Loading...
11 results
Search Results
Now showing 1 - 10 of 11
- Análise de Sobrevivência em RPublication . Sousa-Ferreira, Ivo
- A review of Cox' s model extensions for multiple eventsPublication . Sousa-Ferreira, Ivo; Abreu, AnaIn longitudinal studies, it is usual that a given subject can experience several failures. To analyse multiple failure-time data, we reviewed some extensions of Cox's regression model, which were proposed by: Prentice, Williams and Peterson (PWP); Andersen and Gill (AG); Wei, Lin e Weissfeld (WLW); and Lee, Wei and Amato (LWA). Our main goal is to underline the differences between these extensions, through a brief but careful description, providing also some guidance on how to choose the proper model for each situation. The guidelines presented in this work revealed to be a useful pointer to easily choose the most suitable model. Secondarily, we used the survsim and the survival R packages to illustrate the practical implementation of these models.
- Hybrid model for recurrent event dataPublication . Sousa-Ferreira, Ivo; Abreu, Ana MariaIn the last four decades, there has been an increasing interest in developingsurvivalmodelsappropriateformultipleeventdataand,inparticular,for recurrentevent data. For these situations, several extensionsof the Cox’s regression model have been developed. Some of the most known models were suggested by: Prentice, Williams, and Peterson (PWP); Andersen and Gill (AG); Wei, Lin, and Weissfeld (WLW);andLee,Wei, andAmato(LWA).Thesemodelscanhandlewith situations where exist potentially correlated lifetimes of the same subject (due to the occurrence of more than one event for each subject) which is common in this type of data. In this chapter we present a new model, which we call hybrid model, with the purpose of minimizing some limitations of PWP model. With this model we obtained an improvement in the precision of the parameters estimates and a better fit to the simulated data.
- Modelos para acontecimentos múltiplosPublication . Ferreira, Ivo Miguel Sousa; Abreu, Ana Maria Cortesão Pais Figueira da SilvaAo longo dos últimos anos tem existido um interesse crescente em estudar o tempo até à ocorrência de vários acontecimentos, os quais podem ser observados mais do que uma vez para um mesmo indivíduo. Os dados respeitantes a acontecimentos múltiplos têm como principal característica o facto de se registar mais do que um tempo de vida para cada indivíduo, o que inviabiliza a aplicação direta do modelo de regressão de Cox. Assim, surgiu a necessidade de desenvolver novas extensões deste modelo, sendo que as mais utilizadas na prática foram sugeridas por: Prentice, Williams e Peterson (PWP); Andersen e Gill (AG); Wei, Lin e Weissfeld (WLW); e Lee, Wei e Amato (LWA). Um dos maiores obstáculos na aplicação destes modelos é a forte possibilidade de existir correlação intraindivíduos. Neste ponto, os quatro modelos referidos anteriormente são classificados como modelos marginais, uma vez que o vetor de parâmetros de regressão é estimado com base no ajustamento de um modelo que ignora a correlação entre acontecimentos. Para compensar esse facto, nestes modelos é usado um estimador robusto da matriz de covariância, o qual permite efetuar a correção necessária na estimativa da variância usual. Após realizar uma descrição detalhada de cada modelo marginal procedeu-se à respetiva implementação através do software estatístico R. Para o efeito, recorreu-se à simulação de dados relativos a acontecimentos múltiplos do mesma natureza, isto é, a acontecimentos recorrentes. Os resultados obtidos permitiram realçar e confirmar as características dos modelos estudados.
- Um olhar sobre o desenvolvimento de modelos de sobrevivência para acontecimentos recorrentesPublication . Sousa-Ferreira, Ivo; Abreu, Ana Maria; Rocha, Cristina
- A novel and simpler alkaline hydrolysis methodology for extraction of ferulic acid from brewer’s spent grain and its (partial) purification through adsorption in a synthetic resinPublication . Ideia, Pedro; Sousa-Ferreira, Ivo; Castilho, Paula C.This work aims to develop simpler methodologies of extracting ferulic acid (FA) from brewer’s spent grain (BSG). BSG is produced by brewing companies at high amounts all over the year and does not possess a direct application. Thus, its use as raw material for extraction of bioactive compounds has gained attention in the last years. FA has different interesting applications in cosmetics, food industry, and pharmaceutics. Several studies aim for its extraction from BSG by various methods, namely alkaline hydrolysis. In the present work, we suggest the use of autoclave to process higher amounts of BSG in a lab scale. A simplification of the regular post-hydrolysis procedures is also proposed to decrease the number of experimental steps and energy costs and to simultaneously increase the extraction yield (up to 470 mg of FA per 100 g of BSG). The adsorption of extracted FA in a synthetic resin is suggested as a partial purification method.
