Navegando por Data de Publicação, começando com "2026-03-02"

Filtrar resultados por ano ou mês

Agora exibindo 1 - 10 de 10

A sarcopenia e seus estágios são fatores de risco para declínio do desempenho cognitivo em pessoas idosas?
(Universidade Federal de São Carlos, 2026-03-02) Santos, Jassiely Priscilla de Faria; Alexandre, Tiago da Silva; https://lattes.cnpq.br/5393622641681701; https://orcid.org/0000-0003-3791-9793; https://lattes.cnpq.br/0204064617243377; https://orcid.org/0009-0009-2036-9618; Andrade, Larissa Pires de; Pereira, Daniele Sirineu; https://lattes.cnpq.br/7334391500340646; https://lattes.cnpq.br/2171027311100046
The muscle–brain axis has emerged as a central mechanism underlying the interdependence between sarcopenia and cognitive decline, suggesting that muscle integrity influences neurobiology and the maintenance of cognitive function. Although cross-sectional evidence links losses in muscle strength and mass to poorer cognitive performance, longitudinal research remains limited in demonstrating a gradient of risk across the severity stages of sarcopenia. Moreover, it remains unclear how variability in the criteria used to define low muscle strength (LMS) modulates these associations over time. Thus, this study examined sarcopenia and its stages as risk factors for decline in global cognitive function, memory, and executive function over eight years of follow-up, considering different cut-off points to define LMS (<36, <32, <30, <27, and <26 kg for men; <23, <21, <20, and <16 kg for women). This longitudinal study used data from the English Longitudinal Study of Ageing and followed 4,301 participants aged ≥60 years, without dementia and with baseline cognitive performance (z-score) ≥ −1.5 SD. Sarcopenia was defined according to EWGSOP2 criteria using different LMS cut-offs. Global cognition, memory, and executive function were standardized as z-scores by age and education. Generalized linear mixed models adjusted for sociodemographic, behavioral, and clinical characteristics were used to estimate cognitive trajectories according to sarcopenia stages. Only severe sarcopenia was a risk factor for decline in global cognitive function and its domains, regardless of the LMS cut-off adopted. For global cognition, this corresponded to a decline ranging from −0.92 to −0.69 SD over eight years, depending on the criterion. For executive function, the decline ranged from −0.38 to −0.18 SD, and for memory from −0.30 to −0.11 SD. Severe sarcopenia was consistently associated with decline in global cognition, memory, and executive function in older adults. Because this risk was identified independently of the LMS threshold applied, higher cut-off values may help identify individuals at risk of cognitive decline at earlier stages, enabling more timely interventions.
Segmentação semântica fracamente supervisionada para imagens de fast-food usando arquitetura SegFormer com duplo estudante
(Universidade Federal de São Carlos, 2026-03-02) Carvalho, Matheus Ramos de; Ferrari, Ricardo José; https://lattes.cnpq.br/8460861175344306; https://orcid.org/0000-0003-1197-2553; https://lattes.cnpq.br/7433192248627231; https://orcid.org/0009-0007-8203-5096; Ferrari, Ricardo José; Valejo, Alan Demétrius Baria; Travençolo, Bruno Augusto Nassif; https://lattes.cnpq.br/8460861175344306; https://lattes.cnpq.br/9546164790189830; https://lattes.cnpq.br/2590427557264952
This study focuses on Semantic Segmentation (SS) of fast-food sandwich assembly images using Deep Learning and weakly supervised learning techniques. Semantic segmentation, which involves pixel-wise classification of the image, is a fundamental task in computer vision and relevant for quality control in franchise assembly. When dealing with food, challenges arise such as the high visual diversity of ingredients and class imbalance; moreover, existing models often rely on public datasets that do not reflect scenarios such as Closed-Circuit Television (CCTV) imagery, which is the focus of this work. Automating quality control by distinguishing ingredients at the pixel level contributes to standardization and consistency. Manual annotation for supervised training is costly and time-consuming, which justifies exploring weakly supervised models that require only image-level labels. We propose a weakly supervised SS model based on Class Activation Maps (CAMs) to identify ingredients in low-resolution images captured by CCTV. The architecture is lightweight and low-resource: it uses SegFormer-B0 in a dual-student setup with mutual supervision via cross pseudo-labels, and incorporates Test-Time Augmentation (TTA) and Dynamic Threshold Adjustment (DTA) for pseudo-label refinement. Results show mIoU of 43.9% (comparable to DuPL, 41.2%), with greater stability across folds (standard deviation 0.6% versus 6.8% for DuPL). The model also exhibits superior computational efficiency, using 33% less VRAM and achieving 1.5–5× higher throughput than DuPL, making it suitable for real-time deployment in resource-constrained environments.
