Radiological Physics and Technology最新文献

Abdominal computed tomography (CT) is normally performed with patients raising their arms over abdominal region to prevent arm-induced artifacts that degrade image quality. This study aimed to evaluate the effects of deep learning-based image reconstruction (DLIR) on arm-induced artifacts and image quality in abdominal CT with arms-down positioning, compared to adaptive statistical iterative reconstruction-Veo (ASIR-V) and filtered-backprojection (FBP). A liver nodule phantom with arms from a PBU-60 phantom was scanned in three arms-down positions: alongside the torso, across the abdomen, and crossed over the pelvis. Abdominal CT images of 10 patients in arms-alongside-torso position were also included. Images were reconstructed using DLIRs (L-low, M-medium, and H-high), ASIR-Vs (50% and 100%), and FBP. Phantom images were assessed for artifact strength (location parameter of the Gumbel distribution and standard deviation), signal-to-noise ratio, and contrast-to-noise ratio. Two radiologists qualitatively evaluated patient images for noise, artifacts, sharpness, and overall quality. DLIR-H significantly reduced streak artifacts by 37% in location parameters and by 43% in SD, while improving SNR by 28% and CNR by 29% compared to ASIR-V50%. DLIR-M performed significantly better than ASIR-V50% in all quantitative metrics, except in the arms-alongside-torso position. FBP performed worst, although sharpness was comparable. DLIR-H received the best qualitative scores (low noise and artifacts, minimal blurring, and excellent overall image quality), although ASIR-V100% had lower subjective noise. DLIR outperformed ASIR-V and FBP in arm-induced artifact reduction and image quality and is a preferable reconstruction method for arms-down abdominal CT.

The goal of this study was to assess radiation doses to the eyes of seated patients and medical staff during a videofluoroscopic swallowing study (VFSS), considering the effects of lead-equivalent glasses and the height of the medical staff. Entrance surface air kerma (ESAK) at eye level was measured using nanoDot dosimeters on anthropomorphic phantoms representing medical staff with heights of 150, 165, and 180 cm. ESAK measurements were performed with and without 0.5-0.6 mm lead-equivalent glasses. During a 10-minute fluoroscopy procedure, the mean ESAK at patient's eyes without lead-equivalent glasses was higher in the anteroposterior (AP) position (36.5 mGy) than in the lateral position (14.5 mGy). The corresponding ESAK at medical staff's eyes was notably higher in the 150-cm phantom (0.097 mGy), with an increase of 77.1% compared with that in the 165-cm phantom (0.043 mGy) and 87.4% compared with that in the 180-cm phantom (0.038 mGy). The radioprotective efficiency of lead-equivalent glasses in patients in the anteroposterior position was 92.0-93.4%, which was higher than that in the lateral position (15.8-83.2%). Radiation doses to the eyes were highest for shorter medical staff, highlighting the importance of protective measures, especially for those with shorter stature. The radioprotective efficiency of lead-equivalent glasses was found to be lower in the lateral position, underscoring the need for careful consideration of glasses design.

Intraprocedural visualization of the iceball boundary is often limited at the fat-ice interface, where frozen fat-despite increased computed tomography (CT) values-remains within the negative range, thus yielding limited contrast with non-frozen fat. This limitation is relevant in CT-guided renal cryoablation involving perirenal fat. We evaluated a stepwise CT post-processing method of subtraction and scaled addition with probabilistically adjusted thresholding, using an in situ fat-muscle phantom. This two-step process involved fixed zero-threshold subtraction (Step 1: post-freezing image minus pre-freezing image) and kernel density estimation-based threshold subtraction (Step 2: Step 1 output minus post-freezing image), based on pixel-wise fat-attenuation distributions. Contrast-to-noise ratio improved in both fat and non-fat tissues. In fat tissue, boundary contrast selectively increased by reducing CT values in non-frozen regions, whereas in non-fat tissue, by reducing them in frozen regions. Iceball boundaries aligned with magnetic resonance imaging. This approach may improve iceball demarcation and warrants validation in clinical practice.

This study aimed to evaluate discrepancies between planned, delivered and adapted doses to target and organs at risk (OARs) during bladder cancer radiotherapy, using deformable image registration from CT to CBCT in view of better treatment personalization. Twenty patients with a total of 165 CBCT images were analysed. Intensity modulated techniques (IMRT and VMAT) were simulated in the Monaco planning system. The initial plans were adapted to each CBCT using the Adapt to shape (ATS) function. A reduction in PTV66 and CTV66 coverage was observed upon calculation of treatment dose on CBCTcalc in comparison with plans on CT. The PTV66 coverage was 86.7% IMRT and 86.4% VMAT, respectively. Meanwhile, the optimized plans on CBCTopt provided PTV66 coverage of 97.1% IMRT and 96.4% VMAT, which was similar to the initial planning on CT (97.4% IMRT and 96.6% VMAT). Furthermore, CTV66 showed a coverage of 94.2% IMRT and 94.0% VMAT on CBCTcalc, in comparison to the values on CT (99.8% IMRT and 99.9% VMAT) and the values on CBCTopt (99.8% IMRT and 99.8% VMAT). For OARs, the rectum, bowel bag, and sigmoid exhibited higher values on CBCTcalc than on CT planning and CBCTopt. This study demonstrates that CBCT- guided adaptive radiotherapy enhances treatment precision and personalization in bladder cancer, improving target coverage and reducing radiation exposure to healthy tissues. Next to highlighting the importance of personalizing bladder cancer radiotherapy, the study substantiates that daily reoptimization with ATS constitutes an efficacious strategy in centers with limited resources.

