical characteristics of beam hardening during the transmission of polychromatic x-rays through a material. Most of the associated parameters are numerically obtained from an initial uncorrected CT image and forward projection transformation without additional optimization procedures. Only the unknown parameter related to beam-hardening artifacts is fine-tuned by linear optimization, which is performed only in the reconstruction image domain. The proposed approach was systematically assessed using numerical simulations and phantom data for qualitative and quantitative comparisons. Compared with existing sinogram inpainting-based and model-based approaches, the proposed scheme in conjunction with the constrained beam-hardening estimator not only provided improved image quality in areas surrounding the metal but also achieved fast beam-hardening correction owing to the analytical reconstruction structure. This work may have significant implications in improving dose calculation accuracy or target volume delineation for treatment planning in radiotherapy.

Study Design: Cadaveric experimental study.Purpose: To evaluate the feasibility and accuracy of pedicle screw placement using a custom-developed augmented reality-assisted pedicle screw fixation (ARPSF) system in a porcine spine model. Overview of Literature: Conventional pedicle screw placement techniques face limitations including potential inaccuracy, radiation exposure, and workflow disruption. Augmented reality technology can overlay virtual surgical planning directly onto the operative field while maintaining the surgeon’s focus on the patient.Methods: Five porcine cadaveric lumbar spines were used in this study. A custom-developed head-mounted display system with optical tracking projected three-dimensional reconstructed spine models and planned screw trajectories into the surgeon’s field of view. A single experienced spine surgeon placed 50 pedicle screws (4.5 mm diameter). Registration was performed using a point-pair matching technique with fifteen anatomical landmarks. Accuracy was assessed via postoperative computed tomography scan, measuring entry point deviation, trajectory deviation, and angular difference, and evaluated using the Gertzbein-Robbins classification.Results: Of the 50 pedicle screws placed, 47 (94%) achieved grade A accuracy with complete containment within the pedicle. The remaining three screws (6%) were classified as grade B, with minor breaches less than 2 mm. No unsafe placements (grades C–E) occurred. The mean entry point deviation was 0.55 mm (standard deviation [SD]=0.33 mm), and the mean deviation at the screw tip was 0.71 mm (SD=0.32 mm). The mean axial angular deviation was 2.04° (SD=0.58°). The average placement time was 2.2 minutes per screw.Conclusions: The custom-developed ARPSF system demonstrated high accuracy for pedicle screw placement in a porcine model, achieving submillimeter precision and minimal angular deviation. This experimental study shows the potential of augmented reality technology to enhance spine instrumentation precision.

This paper introduces GA-VAE, a fine-tuning framework that enhances local feature representation in pre-trained Vector Quantized-VAE (VQ-VAE) models through genetic algorithm-based optimization. While VQ-VAE models have shown promise in learning discrete latent representations for complex data distributions, their performance is often limited by underutilized tokens in the codebook, particularly those representing local features, resulting in incomplete feature capture. Our proposed Codebook Genetic Algorithm (CGA) selectively optimizes these underutilized tokens by applying genetic operations to enhance their representation capability while maintaining the well-learned features of frequently used tokens. The framework identifies tokens with low utilization rates and systematically evolves them through carefully designed genetic operations, promoting the emergence of meaningful latent vectors that better capture local features. Experimental results demonstrate that our method effectively improves the representation of local features while maintaining computational efficiency, leading to more balanced codebook utilization and enhanced reconstruction quality across various datasets.

This paper presents a novel framework that can generate a high-fidelity isosurface model of X-ray computed tomography (CT) data. CT surfaces with subvoxel precision and smoothness can be simply modeled via isosurfacing, where a single CT value represents an isosurface. However, this inevitably results in geometric distortion of the CT data containing CT artifacts. An alternative is to treat this challenge as a segmentation problem. However, in general, segmentation techniques are not robust against noisy data and require heavy computation to handle the artifacts that occur in three-dimensional CT data. Furthermore, the surfaces generated from segmentation results may contain jagged, overly smooth, or distorted geometries. We present a novel local isosurfacing framework that can address these issues simultaneously. The proposed framework exploits two primary techniques: 1) Canny edge approach for obtaining surface candidate boundary points and evaluating their confidence and 2) screened Poisson optimization for fitting a surface to the boundary points in which the confidence term is incorporated. This combination facilitates local isosurfacing that can produce high-fidelity surface models. We also implement an intuitive user interface to alleviate the burden of selecting the appropriate confidence computing parameters. Our experimental results demonstrate the effectiveness of the proposed framework.