A rasura de uma história única: espaço e trajetória de uma imigrante negra em Americanah, de Chimamanda Ngozi Adichie
(Universidade Federal de São Carlos, 2026-03-02) Gomes, Jussara Barbosa da Silva; Alavarce, Camila da Silva; https://lattes.cnpq.br/7151151054517258; https://lattes.cnpq.br/7166371043551564; https://orcid.org/0000-0002-5110-8252; Valentim, Jorge Vicente; Barbosa, Sara Rogéria Santos; https://lattes.cnpq.br/4427303771174064; https://lattes.cnpq.br/8830209407070251
This work aims to analyze the novel Americanah, by Nigerian author Chimamanda Ngozi Adichie. The narrative revolves around a young Nigerian couple, Ifemelu and Obinze. After completing high school, both find themselves unable to pursue higher education in Nigeria. Therefore, they decide to leave the country, their destinations being the United States and England, respectively. However, this research will focus solely on Ifemelu. The intention is to understand the relationship between the character, a Black immigrant woman, and the spaces presented in the work, considering the American sociocultural context. Furthermore, it aims to objectively apprehend the conflicts faced by Ifemelu, permeated by racism, sexism, and xenophobia; these oppressions are active in marginalization, especially through language and her hair, central elements in her identity. This research is bibliographic in nature, and in the analysis, we take as fundamental the contributions of theorists of literary criticism, such as Mata (2014, 2020, 2023), Leite (2003), Achebe (2012), Candido (2006), Souza (2021), Morrison (2019), Seligmann-Silva (2018), Pereira (2023) and Gomes (2008), as well as intellectuals linked to the so-called Cultural Studies, among them Bhabha (1998, 2011), Hall (1997, 2003, 2016), Mignolo (2017), and Said (2011). Regarding intersectionality, a concept that will allow us to understand how oppressions operate together in the violence that affects the character Ifemelu in American society, we will focus on Gonzalez (2020), Collins (2019), hooks (2019, 2021), Davis (2016), Alzandúa (2000, 2005), Oyěwùmí (2021), Akotirene (2019), Berth (2020), Nascimento (2022) and Gomes (2019). In addition to these theorists, Fanon (2008, 2022), Munanga (2020, 2008), Mbembe (2018), Césaire (2010) and Santos (2015, 2023) will be important for this work, as their studies are fundamental to understanding spatial segregation and the violence that has permeated Black bodies from the colonization process to the present day. This work will also be guided by the notion of the danger of a single story (Adichie, 2019), as it will allow us to deconstruct eurocentrism surrounding the issues raised throughout the work. Therefore, the aim is to highlight how the literature produced by Chimamanda Ngozi Adichie erases the single narratives about certain bodies and reveals other ways of being in the world.
