https://repositorio.ufba.br/handle/ri/34552 Mon, 04 May 2026 21:52:24 GMT 2026-05-04T21:52:24Z https://repositorio.ufba.br/handle/ri/41432 Título: Identificação de padrões de escoamentos multifásicos em sistemas core annular flow baseados em fluidodinâmica computacional utilizando agrupamentos fuzzy Autor(es): Lima, Patrick Souza Primeiro Orientador: Schnitman, Leizer Abstract: Petroleum, which is composed of a variety of chemical components, exhibits distinct physical and chemical characteristics, which requires the application of specific techniques to ensure adequate flow and avoid operational problems. Among these techniques, Core Annular Flow (CAF) stands out, as it reduces shear stresses by peripherally introducing a low-viscosity fluid. However, the injection of immiscible fluids can result in different multiphase flow patterns, which has been a challenge for the industry, which seeks to identify the flow pattern present in the system. To address this issue, a methodology was developed that combines Computational Fluid Dynamics (CFD) and artificial intelligence with fuzzy groupings. This approach allowed us to effectively identify different flow patterns. The simulations carried out were compared with experimental data already existing in the literature, demonstrating the validity of the proposed methodology. Furthermore, the application of fuzzy clustering enabled the detection of transition regions, providing a more detailed and continuous characterization of multiphase flow patterns, which expands the understanding of these phenomena in the industry. Editora / Evento / Instituição: Universidade Federal da Bahia Tipo: Dissertação Thu, 14 Mar 2024 00:00:00 GMT https://repositorio.ufba.br/handle/ri/41432 2024-03-14T00:00:00Z https://repositorio.ufba.br/handle/ri/40970 Título: Um estudo abrangente sobre segmentação de glomérulos com dados de treinamento limitados em imagens histopatológicas de alta resolução Autor(es): Souza Júnior, Luiz Otávio de Oliveira Primeiro Orientador: Oliveira, Luciano Rebouças de Abstract: The growing availability of scanned whole slide images (WSIs) has expanded digital pathology, enabling medical decision-making and computational analysis directly from high-resolution images. Kidney disease diagnoses using WSIs rely on the analysis of specific tissue structures, and automatic analysis depends on accurately segmenting key components such as glomeruli, tubules, interstitium, and vessels. This thesis focuses on glomeruli, which are essential in assessing WSIs after kidney biopsies. These structures are impacted by lesions related to various diseases. In machine learning-based WSI analysis, glomeruli are often the first regions segmented to guide subsequent tasks. The Bowman’s capsule (BC) is crucial, marking the boundary between glomerular components and surrounding interstitial tissue. This work proposes two studies aimed at addressing the segmentation of glomeruli in high-resolution kidney histopathological images. In the first study, we investigate the feasibility of segmenting glomeruli in human WSIs using deep-learning models trained exclusively on mouse data. Mice and humans share several biological similarities, including genetic, physiological, and structural characteristics, making mice a common model for studying human diseases. While this cross-species knowledge transfer is well-established in medicine, it remains underexplored in computational pathology, where WSIs serve as primary research objects. To address this gap, we evaluated five semantic segmentation models: U-Net, U-Net 3+, Res-U-Net, DeepLabV3+, and MA-Net, using datasets consisting of 18 mouse WSIs and 42 human WSIs. Among these, U-Net 3+ delivered the best performance in intra-dataset evaluation, achieving an average DICE score of 0.930 on HE-stained mouse images. On human data, U-Net 3+ also excelled, attaining DICE scores of 0.772, 0.824, and 0.791 on HE, PAS, and PAMS stains, respectively. Moreover, U-Net 3+ proved promising generalization when trained solely on mouse data and tested across the entire human dataset, achieving a DICE score of 0.798 on HE-stained images. However, while these models performed well on images within the same staining technique, their performance declined when applied across different stains, highlighting a limitation in cross-stain generalization. The second study focuses on the segmentation challenges posed by borderless glomeruli affected by global sclerosis. We developed an automated framework for patch cropping and stitching, eliminating manual intervention to streamline the segmentation process. Our experiments show that while standard segmentation models can achieve state-of-the-art results for normal and partially sclerotic glomeruli, their performance deteriorates significantly for globally sclerotic glomeruli. Notably, segmentation accuracy for these cases was highly dependent on the staining type and generally remained poor across models. We compared non-foundation models (U-Net, U-Net 3+, and SwinTransformer + U-Net) with and without fine-tuning against the SegGPT foundation model. Non-foundation