RESUMEN
Parasitic capacitance represents the main error source in measurement systems based on electrical impedance spectroscopy. The capacitive nature of electrodes' impedance in tetrapolar configuration can give origin to phase errors when electrodes are coupled to parasitic capacitances. Nevertheless, reactive charges in tissue excitation systems are susceptible to instability. Based on such a scenario, mitigating capacitive effects associated with the electrode is a requirement in order to reduce errors in the measurement system. A literature review about the main compensation techniques for parasitic capacitance was carried out. The selected studies were categorized into three groups: (i) compensation in electronic instrumentation; (ii) compensation in measurement processing, and (iii) compensation by negative impedance converters. The three analyzed methods emerged as effective against fixed capacitance. No method seemed capable of mitigating the effects of electrodes' capacitance, that changes in the frequency spectrum. The analysis has revealed the need for a method to compensate varying capacitances, since electrodes' impedance is unknown.
Asunto(s)
Electricidad , Electrónica , Impedancia Eléctrica , Capacidad Eléctrica , ElectrodosRESUMEN
A análise da bioimpedância elétrica de tecidos humanos têm sido utilizada para identificar lesões precursoras ao câncer, auxiliando na prevenção e tratamento precoce. O presente estudo analisou 630 amostras de bioimpedância do tecido cervical de pacientes no estado de Santa Catarina, a fim de identificar lesões no colo do útero. Para tal, comparou-se o desempenho de dois algorítimos de classificação de padrões: Máquina de Vetores de Suporte (SVM) e Redes Neurais (RN) em conjunto com a Análise de Componentes Principais (PCA), separando as amostras entre tecido com lesão ou normal. Ambos classificadores apresentaram resultados satisfatórios. Pode-se concluir que os classificadores de padrões propostos apresentam uma poderosa alternativa para a classificação de tecidos.
The bioelectrical impedance analysis of human tissues have been used to identify early stages of cancer, helpingthe prevention and early treatment. This study analyzed 630 bioimpedance samples of cervical tissue of patients inthe state of Santa Catarina to identify lesions in the cervix. We compared the performance of two pattern recognition algorithms: Support Vector Machines and Neural Networks, in conjunction with Principal Component Analysis, separating the samples of tissue as injured or regular. Both classifiers showed satisfactory results.. It can be concluded that the proposed pattern classifiers might be a powerful alternative tool for tissue characterization.