Radiological Physics and Technology最新文献

The purpose of this study was to investigate the spatial resolution of non-contrast-enhanced (CE) T2prep multi-shot gradient echo planar imaging (MSG-EPI) magnetic resonance angiography (MRA) required to identify peroneal artery perforators and demonstrate its effectiveness in preoperative simulation. Twenty-six legs of 13 volunteers were scanned using non-CE T2prep MSG-EPI-MRA at three spatial resolutions: 1.0-, 0.8-, and 0.6-mm isotropic voxels. The location and number of peroneal artery perforators that could be candidates for free fibula flaps were identified by consensus among three plastic surgeons. Surgeons distinguished between septocutaneous and musculocutaneous perforators using MRA, and confirmed the accuracy of their presence and identification using ultrasonography (US). The ability to detect hypoplasia or stenosis of the anterior tibial, posterior tibial, and peroneal arteries was evaluated by confirming the consistency between the MRA and US results. The number of cutaneous perforators identified using MRA and confirmed using US was 39, 51, and 52 at each respective resolution. The discrimination accuracies between septocutaneous and musculocutaneous perforators were 92.3%, 96.1%, and 96.2%. The number of identified septocutaneous perforators was 1.3 ± 0.6, 1.6 ± 0.8, and 1.7 ± 0.8 at 1.0-, 0.8-, and 0.6-mm data, respectively. All the MRA results, including hypoplasia and stenosis, were consistent with the US results. Non-CE T2prep MSG-EPI-MRA with a spatial resolution of 0.8 mm or less shows promise for identifying septocutaneous perforators of the peroneal artery, suggesting its potential as an alternative to conventional imaging methods for the preoperative planning of free fibula osteocutaneous flap transfers.

Diagnostic reference level (DRL) for mammography for 2015 and 2020 has been published by J-RIME. More new dose studies are needed to revise the next DRL. In preparation for the next revision of the DRL for mammography, this study investigated data from the Japan Central Organization on Quality Assurance of Breast Cancer Screening on the mean average glandular dose (AGD) for institutional image accreditation in 2019-2023 and the relationship between the average at eligible institutions to date and the type of breast X-ray system. The 95th percentile values of the AGD distributions for the Computed Radiography (CR) and Flat Panel Detector (FPD) systems were 2.5 mGy and 2.0 mGy, respectively. Moreover, it is assumed that AGD is decreasing due to the spread of FPD systems, and it is expected that the further spread of FPD systems and systems with W/Rh target/filter will reduce AGD in future.

This study aimed to determine the scan delay for bolus tracking in the hepatic artery phase (HAP) of hepatic dynamic computed tomography (CT) using the cardiothoracic ratio (CTR) from CT scout images. We retrospectively studied 188 patients who underwent hepatic dynamic CT, 24 of whom had scan delays adjusted for CTR. The contrast enhancement of the abdominal aorta, portal vein, hepatic vein, and hepatic parenchyma was calculated for HAP. The adequacy of the scan timing for HAP was assessed using three classifications: early, appropriate, or late. The effect of HAP on scan timing adequacy was determined using multivariate logistic regression analysis, and the optimal cutoff value of CTR was evaluated using receiver operating characteristic analysis. The trigger times for bolus tracking (odds ratio: 1.58) and CTR (odds ratio: 1.23) were significantly affected by the appropriate scan timing of the HAP. The optimal cutoff value of CTR was 59.3%. The scan timing of HAP with a scan delay of 15 s was 14% of early and 86% of appropriate, and the proportion of early in CTR ≥ 60% (early, 52%; appropriate, 48%) was higher than that in CTR < 60% (early, 6%; appropriate, 94%). Adjusting the scan delay to 20 s in CTR ≥ 60% increased the proportion of appropriate (early, 4%; appropriate, 96%). The CTR of a CT scout image is an effective index for determining the scan delay for bolus tracking. Adjusting the scan delay by CTR can provide appropriate HAP images in more patients. Trial registration number: R-080; date of registration: 9 March 2023, retrospectively registered.

In this study, we investigated the application of distributed learning, including federated learning and cyclical weight transfer, in the development of computer-aided detection (CADe) software for (1) cerebral aneurysm detection in magnetic resonance (MR) angiography images and (2) brain metastasis detection in brain contrast-enhanced MR images. We used datasets collected from various institutions, scanner vendors, and magnetic field strengths for each target CADe software. We compared the performance of multiple strategies, including a centralized strategy, in which software development is conducted at a development institution after collecting de-identified data from multiple institutions. Our results showed that the performance of CADe software trained through distributed learning was equal to or better than that trained through the centralized strategy. However, the distributed learning strategies that achieved the highest performance depend on the target CADe software. Hence, distributed learning can become one of the strategies for CADe software development using data collected from multiple institutions.

In normal-pressure hydrocephalus, disturbances in cerebrospinal fluid (CSF) circulation occur; therefore, understanding CSF dynamics is crucial. The two-dimensional phase-contrast (2D-PC) method, a common approach for visualizing CSF flow on MRI, often presents challenges owing to prominent vein signals and excessively high contrast, hindering the interpretation of morphological information. Therefore, we devised a new imaging method that utilizes T₂-weighted high-signal intensification of the CSF and saturation pulses, without requiring specialized imaging sequences. This sequence utilized a T₂-weighted single-shot fast spin-echo combined with multi-phase imaging synchronized with a pulse wave. Optimal imaging conditions (repetition time, presence/absence of fast recovery, and echo time) were determined using self-made contrast and single-plate phantoms to evaluate signal-to-noise ratio, contrast ratio, and spatial resolution. In certain clinical cases of hydrocephalus, confirming CSF flow using 2D-PC was challenging. However, our method enabled the visualization of CSF flow, proving to be useful in understanding the pathophysiology of hydrocephalus.

