Percorrer por autor "Marques, Nuno"
A mostrar 1 - 10 de 15
Resultados por página
Opções de ordenação
- Detection of voicing and place of articulation of fricatives with deep learning in a virtual speech and language therapy tutorPublication . Anjos, Ivo; Maxine, Eskenazi; Marques, Nuno; Grilo, Ana Margarida; Guimarães, Isabel; Magalhães, João; Cavaco, SofiaChildren with fricative distortion errors have to learn how to correctly use the vocal folds, and which place of articulation to use in order to correctly produce the different fricatives. Here we propose a virtual tutor for fricatives distortion correction. This is a virtual tutor for speech and language therapy that helps children understand their fricative production errors and how to correctly use their speech organs. The virtual tutor uses log Mel filter banks and deep learning techniques with spectral-temporal convolutions of the data to classify the fricatives in children’s speech by place of articulation and voicing. It achieves an accuracy of 90:40% for place of articulation and 90:93% for voicing with children’s speech. Furthermore, this paper discusses a multidimensional advanced data analysis of the first layer convolutional kernel filters that validates the usefulness of performing the convolution on the log Mel filter bank.
- Determinants of HIV late presentation among men who have sex with men in Portugal (2014–2019): who’s being left behind?Publication . Abrantes, Ricardo; Pimentel, Victor; Miranda, Mafalda N. S.; Silva, Ana Rita; Diniz, António; Ascenção, Bianca; Piñeiro, Carmela; Koch, Carmo; Rodrigues, Catarina; Caldas, Cátia; Morais, Célia; Faria, Domitília; Gomes da Silva, Elisabete; Teófilo, Eugénio; Monteiro, Fátima; Roxo, Fausto; Maltez, Fernando; Rodrigues, Fernando; Gaião, Guilhermina; Ramos, Helena; Costa, Inês; Germano, Isabel; Simões, Joana; Oliveira, Joaquim; Ferreira, José; Poças, José; Saraiva da Cunha, José; Soares, Jorge; Fernandes, Sandra; Mansinho, Kamal; Pedro, Liliana; Aleixo, Maria João; Gonçalves, Maria João; Manata, Maria José; Mouro, Margarida; Serrado, Margarida; Caixeiro, Micaela; Marques, Nuno; Costa, Olga; Pacheco, Patrícia; Proença, Paula; Rodrigues, Paulo; Pinho, Raquel; Tavares, Raquel; Correia de Abreu, Ricardo; Côrte-Real, Rita; Serrão, Rosário; Sarmento e Castro, Rui; Nunes, Sofia; Faria, Telo; Baptista, Teresa; Simões, Daniel; Mendão, Luís; Martins, M. Rosário O.; Gomes, Perpétua; Pingarilho, Marta; Abecasis, Ana B.Introduction: HIV late presentation (LP) remains excessive in Europe. We aimed to analyze the factors associated with late presentation in the MSM population newly diagnosed with HIV in Portugal between 2014 and 2019. Methods: We included 391 newly HIV-1 diagnosed Men who have Sex with Men (MSM), from the BESTHOPE project, in 17 countrywide Portuguese hospitals. The data included clinical and socio-behavioral questionnaires and the viral genomic sequence obtained in the drug resistance test before starting antiretrovirals (ARVs). HIV-1 subtypes and epidemiological surveillance mutations were determined using different bioinformatics tools. Logistic regression was used to estimate the association between predictor variables and late presentation (LP). Results: The median age was 31 years, 51% had a current income between 501-1,000 euros, 28% were migrants. 21% had never been tested for HIV before diagnosis, with 42.3% of MSM presenting LP. 60% were infected with subtype B strains. In the multivariate regression, increased age at diagnosis, higher income, lower frequency of screening, STI ever diagnosed and higher viral load were associated with LP. Conclusion: Our study suggests that specific subgroups of the MSM population, such older MSM, with higher income and lower HIV testing frequency, are not being targeted by community and clinical screening services. Overall, targeted public health measures should be strengthened toward these subgroups, through strengthened primary care testing, expanded access to PrEP, information and promotion of HIV self-testing and more inclusive and accessible health services.