Dosimetry using SPECT/CT images enables personalized medicine by estimating absorbed doses and optimizing therapy. Differences in organ contouring and calculation algorithms contribute to inter-institutional variability, emphasizing the need for standardization. The present study aimed to investigate factors contributing to inter-institutional variability in kidney dosimetry in Japan. We analyzed four time points in SPECT/CT images of one male and one female patient each from the ¹⁷⁷Lu SNMMI Dosimetry Challenge. Kidney volumes and absorbed doses were calculated at 10 Japanese institutes using their preferred organ-based (OLINDA 2.2, IDAC DOSE 2.1) and voxel-based (Voxel Dosimetry, RT-PHITS, MIM SurePlan MRT, OpenDose3D) software. Reference volumes of interest (VOI) files were distributed to assess the effect of contouring differences on kidney volumes and absorbed doses. Manual VOI contouring revealed substantial inter-institutional variability in kidney volumes, with coefficients of variation (%CVs) up to 16.9%. The reference VOIs reduced volume variability to ≤ 7.4%. Compared to manual VOIs, reference VOIs showed slightly increased doses in both patients with slightly reduced inter-institutional variability. The absorbed doses were generally higher in voxel- than organ-based dosimetry. The %CVs of the right and left kidneys in female patient decreased from 31.36% to 6.26% and 41.28%-3.97%, respectively. Variability in Kidney volume and absorbed doses significantly varied among Japanese institutes. Reference VOIs reduced volume variability but could not fully control dose differences. Voxel-based dosimetry can mitigate inter-institutional variability independent of contouring. Our findings emphasize the importance of algorithm standardization for reliable ¹⁷⁷Lu-DOTATATE kidney dosimetry in Japan.

Previous studies have shown that high kilovoltage (kV) angiographic imaging techniques can reduce radiation doses to patients more effectively than using low kV techniques. While radiologists often accept the resulting image quality, a detailed quantitative comparison between these techniques remains limited. This study aimed to evaluate and compare the quality of cerebral angiographic images acquired using high kV (79-90 kV) and low kV (68-82 kV) techniques on a biplane digital subtraction angiography (DSA) system. Images were analyzed from patients with cerebral aneurysms as well as a quality assurance phantom (TO DSA), focusing on 2-dimensional angiography (2D-DSA). The contrast-to-noise ratio (CNR) and signal-to-noise ratio (SNR) were measured at various vascular locations in posteroanterior (PA) axial and lateral views. While demographic data did not differ between groups, CNR for PA axial view and PA phantom images produced with high kV was significantly lower than that with low kV. In contrast, the high kV technique demonstrated higher SNR values in both PA and lateral views compared to the low kV technique. Radiation dose per frame confirmed a reduction in dose for the high kV protocol. Conversely, TO DSA images acquired using high kV had a lower SNR than those from low kV. The low kV technique achieved better vessel contrast, as evidenced by its higher CNR compared to the high kV technique. However, it also resulted in a lower SNR in patient images and a higher radiation dose. Protocol selection should, therefore, aim to optimize the trade-off between image quality and radiation exposure.

This study aimed to propose a definition of linearity in image reconstruction and demonstrate, by reductio ad absurdum, that the row-action maximum likelihood algorithm (RAMLA) and ordered subset expectation maximization (OSEM) are nonlinear when the number of iterations is low and linear approximation when the number of iterations increases. Block sequential regularized expectation maximization (BSREM) and one-step late maximum a posteriori expectation maximization (OSLEM), which serve as regularized versions of RAMLA and OSEM, respectively, remain nonlinear regardless of the number of iterations. Simulations using ideal two-dimensional (2D) parallel beam projections validated the results of the reductio ad absurdum proof. The three numerical phantoms were point source ${\overline{x}}_1$ , represented by 2D Gaussian with a full width at half maximum of 3 pixels positioned at the center of disk background; point source ${\overline{x}}_2$ , separated by 24 pixels along the x-axis; and point source ${\overline{x}}_3$ , is the sum of ${\overline{x}}_1$ and ${\overline{x}}_2$ . In numerical experiment, when the difference of the area under the curve (AUC) or recovery for reconstructed image of ${\overline{x}}_3$ and the summed reconstructed images of ${\overline{x}}_1$ and ${\overline{x}}_2$ is within reference values, or when AUC profiles are visually consistent, we defined image reconstruction as linear approximation. RAMLA and OSEM were deemed nonlinear when less than 20 iterations were performed with 64 subsets and linear approximation when $\ge 20$ iterations were used. By contrast, BSREM and OSLEM remained nonlinear. Algebraic reconstruction technique is linear and its regularized variant has a tendency of linear approximation, indicating that the same regularization function works differently in linear and nonlinear image reconstructions.