Journal of Radiation Research and Applied Sciences最新文献

Background

Image-guided radiation therapy (IGRT) involves online medical imaging systems. CT-guided imaging systems not only have high image quality but also allow for direct dose calculation for radiation treatment planning, making it an ideal image guidance method. However, the imaging dose from CT images may pose additional risks to patients. Therefore, the aim of this study was to evaluate the substitutability of using low-dose CT images for IGRT.

Materials and methods

The CIRS-062 phantom, Catphan 604 phantom, CT dose index phantom were used to obtain the conversion curves of the CT values to the relative electron density (RED), image quality, and imaging dose of the low-dose CT, respectively. Catphan 604 phantom and clinical cases were used to analyze the feasibility of low-dose CT for dose calculation.

Results

No significant differences were observed in the conversion curves of CT values to RED between low-dose and standard-dose CT scans. In terms of image quality, the geometric distortion of low-dose CT was 0.1 mm, which was the same as that of standard-dose CT. The spatial resolution of the images was 0.42–0.45 lp/mm, which was slightly better than that of standard-dose CT (0.40–0.44 lp/mm). The image uniformity was 98.4%–99.0%, slightly worse than that of standard-dose CT (99.4%–99.5%), but still met the clinical requirements. The imaging dose was 2.96–4.62 mGy, significantly lower than that of standard-dose CT (9.71–26.68 mGy), especially for head CT protocol. In terms of dose calculation, the γ passing rates were not lower than 98% for either the Catphan 604 phantom or the clinical cases.

Conclusion

Low-dose CT can significantly reduce the imaging dose without a significant loss of image quality and can be directly used for dose calculation in radiation treatment planning, providing a safer and more effective image guidance method for IGRT.

In this paper, we investigated different cement mortars for gamma-rays shielding. Six composite mortars were prepared by replacing fine aggregate (sand) with 0, 20, 40, 60 and 100% of cathode ray tube (CRT) glass. The composites represented by CM-CRT-0 (without CRT), CM-CRT-20, CM-CRT-40, CM-CRT-60, CM-CRT-80, and CM-CRT-100 (without sand). The mass attenuation coefficient, or MAC, of these materials has been determined experimentally at a wide range of energy ranges. At 0.060 MeV and 0.662 MeV, the samples with highest CRT concentrations had the greatest MAC values. However, at energies above 1 MeV, the opposite trend between MAC and CRT content is observed. The prepared composites' linear attenuation coefficient was also determined with the results showing that the CM-CRT-100 sample, that with the greatest concentration of CRT, is characterized at all tested energies by the optimum shielding capabilities. The half value layer (HVL) values are observed to increase with energy, with the lowest occurring at 0.060 MeV, and continuously increasing until 1.333 MeV. For example, the HVL of CM-CRT-20 is 0.616, 3.815, 5.063, and 5.408 cm at 0.060, 0.662, 1.173, and 1.333 MeV, respectively, while that of CM-CRT-60 is 0.303, 3.661, 4.972, and 5.321 cm, respectively. At 0.060 MeV, the CM-CRT-60, CM-CRT-80, and CM-CRT-100 samples have radiation absorption rate, or RAR values of close to 100%. At elevated energies like 1.173 MeV, the RAR values range between 49.24% for CM-CRT-0 to 50.93% for CM-CRT-100.

In this paper, borosilicate glasses with varying concentrations of sodium oxide were synthesized using the melt-quenching method. The amorphous nature of the glasses under investigation, with a composition of (20 + x)Na₂O-(60-x)B₂O₃–20SiO₂ (where x = 0, 3, 5, 7, 8, 10, 13, 15, 17, and 20 mol%), was confirmed through X-ray diffraction (XRD) analysis. Fourier-transform infrared (FT-IR) spectroscopy was employed to analyze the glasses and revealed an increase in the concentration of non-bridging oxygen atoms (NBO) with increasing amounts of Na₂O. The N₄ parameter, which represents the fraction of NBOs, was calculated and found to decrease from 74% to 52% as the concentration of Na₂O increased. The density of the glasses increased from 2.391 g/cm₃ to 2.744 g/cm³ when B₂O₃ was replaced with Na₂O, while the molar volume decreased from 27.68 cm³/mol to 23.56 cm³/mol. Theoretically, theoretical calculations were performed using newly developed computer software called Phys-PSD to assess their shielding properties, including attenuation coefficients, half value layer, and buildup factors. The calculations were conducted in the energy range of 0.015–15 MeV. The results indicated that the synthesized glasses exhibited promising potential as radiation shields.

