https://repositorio.ufu.br/handle/123456789/18979 2026-08-13T22:00:30Z 2026-08-13T22:00:30Z https://repositorio.ufu.br/handle/123456789/49507 2026-08-13T12:26:25Z 2026-07-15T00:00:00Z

Title: Modelagem e validação Monte Carlo do acelerador Halcyon™ para caracterização do Beam Stopper Abstract: The Varian Halcyon™ linear accelerator, featuring an O-Ring architecture and fully jawless collimation by a dual-layer multileaf collimator, has a high-atomic-number beam stopper positioned opposite the isocenter, whose radiation-protection function still requires quantitative characterization. This work aimed to develop, validate, and apply a geometric model of the Halcyon™ treatment head in the MCNP6.2 radiation transport code, based on the stochastic Monte Carlo method, in order to characterize the performance of the beam stopper in structural shielding and in patient dose. The model, comprising the target, the primary collimator, the dual-layer multileaf collimator, and the lead–antimony alloy beam stopper (11.34 g/cm³), was validated against the clinical commissioning data of the Eclipse treatment planning system. The gamma index was used, with the 3%/3 mm and 2%/2 mm criteria, on percentage depth-dose curves and lateral profiles for seven field sizes, from 2 × 2 to 28 × 28 cm². The validation showed satisfactory agreement between the model and the clinical data, with passing rates above 95% under the 3%/3 mm criterion in most configurations and deviations expected only for the smallest field sizes. Once the model was validated, the attenuation of the primary beam by the beam stopper and its contribution to the primary barrier of the bunker were estimated. A real clinical prostate volumetric modulated arc therapy (VMAT) plan was then reconstructed from the accelerator's trajectory log and the ICRP 145 AM anthropomorphic phantom, in order to quantify the peripheral and surface dose associated with backscatter from the component. The beam stopper proved to be an effective attenuator of the primary beam transmitted along the central axis, equivalent to 2.58–2.98 tenth-value layers (TVL). Its effect on patient dose was found to be real but clinically negligible: below 1.4%, restricted to distal cortical bone structures outside the treatment field, corresponding to a small increase in dose due to backscatter and with no statistical significance in the irradiated regions. It is concluded that the beam stopper acts essentially as a shielding element, with trivial influence on the therapeutic dose, and that the developed model constitutes a validated tool for dosimetric and radiation-protection studies in accelerators with an architecture similar to that of the Halcyon™.

2026-07-15T00:00:00Z https://repositorio.ufu.br/handle/123456789/49159 2026-07-30T06:26:09Z 2025-12-16T00:00:00Z

Title: Desenvolvimento e aprimoramento de Phantoms antropomórficos mamográficos Abstract: Considering the high cost, limited accessibility, and low reproducibility of commercially available mammographic phantoms, several studies have proposed alternative models using more accessible materials, such as the Maria Phantom. This model was originally designed to simulate different breast densities using layers of gel paraffin and PVC (polyvinyl chloride) films arranged randomly. However, the plates deformed over time and varied in size, resulting in irregular edges during image acquisition. To address these limitations, this study aimed to improve the Maria Phantom by reducing layer deformation using beeswax, standardizing plate geometry, and minimizing irregularities using 3D-printed molds. The new model was named “BeeMamma” Phantom. Gel paraffin and beeswax were tested as adipose-equivalent materials, with beeswax demonstrating greater durability, ease of handling, and radiological behavior like adipose tissue and gel paraffin. Plates representing different breast densities were fabricated by varying PVC film quantities, and additional plates containing simulated microcalcifications and 3D-printed nodules were incorporated. BeeMamma Phantom images were acquired on different mammography systems and analyzed using texture descriptors, including the 14 Haralick features, Skewness, and Kurtosis. Initial analyses showed that Beeswax presents radiographic characteristics comparable to adipose tissue and offers improved structural stability during manufacturing and storage. Attribute analysis indicated that BeeMamma Phantom reproduces breast tissues of varying densities effectively, preserving the textural characteristics observed in the Maria Phantom while overcoming limitations related to shape irregularities, deformation and handling difficulties. Furthermore, BeeMamma Phantom exhibited texture patterns like those of real mammographic images, supporting its suitability for testing new image-processing techniques and validating CAD and CADx schemes.