- HIV-1-Transmitted Drug Resistance and Transmission Clusters in Newly Diagnosed Patients in Portugal Between 2014 and 2019Publication . Pingarilho, Marta; Pimentel, Victor; Miranda, Mafalda N. S.; Silva, Ana Rita; Diniz, António; Ascenção, Bianca Branco; Piñeiro, Carmela; Koch, Carmo; Rodrigues, Catarina; Caldas, Cátia; Morais, Célia; Faria, Domitília; da Silva, Elisabete Gomes; Teófilo, Eugénio; Monteiro, Fátima; Roxo, Fausto; Maltez, Fernando; Rodrigues, Fernando; Gaião, Guilhermina; Ramos, Helena; Costa, Inês; Germano, Isabel; Simões, Joana; Oliveira, Joaquim; Ferreira, José; Poças, José; da Cunha, José Saraiva; Soares, Jorge; Henriques, Júlia; Mansinho, Kamal; Pedro, Liliana; Aleixo, Maria João; Gonçalves, Maria João; Manata, Maria José; Mouro, Margarida; Serrado, Margarida; Caixeiro, Micaela; Marques, Nuno; Costa, Olga; Pacheco, Patrícia; Proença, Paula; Rodrigues, Paulo; Pinho, Raquel; Tavares, Raquel; de Abreu, Ricardo Correia; Côrte-Real, Rita; Serrão, Rosário; Castro, Rui Sarmento e; Nunes, Sofia; Faria, Telo; Baptista, Teresa; Martins, Maria Rosário O.; Gomes, Perpétua; Mendão, Luís; Simões, Daniel; Abecasis, AnaObjective: To describe and analyze transmitted drug resistance (TDR) between 2014 and 2019 in newly infected patients with HIV-1 in Portugal and to characterize its transmission networks. Methods: Clinical, socioepidemiological, and risk behavior data were collected from 820 newly diagnosed patients in Portugal between September 2014 and December 2019. The sequences obtained from drug resistance testing were used for subtyping, TDR determination, and transmission cluster (TC) analyses. Results: In Portugal, the overall prevalence of TDR between 2014 and 2019 was 11.0%. TDR presented a decreasing trend from 16.7% in 2014 to 9.2% in 2016 (p for-trend = 0.114). Multivariate analysis indicated that TDR was significantly associated with transmission route (MSM presented a lower probability of presenting TDR when compared to heterosexual contact) and with subtype (subtype C presented significantly more TDR when compared to subtype B). TC analysis corroborated that the heterosexual risk group presented a higher proportion of TDR in TCs when compared to MSMs. Among subtype A1, TDR reached 16.6% in heterosexuals, followed by 14.2% in patients infected with subtype B and 9.4% in patients infected with subtype G. Conclusion: Our molecular epidemiology approach indicates that the HIV-1 epidemic in Portugal is changing among risk group populations, with heterosexuals showing increasing levels of HIV-1 transmission and TDR. Prevention measures for this subpopulation should be reinforced.
- HIV-1-transmitted drug resistance and transmission clusters in newly diagnosed patients in Portugal between 2014 and 2019Publication . Pingarilho, Marta; Pimentel, Victor; Miranda, Mafalda N. S.; Silva, Ana Rita; Diniz, António; Ascenção, Bianca Branco; Piñeiro, Carmela; Koch, Carmo; Rodrigues, Catarina; Caldas, Cátia; Morais, Célia; Faria, Domitília; Silva, Elisabete Gomes da; Teófilo, Eugénio; Monteiro, Fátima; Roxo, Fausto; Maltez, Fernando; Rodrigues, Fernando; Gaião, Guilhermina; Ramos, Helena; Costa, Inês; Germano, Isabel; Simões, Joana; Oliveira, Joaquim; Ferreira, José; Poças, José; Cunha, José Saraiva da; Soares, Jorge; Henriques, Júlia; Mansinho, Kamal; Pedro, Liliana; Aleixo, Maria João; Gonçalves, Maria João; Manata, Maria José; Mouro, Margarida; Serrado, Margarida; Caixeiro, Micaela; Marques, Nuno; Costa, Olga; Pacheco, Patrícia; Proença, Paula; Rodrigues, Paulo; Pinho, Raquel; Tavares, Raquel; Abreu, Ricardo Correia de; Côrte-Real, Rita; Serrão, Rosário; Castro, Rui Sarmento e; Nunes, Sofia; Faria, Telo; Baptista, Teresa; Martins, Maria Rosário O.; Gomes, Perpétua; Mendão, Luís; Simões, Daniel; Abecasis, Ana; on behalf of the BESTHOPE Study GroupObjective: To describe and analyze transmitted drug resistance (TDR) between 2014 and 2019 in newly infected patients with HIV-1 in Portugal and to characterize its transmission networks. Methods: Clinical, socioepidemiological, and risk behavior data were collected from 820 newly diagnosed patients in Portugal between September 2014 and December 2019. The sequences obtained from drug resistance testing were used for subtyping, TDR determination, and transmission cluster (TC) analyses. Results: In Portugal, the overall prevalence of TDR between 2014 and 2019 was 11.0%. TDR presented a decreasing trend from 16.7% in 2014 to 9.2% in 2016 (pfor–trend = 0.114). Multivariate analysis indicated that TDR was significantly associated with transmission route (MSM presented a lower probability of presenting TDR when compared to heterosexual contact) and with subtype (subtype C presented significantly more TDR when compared to subtype B). TC analysis corroborated that the heterosexual risk group presented a higher proportion of TDR in TCs when compared to MSMs. Among subtype A1, TDR reached 16.6% in heterosexuals, followed by 14.2% in patients infected with subtype B and 9.4% in patients infected with subtype G. Conclusion: Our molecular epidemiology approach indicates that the HIV-1 epidemic in Portugal is changing among risk group populations, with heterosexuals showing increasing levels of HIV-1 transmission and TDR. Prevention measures for this subpopulation should be reinforced.