models, trained exclusively on normal glomeruli with HE, PAS, and PAMS stains, achieved high performance on normal glomeruli (mDice > 0.92) and moderate performance on partially sclerotic glomeruli (mDice > 0.72). However, their performance dropped sharply to mDice > 0.02 for globally sclerotic glomeruli, with minimal improvements even after fine-tuning. In contrast, SegGPT demonstrated substantial improvement, achieving a significantly higher mDice score (> 0.37) for globally sclerotic glomeruli by leveraging only a few query samples. This result highlights the potential of foundation models in addressing segmentation challenges for glomeruli affected by severe lesions. In summary, the studies presented in this thesis represent a significant step forward in the segmentation of glomeruli in WSIs. Our findings offer a comprehensive analysis of glomerulus segmentation with limited training data, demonstrating the potential of mouse-to-human transfer learning, as well as the use of foundation models to improve segmentation accuracy for glomeruli affected by sclerosis. Editora / Evento / Instituição: Universidade Federal da Bahia Tipo: Tese Fri, 06 Dec 2024 00:00:00 GMT https://repositorio.ufba.br/handle/ri/40970 2024-12-06T00:00:00Z https://repositorio.ufba.br/handle/ri/40807 Título: Novo radiômetro solar em topologia fly-eye produzido por manufatura aditiva Autor(es): Gomes, Juliane Grasiela de Carvalho Primeiro Orientador: Pepe, Iuri Muniz Abstract: The use of solar generation has been increasingly widespread in Brazil in recent years due to it being a clean generation method, as well as the fact that Brazil is located in a region where solar irradiance is conducive to electricity production. For a photovoltaic system to perform well, the choice of location for its installation is crucial. This work proposes a solar radiometer with a fly-eye topology, resembling the eye of a fly. This technique, which draws inspiration from nature for technological development, is called biomimetics. The developed radiometer is of a reduced scale, allowing for greater mobility of the equipment and versatility of the structure since it is modeled using additive manufacturing and is easy to assemble and disassemble. It also includes a commercial pyranometer as a self-calibrator for the cells. With this equipment, it is possible to understand the dynamics of solar radiation during different times of the year and to confirm that the location where a photovoltaic generation system is intended to be installed is favorable for achieving the expected system performance. This equipment is an improvement on the work of student Lucas Barbosa da Silva. Editora / Evento / Instituição: Universidade Federal da Bahia Tipo: Dissertação Fri, 23 Aug 2024 00:00:00 GMT https://repositorio.ufba.br/handle/ri/40807 2024-08-23T00:00:00Z https://repositorio.ufba.br/handle/ri/40806 Título: SAMOM: uma arquitetura para operação de sistemas de manufatura com habilidades inteligentes e adaptativas Autor(es): Cruz, Jhaidan Ribeiro Primeiro Orientador: Lepikson, Herman Augusto Abstract: Manufacturing systems commonly require adequate management and control (M&C) for decision-making and consequent meeting of production demands. Systems prior to Industry 4.0 (traditional) are limited in terms of the flexibility required to adapt to emerging demands, such as process diversity and product customization. In these systems, the typical characteristics of dedicated automation focused on islands do little to favor adequate integration. As a result, M&C activities require large amounts of resources. Industry 4.0 (I4.0) brings concepts and technologies that enable new manufacturing systems (4.0 manufacturing systems) with greater flexibility and M&C efficiency. This paper aims to propose an architecture for integrated management and control of discrete and batch manufacturing systems with distributed characteristics and predictive and adaptive capabilities. To this end, the architecture uses enabling technologies and concepts from I4.0, focusing on Data Analysis, AI, Systems Integration and Service-Oriented Architecture (SOA). The proposed model unfolds into a SOA, based on Multi-Agent Systems (MAS) and Manufacturing Execution Systems (MES). It also describes components, interactions, concepts and implementation particularities of such architecture. The proposed model was validated in a pilot plant of 4.0 manufacturing, where it was possible to demonstrate that the devices of such system were managed and controlled, with the help of digital representations (agents), in a decentralized, distributed and heterarchical way. It was also possible to demonstrate the realization, by the MAS in question, of predictions of the availability of the devices and of decision-making about the production performance. It is concluded that the work proves to be a reference for the implementation of architectures compatible with the G&C of manufacturing systems 4.0. Editora / Evento / Instituição: Universidade Federal da Bahia Tipo: Dissertação Mon, 12 Aug 2024 00:00:00 GMT https://repositorio.ufba.br/handle/ri/40806 2024-08-12T00:00:00Z