2025-12-16T00:00:00Z https://repositorio.ufu.br/handle/123456789/49123 2026-07-28T06:21:53Z 2026-07-20T00:00:00Z

Title: Marcapasso na radioterapia mamária: caracterização computacional das alterações dosimétricas Abstract: Cancer is one of the leading causes of death worldwide, and breast cancer is the second most frequent type and the most common among women. Population aging and unhealthy lifestyle habits contribute to the development of cancer and cardiovascular diseases (CVD), which often make the implantation of a cardiac electronic device necessary. Thus, patients with breast cancer and an implanted pacemaker may undergo radiotherapy, which is one of the treatment modalities for breast cancer. It is well established that pacemakers contain radiation-sensitive components and, because they are composed of metallic structures, may generate radiation scattering when directly exposed to the primary treatment beam. Therefore, the present study aims to analyze the dosimetric influence of the pacemaker on dose coverage in the breast, as well as to determine the absorbed dose by the device under the scenarios investigated. The Monte Carlo code TOPAS was used to simulate a breast cancer radiotherapy scenario with tangential fields on a TrueBeam linear accelerator using a 6 MV photon beam. The TET2DICOM software was used to convert the female mesh-type anthropomorphic reference phantom (MRCP) from ICRP Publication 145 into a set of DICOM images for implementation in TOPAS. Dose–volume histograms were generated, and the main dose metrics were extracted using the 3D Slicer software. The results showed that the presence of the pacemaker has a negligible impact on dose delivery to the treatment target volume. The prescription dose, D95%, was slightly reduced from 47.64 to 47.44 Gy, while the mean absorbed dose to the breast remained unchanged. Collimating the pacemaker as a strategy to reduce the absorbed dose to the device decreased its dose from 26.04 to 1.58 Gy; however, it also reduced breast dose coverage (D95% = 17.56 Gy), making this approach of limited clinical effectiveness. Advanced techniques, such as pacemaker-sparing Volumetric Modulated Arc Therapy (VMAT), should be investigated in future studies.

2026-07-20T00:00:00Z https://repositorio.ufu.br/handle/123456789/48950 2026-07-22T06:27:49Z 2025-12-17T00:00:00Z

Title: Caracterização de imagens de tecido mamário felino corado com Picrosirius Red Abstract: Feline mammary carcinoma represents the third most frequent tumor in queens. It exhibits aggressive behavior and a high mortality rate due to its invasive and metastatic pattern, making it a relevant model for comparative studies with breast cancer. Stromal alterations, such as increased collagen deposition, are associated with tumor aggressiveness and can be evaluated using histological techniques such as Picrosirius Red (PSR), complementing the routine Hematoxylin–Eosin (H&E) staining used for histopathological diagnosis. Computational methods, including feature extraction and artificial intelligence, enable image analysis to improve diagnostic accuracy. In this context, the present study aimed to classify images of healthy and tumoral feline mammary tissue stained with PSR according to histopathological grading. To this end, methods involving collagen quantification in the images were employed, as well as image characterization using Haralick texture descriptors and statistical analysis, with Gaussian curves applied for feature selection. Finally, machine learning models based on Random Forest (RF), XGBoost (XGB), and Multilayer Perceptron (MLP) were used with the selected features. Seven features were selected to classify healthy tissue and tumor images, and five features were selected to distinguish low- and high-grade tumors. The MLP model achieved the best performance in classifying healthy tissue (26 of 28 images) and tumor tissue (30 of 30 images), reaching 96.4% accuracy on the test set, whereas XGB demonstrated superior performance in differentiating low-grade (30 of 38 images) and high-grade tumors (36 of 27 images), achieving 88.0% accuracy. Skewness and kurtosis contributed to the classification between healthy and altered tissue images but were not effective in distinguishing low- and high-grade carcinogenic tissue. A subset of Haralick descriptors contributed to both classification tasks, mainly due to their ability to capture the heterogeneity of tumor tissues regarding collagen deposition during carcinogenesis, in contrast to the homogeneity of thinner collagen fibers in healthy feline mammary tissue. These results reinforce the potential of texture analysis combined with artificial intelligence methods as a complementary tool to histopathological evaluation and for improving diagnostic accuracy in veterinary oncology. With further refinements, the same approach may also assist in the diagnosis of breast cancer.

2025-12-17T00:00:00Z