- Lipid biosignature of breast cancer tissues by matrix-assisted laser desorption/ionization time-of-flight mass spectrometryPublication . Silva, Catarina L.; Perestrelo, Rosa; Sousa-Ferreira, Ivo; Capelinha, Filipa; Câmara, José S.; Petković, MarijanaPurpose One of the hallmarks of cancer cells is the demand of supply for the synthesis of new membranes involved in cell proliferation and lipids have an important role in cellular structure, signaling pathways and progression of cancer. In this sense, lipid studies have become an essential tool allowing the establishment of signatures associated with breast cancer (BC). In this regard, some metabolic processes including proteins, nucleic acids and lipid synthesis are enhanced as part of cancer-associated metabolic reprogramming, as a requirement for cell growth and proliferation. Methods Pairwise samples of breast active carcinoma (BAC) and breast cancer-free tissues were collected from n=28 patients and analyzed by MALDI-TOF MS. Results Major lipid species are identifed in the MALDI-TOF mass spectra, with certain phosphatidylinositols (PIs) detect able only in BAC. Statistical analysis revealed signifcant diferences (p<0.05) between ratios lysophosphatidylcholine (LPC) 16:0/phosphatidylcholine (PC) 16:0_18:2 between AC and CF groups as well as for BC stages II and III. The ratio PC 16:0_18:2/PC16:0_18:1 was statistically diferent between AC and CF groups. The one-way ANOVA revealed that there are no statistical diferences among BC stages (I, II and III) within AC group. Comparing BC stages, the signifcance impact increased (p<0.05) with stage. Conclusion The obtained data revealed MALDI-TOF MS as a powerful tool to explore lipid signatures and the enzyme activity associated with BC and possibly establish novel disease markers.
- A review of velocity-type PSO variantsPublication . Sousa-Ferreira, Ivo; Sousa, DuarteThis paper presents a review of the particular variants of particle swarm optimization, based on the velocity-type class. The original particle swarm optimization algorithm was developed as an unconstrained optimization technique, which lacks a model that is able to handle constrained optimization problems. The particle swarm optimization and its inapplicability in constrained optimization problems are solved using the dynamic-objective constraint-handling method. The dynamic-objective constraint-handling method is originally developed for two variants of the basic particle swarm optimization, namely restricted velocity particle swarm optimization and self-adaptive velocity particle swarm optimization. Also on the subject velocity-type class, a review of three other variants is given, specifically: (1) vertical particle swarm optimization; (2) velocity limited particle swarm optimization; and (3) particle swarm optimization with scape velocity. These velocity-type particle swarm optimization variants all have in common a velocity parameter which determines the direction/movements of the particles.
- Parametric regression models for recurrent events analysis based on Chen distributionPublication . Sousa-Ferreira, Ivo; Abreu, Ana Maria; Rocha, CristinaIn this paper, two parametric regression models based on Chen dis tribution are proposed for situations where recurrent events have the same or dif ferent risks of occurrence. Inference is based on a maximum likelihood approach, which ensures consistent parameter estimators. However, since in recurrent event data there is within-subject correlation, the “one step” jackknife estimator is used. An application on a real data set is also provided for illustrative purposes.
- Modelos de sobrevivência aplicados à análise de acontecimentos múltiplosPublication . Sousa-Ferreira, Ivo; Abreu, Ana Maria; Rocha, CristinaNa modela¸c˜ao de acontecimentos m´ultiplos, uma abor dagem muita utilizada consiste em desenvolver extens˜oes do modelo semiparam´etrico de Cox. Contudo, quando se considera que o co nhecimento da distribui¸c˜ao do tempo ´e importante para o estudo, a abordagem param´etrica revela-se mais adequada. Neste trabalho, s˜ao apresentados dois novos modelos param´etricos baseados na dis tribui¸c˜ao de Weibull (modelos WNE e WE), como alternativas a duas das extens˜oes do modelo de Cox para acontecimentos recorren tes (modelos AG e PWP). De forma a ilustrar as diferen¸cas entre estas duas abordagens, ´e considerado um exemplo de aplica¸c˜ao com dados simulados. Os resultados obtidos com os modelos propostos permitem acrescentar mais duas op¸c˜oes ao leque de escolhas de mo delos para acontecimentos recorrentes.