Promoção da saúde mental e qualidade de vida do residente médico de psiquiatria: revisão integrativa
(Universidade Federal de São Carlos, 2026-03-02) Melo, Fellipe Oliveira; Rocha, Sheyla Ribeiro; https://lattes.cnpq.br/4784568943530426; https://orcid.org/0000-0003-1926-4439; Aquilante, Aline Guerra; https://lattes.cnpq.br/5699716799010108; https://orcid.org/0000-0001-5925-7642; https://lattes.cnpq.br/8198235037188111; https://orcid.org/0000-0001-6807-2391; Soeiro, Everton; Barbosa, Larissa Campagna Martini; https://lattes.cnpq.br/0512926084072811; https://lattes.cnpq.br/8969823407559809
Introduction: Medical residency, characterized by on-the-job training, is the gold standard of specialization for medical graduates. However, this type of postgraduate training can be associated with feelings of insecurity, uncertainty, and frustration, linked to the high level of dedication, growth, and maturation required. If this process is not counterbalanced in an empathetic and supervised manner, it can be traumatic and lead to mental health disorders. Added to this are the peculiar characteristics of psychiatric residency, such as treating the mental suffering of another human being and its challenges, such as exposure to violence, suicide attempts and acts, empathic suffering, and stigmatization. Objective: To identify and propose strategies to promote the quality of life and mental health of psychiatric residents. Method: Integrative review, conducted in electronic databases of scientific publications—Medline/PubMed, SciELO, SCOPUS, Web of Science, ERIC, Virtual Health Library/BIREME (VHL), and Cochrane Library—and in the gray literature – CAPES Thesis Database, Brazilian Digital Library of Theses and Dissertations (BDTD), and Networked Digital Library of Theses and Dissertations (NDLTD), with selection using search strings by two independent researchers, with disagreements decided by a third party. Results: Of the 710 studies found, eight articles were selected, separated according to the type of modality used in the proposed intervention: virtual (online) and in-person. Remote interventions used cognitive behavioral therapy for suicide prevention and offered interventions through an app at strategic times of the day to improve behavior and mental health, both targeting first-year residents. In-person interventions focused on implementing wellness service strategies for resident physicians, organizational change strategies, post-suicide prevention protocols, meditation practices such as heartfulness and mindfulness, and electroacupuncture. Conclusion: Although studies show that the interventions found are low-cost and easily accessible, the studies do not show specific data such as financial values and reflect a reality that is often not found in countries such as Brazil. It is urgent to raise awareness among managers and the general public about this issue, in order to contribute to the development of public policies focused on care within the training settings for health professionals, which is closely linked to the field of public health.
Utilizando redes neurais informadas pela física para encontrar hamiltonianos efetivos em sistemas quânticos
(Universidade Federal de São Carlos, 2026-03-02) Silva, Rodrigo Pereira; Villas-Bôas, Celso Jorge; https://lattes.cnpq.br/5700887540085418; https://lattes.cnpq.br/1701487073029388; Dias, Rodrigo Alves; Richard, Victor Lopez; https://lattes.cnpq.br/3479575257091790; https://lattes.cnpq.br/4492188799583104
Effective Hamiltonians are a cornerstone of quantum optics and cavity quantum electro-dynamics (QED). By identifying and discarding terms that do not contribute significantly to the overall dynamics in a given parameter regime, they provide tractable models that still reproduce the experimentally relevant dynamics. In light-matter systems, a canonical example is the reduction of the quantum Rabi Hamiltonian (RH) to the effective Jaynes-Cummings Hamiltonian (JCH) under the rotating-wave approximation (RWA), where fast-oscillating terms are discarded from the system's Hamiltonian due to little contribution to dynamics. While effective-model selection is often justified analytically, as in the RWA, this approach requires a deep understanding of the physical system at hand, and is often heavily handmade. Hence, the central problem addressed in this dissertation is whether one can infer an effective Hamiltonian directly from limited dynamical data and physics a priori knowledge shared by many common physical systems. To this end, we leverage an inverse physics-informed neural network (PINN) framework to reproduce the passage of the RH to the JCH. In this setting, the complex-valued vector state of the composed atom-cavity system is represented by two real-valued neural networks (for the real and imaginary components) trained using a loss function with terms that (i) enforces Schrödinger dynamics through an ordinary differential equation (ODE) residual evaluated at fixed time instants, (ii) matches data observation from experimentally accessible expectation values (photon number and excited-state population), and (iii) imposes physical constraints through initial-condition and normalization penalties. Unknown atom-cavity coupling parameters present in the Hamiltonian are treated as trainable neural network parameters and learned jointly with the state trajectory. After validating the approach in a matched-model baseline (simulating train data with JCH and using the Jaynes-Cummings ODE residual) across distinct cavity initial states, we tackle effective-model identification by training with a split Rabi interaction containing separate rotating and counter-rotating couplings. Results show that in a parameter regime where RWA is feasible, training consistently recovers the Jaynes-Cummings coupling while suppressing the counter-rotating coupling toward zero, both when data are generated by full Rabi dynamics and when Jaynes-Cummings data are fit using the more expressive RH. We further show that the chosen non-negativity parameterization of the learned coupling parameter (absolute value versus softplus functions) influences how readily weakly identifiable terms collapse toward zero. These results support inverse PINNs as a practical workflow for testing whether candidate terms in full Hamiltonians are warranted by data and for extracting effective Hamiltonians from partial expectation-value measurements.