Electrometers are important devices that are part of the standard dosimetry system. Therefore, we evaluated the variation of electrometer calibration coefficients (k_elec) over 1 year in this study. We investigated two types of electrometers: a rate mode and an integrate mode. Each electrometer was connected to a charge generator, a constant charge was applied, and k_elec was determined by measuring the current. The current measurements were repeated once a month. For electrometers with multiple ranges, measurements were taken at low and medium ranges. Almost all k_elec measurements agreed within 0.2% of the initial measurements. However, the low range of the electrometer with an integrate mode showed seasonal variation, with a variation greater than 0.2%. This study shows that electrometers may exhibit errors that cannot be detected through annual inspections. The importance of quality assurance using a charge generator at one's own institution was demonstrated.

In positron emission tomography (PET), measurements of depth-of-interaction (DOI) information and time-of-flight (TOF) information are important. DOI information reduces the parallax error, and TOF information reduces noise by measuring the arrival time difference of the annihilation photons. Historically, these have been studied independently, and there has been less implementation of both DOI and TOF capabilities because previous DOI detectors did not have good TOF resolution. However, recent improvements in PET detector performance have resulted in commercial PET scanners achieving a coincidence resolving time of around 200 ps, which result in an effect even for small objects. This means that TOF information can now be utilized even for a brain PET scanner, which also requires DOI information. Therefore, various methods have been proposed to obtain better DOI and TOF information. In addition, the cost of PET detectors is also an important factor to consider, since several hundred detectors are used per PET scanner. In this paper, we review the latest DOI-TOF detectors including the history of detector development. When put into practical use, these DOI-TOF detectors are expected to contribute to the improvement of imaging performance in brain PET scanners.

Deep learning, particularly convolutional neural networks (CNNs), has advanced positron emission tomography (PET) image reconstruction. However, it requires extensive, high-quality training datasets. Unsupervised learning methods, such as deep image prior (DIP), have shown promise for PET image reconstruction. Although DIP-based PET image reconstruction methods demonstrate superior performance, they involve highly time-consuming calculations. This study proposed a two-step optimization method to accelerate end-to-end DIP-based PET image reconstruction and improve PET image quality. The proposed two-step method comprised a pre-training step using conditional DIP denoising, followed by an end-to-end reconstruction step with fine-tuning. Evaluations using Monte Carlo simulation data demonstrated that the proposed two-step method significantly reduced the computation time and improved the image quality, thereby rendering it a practical and efficient approach for end-to-end DIP-based PET image reconstruction.

Non-alcoholic fatty liver disease (NAFLD) and its advanced stage, non-alcoholic steatohepatitis (NASH), have become increasingly prevalent owing to the rise in metabolic syndromes. Accurate assessment of hepatic fat deposition and inflammation is crucial for diagnosing and managing NAFLD/NASH. We investigated the influence of Gd-EOB-DTPA, (EOB) on proton-density fat fraction (PDFF) measurements using chemical shift-encoded magnetic resonance imaging (CSE-MRI) at 3-T. In total, 431 patients who underwent EOB contrast-enhanced MRI were included. PDFF measurements were obtained from pre- and post-contrast CSE-MRI. Linear regression and Bland-Altman analyses were performed to assess the correlation and agreement between pre- and post-EOB PDFF measurements. Relative enhancement (RE) of the liver was calculated as an EOB uptake index. There was a significant decrease in PDFF following EOB administration compared with the pre-contrast values (P < 0.0001), which was observed across all PDFF ranges (< 10% and ≥ 10%). Linear regression analysis revealed high correlation between pre- and post-EOB PDFF measurements. Bland-Altman analysis indicated a small bias between pre- and post-EOB PDFF values. Subgroup analysis based on RE showed a significant difference in ΔPDFF between patients with high RE (> 120%) and those with lower RE levels. EOB administration resulted in a slight decrease in PDFF measurements obtained using CSE-MRI at 3-T. We were able to generalize and clarify that the PDFF of the liver on 3D CSE-MRI at 3-T was slightly decreased after EOB administration as we used a larger group of patients compared to previous studies.

We investigated whether magnetic resonance imaging can visualize the distribution in the vitreous cavity via eye drops of ophthalmic solutions, gadolinium-based contrast agent, and ¹⁷O-water, and to clarify the usefulness of ultra-heavily T2-weighted sequences in the research of intraocular distribution. Five different solutions (V-ROHTO, TRAVATANZ, gadobutrol, H₂¹⁷O, and saline) were administered to excised pig eye specimens. The samples were scanned using T1 mapping, T2 mapping, 3D T2-weighted (echo times (TE): 500, 3200, and 4500 ms), a half-Fourier single-shot turbo-spin echo sequence (HASTE; TE: 440 and 3000 ms), and 3D-real inversion-recovery before eye drops administration. Subsequently, we used a plastic dropper to drop a 0.5 mL solution each, and images were obtained up to 26 h later. Temporal changes in the T1 and T2 values of the anterior chamber and vitreous cavity were compared. The other sequences were evaluated by determining temporal signal changes as signal intensity ratio (SIR) compared to "No drop." The T1 and T2 values of samples treated with gadobutrol and H₂¹⁷O decreased over time. The SIR of samples treated with gadobutrol and H₂¹⁷O showed remarkable changes in the 3D T2-weighted images, whereas no remarkable temporal changes were observed in the other solutions. Longer TEs resulted in remarkable changes. We demonstrated that visualization of distribution in the vitreous cavity via eye drops could be achieved with excised pig eyes using gadobutrol and H₂¹⁷O, but not with ophthalmic solutions. Ultra-heavily T2-weighted sequences may be promising for the early and highly sensitive visualization of the intraocular distribution of eye drops.