Radiotherapy, a mainstay cancer treatment for patients worldwide, utilizes medical linear accelerators (linacs) with photon energies ranging from 4 MV to 25 MV. However, concerns arise at higher energies (>6 MV, particularly >10 MV). These high-energy photons interact with high-atomic number materials in the linac head and collimation system, generating unwanted neutrons through (γ,n) reactions. This neutron contamination, present in both photon and electron beams, is a significant issue. Neutrons, with their high Linear Energy Transfer (LET), are more effective at causing clustered DNA damage (single and double-strand breaks). These neutrons not only impact shielding requirements in treatment rooms but also increase out-of-field radiation doses for patients receiving high-energy photon therapy. Therefore, for radiotherapy treatments exceeding 6 MV, additional precautions become crucial, including enhanced door shielding and optimized treatment planning. This review discusses in detail the multifaceted aspects of neutron production and shielding requirements during radiotherapy.

Background

The retrieval of magnetic resonance imaging (MRI) data holds paramount importance in clinical settings and sports medicine due to the limitations of conventional methods, such as slow speed, low accuracy, and limited learning capabilities. Enhancing this retrieval process is critical for advancing sports injury diagnostics and treatment outcomes. Overcoming these challenges is vital for improving healthcare practices and sports medicine methodologies.

Method

This study investigates the utilization of autoencoders in deep learning to efficiently retrieve MRI data from databases for sports injury diagnosis and treatment, with a focus on the model's ability to be trained with a small amount of labeled data. This research aims to enhance the MRI data retrieval process by leveraging autoencoders, showcasing the potential of deep learning technologies in sports injury diagnostics without the necessity of extensive labeled datasets for training.

Results

Findings have showcased the remarkable benefits of this approach for MRI data retrieval tasks, achieving an average accuracy of 99.09%. This signifies the exceptional performance of the technique within this specific domain, demonstrating its effectiveness and reliability in extracting MRI data.

Conclusions

This innovative methodology can enhance the management of archival data and diagnostic capabilities of medical images in sports injury contexts, offering an efficient and dependable solution for MRI data retrieval. It not only facilitates rapid clinical diagnosis and sports medicine research but also proposes a convenient approach for medical image file management.

In this research, we investigated the unsteady magnetohydrodynamic (MHD) convective flow of Casson hybrid nanofluids over an oscillating plate, considering the effects of a porous medium and thermal radiation. These hybrid nanofluids, composed of multiple nanoparticles dispersed in a base fluid, offer advantages over single nanoparticle suspensions, including improved heat transfer properties and enhanced thermal conductivity. The study focused on Ag–Au/blood hybrid nanofluids across an oscillating plate. The drug-carrying fluid can navigate complex vascular structures effectively by using hybrid nanofluids e.g., silver and gold nanoparticles in blood. Also, by employing external magnetic fields to guide and concentrate the drug-carrying nanofluid to the target area. The Casson fluid, known for its elasticity behavior, is a non-Newtonian fluid with diverse applications in industrial and engineering sectors. Mathematically, we incorporated the Casson fluid model to express blood flow, accounting for convection, thermal radiation, and porous medium effects. By transforming the governing partial differential equations into dimensionless form using dimensionless parameters, we employed the Laplace transform method to find the exact solutions for velocity and temperature. Graphical analysis revealed that fluid velocity decreases with increasing magnetic field parameter but increases with Darcy parameter, Casson fluid parameter, Grashof number, nanoparticle concentration, and unsteady parameter. Additionally, the hybrid nanofluid temperature rises proportionally with the radiation parameter, nanoparticle volume fraction, and unsteady parameter.

Objective

Diabetic retinopathy (DR) poses a significant challenge as a leading cause of vision impairment among diabetic individuals. Previous endeavors in optical coherence tomography (OCT) image segmentation using conventional deep learning methodologies have exhibited limitations in achieving robust generalization. Our study endeavors to explore the application of deep transfer learning models on OCT images for DR identification, juxtaposing their performance against conventional deep learning approaches.

Methods

Our investigation involved a cohort of 103 DR patients admitted to the ophthalmology department of our institution spanning from January 2023 to January 2024. Through a randomized allocation process, these patients were partitioned into distinct training and validation sets at a ratio of 7:3. Two convolution models, VGG19 and DenseNet, were constructed and transfer learning was carried out. The recognition effect of the traditional model and transfer model is compared and verified.

Results

Our findings demonstrate that both the VGG19 and DenseNet prediction models exhibit notable segmentation performance following transfer learning compared to their non-transfer learning counterparts. Post-transfer learning, the VGG model achieved accuracy, precision, recall, and F1-score values of 0.890, 0.924, 0.950, and 0.867, respectively, while the DenseNet model achieved corresponding values of 0.897, 0.900, 0.931, and 0.859. Furthermore, in the test set, the area under the curve (AUC) improved significantly for both models post-transfer learning, with the VGG model showcasing an AUC of 0.9118 and the DenseNet model exhibiting an AUC of 0.951.