Capacitação aos profissionais da Atenção Primária à Saúde por meio da AMPI-AB para fortalecimento da Saúde da Pessoa Idosa: estudo quase-experimental
(Universidade Federal de São Carlos, 2026-03-02) Santos, Claudia Aparecida dos; Jesus, Isabela Thaís Machado de; https://lattes.cnpq.br/6262935093713981; https://orcid.org/0000-0002-3752-8867; Gramani-Say, Karina; http://lattes.cnpq.br/5096508613057074; https://orcid.org/0000-0002-2451-8109; https://lattes.cnpq.br/2072357265910195; https://orcid.org/0009-0008-9202-3758; Gratão, Aline Cristina Martins; Ramos, Nádia Placideli; https://lattes.cnpq.br/7873339343111119; https://lattes.cnpq.br/7300140470020309; https://orcid.org/0000-0002-8508-0251; https://orcid.org/0000-0002-8204-0668
Introduction: Multidimensional assessment of older adults (MDA) is an interdisciplinary process that contributes to health promotion. The Multidimensional Assessment of Older Adults in Primary Care (AMPI-AB) instrument allows for the identification of the functional stratification of older adults based on vulnerabilities, thus enabling the development of a longitudinal care plan. Although MDA is integrated into the Primary Health Care (PHC) level, its application is still limited. The qualification of professionals and the management of the work process are improved through continuing education strategies. Objective: To evaluate the level of knowledge of Primary Health Care professionals about AMPI-AB before and after a continuing education intervention and the recording of MDA results in ESUS-AB. Methodology: A quasi-experimental, single-group study was conducted before and after the intervention, developed in the PHC of a municipality in the interior of the state of São Paulo from August to December 2025. Health professionals were invited to participate in MDA and AMPI-AB training. Different professionals from family health and multidisciplinary teams who agreed to participate in the study were included, excluding non-permanent professionals and those who did not participate in the training. A knowledge level assessment questionnaire with four questions was applied at two time points (pre- and post-intervention), and a questionnaire evaluating the implementation and use of AMPI-AB was conducted in a 2-month follow-up after the training, with fourteen questions. The intervention consisted of theoretical-practical training and a simulated case study. Quantitative data were presented using descriptive statistics and compared using the Wilcoxon test for paired samples. Qualitative data were analyzed using content analysis. Results: 125 professionals participated in the training, 92.6% female, 61.6% Community Health Agents (CHAs), with an average of 7 years of experience in primary health care. The level of knowledge about AMPI-AB after training showed a significant increase in the median (2.00, 25%-75% pre vs. 4.00, 25%-75% post). During the subsequent months, 457 AMD records were made. Regarding the professionals' perception, most agreed that the training influenced changes in conduct and care support, recognizing the importance of AMPI-AB for the quality of care. The main barrier reported was lack of time, while facilitators included the receptiveness of older people. The sustainability of maintaining the use of the assessment during care was affirmative. Conclusion: The training highlighted the potential of continuing education as a strategy to qualify professionals, who came to recognize AMPI-AB as an accessible screening tool in the work process, in addition to contributing to care practices focused on longitudinal, comprehensive care centered on the needs of older people and assisting management in organizing the health network for older people from primary health care. Key words: Primary Health Care; Continuing Education; Evaluation of Processes in Health Care; Elderly Health.
