Authors
Abstract(s)
A análise e classificação da condição física e da performance desportiva constituem uma necessidade emergente e cada vez mais requerida, daí que a recolha de informação relevante e o seu processamento sejam essenciais para apoiar uma gestão orientada e personalizada da dieta e do treino praticados pelos atletas. Neste trabalho, é proposta uma metodologia que permite realizar a monitorização dos atletas em point-of-care, baseada na técnica de espetroscopia de impedância elétrica (EIS), a partir da qual se pretende classificar o nível de esforço exigido durante uma sessão de treino, em função da dieta implementada e do tipo de treino realizado. Além disso, esta metodologia visa a avaliação dos efeitos isolado e combinado dos fatores referidos, com o objetivo de compreender que alterações ocorrem ao nível dos tecidos biológicos e dos fluídos corporais nos atletas de Crossfit sob análise, durante as sessões de treino.
A metodologia proposta inclui a avaliação de diferentes abordagens de processamento e a análise estatística dos dados obtidos após aplicação das mesmas, análise esta que permitiu aferir qual o processamento mais adequado para as avaliações a efetuar, qual o intervalo de frequência que permite obter resultados mais satisfatórios e, por fim, qual ou quais as frequências que permitem diferenciar as dietas implementadas e os treinos praticados e avaliar o efeito combinado destes fatores. O método discriminativo utilizado permitiu verificar que a autocorrelação do sinal no intervalo de 48000-52000 Hz permite diferenciar os momentos anterior e posterior ao treino com significância estatística e, apesar de se ter concluído que não havia interação entre a dieta e o treino praticados, foi possível detetar diferenças estatisticamente significativas entre as dietas, quando analisadas as componentes dos sinais de EIS adquiridos em todas as frequências avaliadas entre 49700 e 51800 Hz, assim como para a soma total da energia espectral enfatizada, e entre os treinos praticados, para todas as frequências contidas no intervalo de 48500-52000 Hz e para a soma total da energia enfatizada do sinal. Posto isto, o presente estudo veio demonstrar que, a partir dos espetros de impedância adquiridos pelo sensor AD5933, é possível discriminar as alterações nos valores de impedância induzidas pelo esforço exigido pelo treino e caracterizar os efeitos isolado e combinado, da dieta e do tipo de treino praticados.
The analysis and classification of physical condition and sports performance are an evolving and increasingly required need, hence the gathering of relevant information and its processing are essential to support a guided and customized diet and training management by the athletes. In this work, a methodology based on electrical impedance spectroscopy (EIS) that allows the point-of-care monitoring of these athletes is proposed, which is intended to classify the level of effort required on a training session, according to the implemented diet and the training. Furthermore, this methodology aims the evaluation of the isolated and combined effects of the referred factors, in order to understand which alterations occur in the biological tissues and body fluids of Crossfit athletes under analysis, during the training sessions. The proposed methodology includes an evaluation of different processing approaches and the statistic analysis of data, that allows the understanding about which is the most suitable processing method, which is the frequency range that comes whit the best results, and, finally, which are the frequencies that differentiate the implemented diets and the training, as well as the combined effect of both of these factors. The applied discriminative method found that the autocorrelation of the signal in the range of 48000-52000 Hz allows the differentiation of the previous and the latest moments of the training sessions with high statistical significance and, despite the fact that it has been concluded that there was no interaction between the diets and the type of training, it was possible to detect some statistically significant differences between diets when the signal components in the range of 48500-52000 Hz were analyzed, as for the sum of the total spectral energy, and between the training sessions for all the signal components in the tested range, as for the sum of total spectral energy. Therefore, the present study shows that it is possible to discriminate the alterations in the impedance data, induced by the effort demanded on a training session, and to characterize the isolated and combined effects of the hold diet and training, from the impedance spectra acquired using the AD5933 sensor.
The analysis and classification of physical condition and sports performance are an evolving and increasingly required need, hence the gathering of relevant information and its processing are essential to support a guided and customized diet and training management by the athletes. In this work, a methodology based on electrical impedance spectroscopy (EIS) that allows the point-of-care monitoring of these athletes is proposed, which is intended to classify the level of effort required on a training session, according to the implemented diet and the training. Furthermore, this methodology aims the evaluation of the isolated and combined effects of the referred factors, in order to understand which alterations occur in the biological tissues and body fluids of Crossfit athletes under analysis, during the training sessions. The proposed methodology includes an evaluation of different processing approaches and the statistic analysis of data, that allows the understanding about which is the most suitable processing method, which is the frequency range that comes whit the best results, and, finally, which are the frequencies that differentiate the implemented diets and the training, as well as the combined effect of both of these factors. The applied discriminative method found that the autocorrelation of the signal in the range of 48000-52000 Hz allows the differentiation of the previous and the latest moments of the training sessions with high statistical significance and, despite the fact that it has been concluded that there was no interaction between the diets and the type of training, it was possible to detect some statistically significant differences between diets when the signal components in the range of 48500-52000 Hz were analyzed, as for the sum of the total spectral energy, and between the training sessions for all the signal components in the tested range, as for the sum of total spectral energy. Therefore, the present study shows that it is possible to discriminate the alterations in the impedance data, induced by the effort demanded on a training session, and to characterize the isolated and combined effects of the hold diet and training, from the impedance spectra acquired using the AD5933 sensor.
Description
Keywords
Classificação Condição física Espetroscopia de impedância elétrica Performance desportiva Classification Electrical impedance spectroscopy Physical condition Sport performance