- Hyperparameter Optimization of a Convolutional Neural Network Model for Pipe Burst Location in Water Distribution NetworksPublication . Antunes, André; Ferreira, Bruno; Marques, Nuno; Carriço, NelsonThe current paper presents a hyper parameterization optimization process for a convolutional neural network (CNN) applied to pipe burst locations in water distribution networks (WDN). The hyper parameterization process of the CNN includes the early stopping termination criteria, dataset size, dataset normalization, training set batch size, optimizer learning rate regularization, and model structure. The study was applied using a case study of a real WDN. Obtained results indicate that the ideal model parameters consist of a CNN with a convolutional 1D layer (using 32 filters, a kernel size of 3 and strides equal to 1) for a maximum of 5000 epochs using a total of 250 datasets (using data normalization between 0 and 1 and tolerance equal to max noise) and a batch size of 500 samples per epoch step, optimized with Adam using learning rate regularization. This model was evaluated for distinct measurement noise levels and pipe burst locations. Results indicate that the parameterized model can provide a pipe burst search area with more or less dispersion depending on both the proximity of pressure sensors to the burst or the noise measurement level.
- Intervenções de enfermagem pré-hospitalar : revisão narrativaPublication . Mota, Mauro; Cunha, Madalena; Reis Santos, Margarida; Cunha, Isabel Cristina; Alves, Mónica; Marques, NunoConstruir algoritmos de intervenção de enfermagem pré-hospitalar para vítimas de trauma. Metodologia: Revisão Narrativa da Literatura, entre 2008 e 2019, nas principais bases de dados. Dois revisores independentes realizaram a avaliação crítica, extração e síntese dos dados. A construção dos algoritmos resultou do processo interpretativo da revisão narrativa por três peritos na área. Utilizou-se o modelo teórico de Virgínia Henderson. Resultados: Obtiveram-se 17 documentos, seis foram incluídos no desenvolvimento dos metaparadigmas Saúde, Pessoa e Ambiente e 16 na elaboração e construção de Algoritmos de avaliação, diagnóstico e intervenções de enfermagem às vítimas de trauma. Conclusões: A revisão possibilitou a operacionalização do modelo teórico de Henderson para a assistência pré-hospitalar permitindo a criação de algoritmos orientadores da prática de enfermagem.
- A obstenção em Portugal como fenómeno de não consumo político: diagnóstico das causas numa perspetiva de marketingPublication . Marques, Nuno; Ferreira, Mafalda
- Sibilant consonants classification with deep neural networksPublication . Anjos, Ivo; Marques, Nuno; Grilo, Ana Margarida; Guimarães, Isabel; Magalhães, João; Cavaco, SofiaAbstract. Many children su ering from speech sound disorders cannot pronounce the sibilant consonants correctly. We have developed a serious game that is controlled by the children's voices in real time and that allows children to practice the European Portuguese sibilant consonants. For this, the game uses a sibilant consonant classi er. Since the game does not require any type of adult supervision, children can practice the production of these sounds more often, which may lead to faster improvements of their speech. Recently, the use of deep neural networks has given considerable improvements in classi cation for a variety of use cases, from image classication to speech and language processing. Here we propose to use deep convolutional neural networks to classify sibilant phonemes of European Portuguese in our serious game for speech and language therapy. We compared the performance of several diferent arti cial neural networks that used Mel frequency cepstral coefcients or log Mel lterbanks. Our best deep learning model achieves classi cation scores of 95:48% using a 2D convolutional model with log Mel lterbanks as input features.
- Sibilant consonants classification with deep neural networksPublication . Anjos, Ivo; Marques, Nuno; Grilo, Ana Margarida; Guimarães, Isabel; Magalhães, João; Cavaco, SofiaAbstract. Many children su ering from speech sound disorders cannot pronounce the sibilant consonants correctly. We have developed a serious game that is controlled by the children's voices in real time and that allows children to practice the European Portuguese sibilant consonants. For this, the game uses a sibilant consonant classi er. Since the game does not require any type of adult supervision, children can practice the production of these sounds more often, which may lead to faster improvements of their speech. Recently, the use of deep neural networks has given considerable improvements in classi cation for a variety of use cases, from image classication to speech and language processing. Here we propose to use deep convolutional neural networks to classify sibilant phonemes of European Portuguese in our serious game for speech and language therapy. We compared the performance of several diferent arti cial neural networks that used Mel frequency cepstral coefcients or log Mel lterbanks. Our best deep learning model achieves classi cation scores of 95:48% using a 2D convolutional model with log Mel lterbanks as input features.
- Simulating Price Interactions by Mining Multivariate Financial Time SeriesPublication . Silva, Bruno; Cavique, Luis; Marques, NunoThis position paper proposes a framework based on a feature clustering method using Emergent Self-Organizing Maps over streaming data (Ubi-SOM) and Ramex-Forum – a sequence pattern mining model for financial time series modeling based on observed instantaneous and long term relations over market data. The proposed framework aims at producing realistic monte-carlo based simulations of an entire portfolio behavior over distinct market scenarios, obtained from models generated by these two approaches.