Transferência de aprendizado profundo para previsão da irradiação solar: uma análise para diferentes bioclimas e escassez de dados
(Universidade Federal de São Carlos, 2026-03-02) Diniz, Leonardo; Barbosa, Daniel; https://lattes.cnpq.br/7843695359589199; https://orcid.org/0000-0002-0542-1723; Fernandes, Ricardo Augusto Souza; https://lattes.cnpq.br/0880243208789454; https://orcid.org/0000-0003-2361-6505; https://lattes.cnpq.br/6784659523460869; https://orcid.org/0009-0001-0657-5409; Machado, Vinicius Ponte; Palito, Thamyres Tamulla Cavalcante; https://lattes.cnpq.br/9385561556243194; https://lattes.cnpq.br/0882441206072709; https://orcid.org/0000-0003-3391-8443
The intermittent nature of solar irradiation has been a major challenge in planning and operating electrical power systems. In Brazil, due to the distinct bioclimatic zones, each with its own peculiarities and climatic behaviors, the sce- nario becomes complex. For this reason, a methodology based on transfer learning combined with state-of-the-art solar irradiation forecasting can bring advantages, especially when there is a low amount of data from certain climatological stations. In this sense, transfer learning can be used both between data from stations be- longing to the same bioclimatic zone (aiming to solve the problem of missing data) and between data from stations located in different bioclimatic zones (seeking to solve the problem of climatic disparity). This work proposes, based on a baseline constructed with the seasonal autoregressive integrated moving average model, to evaluate the performance of LSTM (long-short term memory) neural networks and their hybridization with convolutional neural networks (CNNs) in order to perform hourly solar irradiance forecasting. The results demonstrate that, in relation to the test of models constructed solely by LSTMs, compared to the CNN-LSTM hybrid models, the former showed slightly better performance, by approximately 1% on av- erage. Regarding the learning transfer tests, the feature extraction and fine-tuning approaches performed reasonably well when compared to zero-shot adaptation, be- ing comparable to the results of the LSTM models for each climatological station, especially in the configuration of stations by distance. Regarding the configuration of stations with a low amount of data, the approach performed well, mainly within the MAE metric.
Efeitos do ensino do ecoico generalizado na emergência de prosódia para crianças com Autismo
(Universidade Federal de São Carlos, 2026-03-02) Magalhães, Aline de Fátima Rosa; Goyos, Antônio Celso de Noronha; https://lattes.cnpq.br/3407008711535651; https://orcid.org/0000-0002-6150-1074; https://lattes.cnpq.br/0680686871478874; https://orcid.org/0009-0000-4815-1148; Goyos, Antonio Celso de Noronha; Costa, Maria da Piedade Resende da; Martone, Maria Carolina Correa; Garcia, Marcos Roberto; https://lattes.cnpq.br/3407008711535651; https://lattes.cnpq.br/2791098416503955; https://lattes.cnpq.br/4358052434870706; https://lattes.cnpq.br/1056949553338542; https://orcid.org/0000-0002-8746-4339; https://orcid.org/0000-0002-7420-5602; https://orcid.org/0000-0003-4595-9541; https://orcid.org/0000-0002-8746-4339
The objective of the study was to experimentally evaluate the effects of teaching generalized echoic verbal operants on the emergence of prosody in children with ASD. To this end, a study was conducted using a non-concurrent multiple baseline design between participants. The dependent variables (DVs) in this study correspond to echoic responses, direct teaching of multiple echoic samples, and prosody, the acquisition of which will not be taught directly, but only measured to verify possible side effects or generativity processes resulting from the teaching of multiple echoic samples. The independent variable (IV) consisted of the direct teaching of multiple examples of verbal auditory stimuli, implemented systematically throughout the sessions, investigating their effects on both the emergence and acquisition of prosodic repertoires. Five children diagnosed with ASD, aged between four and seven years, participated in the study. The results indicated that teaching multiple instances of echolalia produced consistent increases in prosodic performance percentages throughout the sessions. The observed effects can be explained by the formation of auditory-auditory equality relations and the transfer of control of this relation across new stimuli. Point-to-point correspondence expanded from words to new and socially relevant acoustic properties, and such changes have direct implications for the social function of language, including reading, writing, and literacy.