Conclusion

The neural network model leveraging deep transfer learning demonstrates a notable enhancement in the recognition capability of DR based on OCT images. Furthermore, it effectively streamlines the workflow of ophthalmologists, thus warranting further promotion and adoption in clinical practice.

Objectives

The rising prevalence of obesity throughout the world is a major contributor to numerous chronic diseases, including heart disease. This study aimed to explore the relationship between obesity and cardiac geometry using echocardiography.

Methods

A retrospective cohort study was conducted from January 2022 to December 2022 at the Cardiology Department in King Abdulaziz University Hospital in Jeddah, Saudi Arabia. This study included 224 patients, who were referred for echocardiographic examinations. The inclusion criteria of the study were patients of both sexes older than 20 without cardiac disease. The patients were classified into four main groups based on their BMI: underweight, normal weight, overweight, and obese. The echocardiogram scanning was performed according to the protocol of the King Abdulaziz University Hospital echocardiography exam.

Result

A total of 224 patients were included in this study. The results demonstrated a significant increased thickness in the interventricular septal and left ventricular posterior wall between the different body mass index groups, with the obese group having thickened measurements (p < 0.001 and 0.007 respectively). The left atrial diameter was statistically significantly higher in the obese group compared to the non-obese group (p-value = 0.019).

Conclusion

A strong relationship was demonstrated between obesity and changes in cardiac geometry. While left ventricular function remained unaffected in this study, these findings reinforce the concern that obesity poses a significant threat to cardiac health. These changes in cardiac geometry can be detected by echocardiography, which is an essential part of cardiological practice. Early intervention through lifestyle modifications is crucial to mitigate the adverse effects of obesity on the heart.

Purpose

Brachytherapy (BT) plays a crucial role in cervical cancer treatment. This study aimed to develop a 3D dose prediction model for cervical BT using Convolutional Neural Network (CNN).

Methods

In this study, we introduced a dose prediction model guided to generate dose distributions with explicit anatomical mask guidance. The model encompassed 224 clinical cases, including 190 for training-validation and 34 for testing. For performance evaluation, DVH metrics and 3D Gamma analysis were employed. The results were compared with those obtained using a 3D U-net model.

Results

DVH metrics for the test set, including HRCTV D90, HRCTV D95, HRCTV D100, bladder D2CC, sigmoid D2CC, rectum D2CC, and intestine D2CC, yielded values of 5.44 ± 0.91, 5.05 ± 0.88, 3.34 ± 0.79, 4.39 ± 1.53, 3.24 ± 1.31, 3.03 ± 1.87, and 2.71 ± 1.79, respectively. The DVH metrics of dose differences between the predicted dose distribution and the ground-truth plan were 0.63 ± 0.63, 0.60 ± 0.61, 0.53 ± 0.61, 1.21 ± 0.85, 0.71 ± 0.61, 1.16 ± 1.09, and 0.86 ± 0.58, respectively. The 3D gamma passing rates for the 3%/3 mm criteria of HRCTV, bladder, sigmoid, rectum, and intestine were 0.95 ± 0.04, 0.99 ± 0.02, 1.00 ± 0.02, 1.00 ± 0.01, and 1.00 ± 0.00, respectively.

Conclusion

The 3D BT dose prediction system, based on a 3D anatomical mask-guided deep learning network, could accurately generate 3D dose distributions, offering decision support for automatic clinical BT treatment planning.

Chest X-ray images are known to be extremely helpful in the investigation of a numerous pulmonary conditions such as COVID-19 and pneumonia. Many affected individuals may be protected from such pulmonary conditions through early identification. Unfortunately, COVID-19 can be misdiagnosed as pneumonia, which can rapidly worsen and lead to death. The sensitivity of RT-PCR-based COVID-19 detection is also not satisfactory. Herein, a deep-learning (DL) model for predicting three distinct classes, i.e., COVID-19, pneumonia, and normal, is presented, which achieves good classification performance. Three separate publicly available datasets and an additional dataset with their merged form were used to confirm the efficacy of the proposed model. The DL-based techniques compute features from original raw input spatial images and do not directly provide much information on extremely fine image details, which is quite important for biomedical image analysis. As the individual image bit planes (BPs) carry extremely-fine-to-coarse image information and the effective handcrafted pattern maps created from these BPs may incorporate important discriminating information, the deep features computed from such handcrafted pattern maps may provide complementary information regarding the deep features computed using raw spatial input images. Therefore, we propose a blend of deep features computed with raw spatial images and deep features computed using the proposed local bit plane-based pattern maps to predict the three classes. It is demonstrated that the blend of such features supplies improved discrimination potential and is complementary to the sole features. We incorporated multiscale information computed in each BPs along with interscale details to generate the final bit plane-based pattern maps. The proposed model achieved an average accuracy of 100%, 99.9%, 98.8%, and 98.8% for datasets 1,2, and 3 and their combined form, respectively, outperforming the existing methods.