Saberes tradicionais nos cursos de Pedagogia das universidades públicas do Estado de São Paulo
(Universidade Federal de São Carlos, 2026-03-02) Mota, Maria Elizabete Pereira; Monteiro, Rosana Batista; https://lattes.cnpq.br/2567706339313166; https://orcid.org/0000-0003-1672-8912; https://lattes.cnpq.br/5598639076383646; https://orcid.org/0009-0000-2920-1306; Carril, Lourdes de Fátima Bezerra; Silva, Mariana Martha de Cerqueira; https://lattes.cnpq.br/0328730975862213; https://lattes.cnpq.br/6083262350667325; https://orcid.org/0000-0001-9054-6880; https://orcid.org/0000-0003-3445-7271
This dissertation investigates the presence and modes of incorporation of traditional quilombola and indigenous knowledge in the curricula of Pedagogy degree programs. The study is grounded in an anti-racist, decolonial, and contracolonial perspective, understanding the curricula as a field of political, symbolic, and epistemological disputes. Methodologically, this is a qualitative documentary research that analyzed the syllabi of mandatory courses on Education on Ethnic-Racial Relations (ERER), extracted from the Pedagogical Course Projects of public universities in the state of São Paulo. The results show that the incorporation of traditional knowledge occurs unevenly in the prescribed curricula. While indigenous knowledge appears explicitly named in all the syllabi analyzed, quilombola knowledge tends to be addressed implicitly, diluted within the field of Afro-Brazilian culture, being explicitly mentioned in only one of the courses. This asymmetry indicates epistemological limits in the recognition of this knowledge in initial teacher education. Although the presence of mandatory ERER courses constitutes a relevant condition for addressing these topics, it is observed that their mere existence does not guarantee the systematic valorization of traditional knowledge in the curricula. It is concluded that their effective incorporation requires political and epistemological choices that recognize them as legitimate forms of knowledge production, beyond their inclusion as complementary content, contributing to the construction of teacher education committed to anti-racist and contracolonial education.
Representações sociais da gestação e parto de mulheres que tiveram essa vivência durante a pandemia do COVID-19
(Universidade Federal de São Carlos, 2026-03-02) Costa, Kemily Benini; Ogata, Márcia Niituma; https://lattes.cnpq.br/3274294833403570; https://lattes.cnpq.br/1424644730814556; https://orcid.org/0000-0003-1410-7637; Bussadori, Jamile Claro de Castro; Melo, Flávio Adriano Borges; Binotto, Cibele Correia Semeão; Chirelli, Mara Quaglio; https://lattes.cnpq.br/4549657176930635; https://lattes.cnpq.br/2146448542549482; https://lattes.cnpq.br/2461303864889769; https://lattes.cnpq.br/8721545236460034
Objective: Analyze the social representations of pregnancy and childbirth for women who experienced this, during the COVID-19 pandemic. Method: Is a qualitative research study, composed of two stages. A systematic mapping review was conducted in March 2024, using the PubMed, Scopus, and Web of Science databases. After removing duplicates, the selection of studies was first carried out by analyzing the abstract and, subsequently by reading the full text. Excel spreadsheets were created to categorize the variables of the included studies, and the data were presented descriptively. Empirical data collection was carried out through semi-structured and audio-recorded interviews with women from three cities in the interior of São Paulo, who were pregnant and gave birth between 2020 and 2021. After transcription of the collected material, it was subjected to content analysis, resulting in categories of analysis based on the most frequent textual excerpts. Results: 65 primary studies on perceptions of pregnancy and childbirth during the COVID-19 pandemic were mapped, and 14 women who had these experiences during the aforementioned period were interviewed. The social representations of pregnancy during the COVID-19 pandemic were anchored in fear, sadness, and loneliness, where there was a discontinuity in their conceptual references of pregnancy, and therefore, a mismatch between expectations and reality related to the gestational period. Regarding childbirth, social representations were anchored in cesarean sections, obstetric violence, and unsuccessful experiences in vaginal delivery. Final considerations: The social representations of childbirth, compared to the pre-pandemic period, underwent significant transformations, with a shift in images from joy to fear, from sharing to loneliness and sadness, resulting in intense psychological suffering for women during pregnancy. Meanwhile, the social representations of childbirth remained aligned with the previously reified universe before the virus, to the traditions of birth and its generally consensual entirety, where being in labor usually means being subjected to a cesarean section or experiencing some degree of obstetric violence, where the covid-19 pandemic intensified vulnerabilities related to healthcare and the mother's satisfaction with the care received