RESUMO
OBJECTIVE: The aim of this work was to develop a novel technique for digital processing of Doppler ultrasound blood flow sensor data from noisy blood flow velocity waveforms. METHODS: To evaluate the fluctuating blood flow parameters, various nonlinear dynamics methods and algorithms are often being used. Here, for identification of chaotic and noise components in a fluctuating coronary blood flow, for the first time the Allan variance technique was used. Analysis of different types of noises (White, Brownian, Flicker) was carried out and their strong correlation with fractality of time series (the Hurst exponent) was revealed. RESULTS: Based on a specialized software realizing the developed technique, numerical experiments with real clinical data were carried out. Recommendations for identification of noisy patterns of coronary blood flow in normal and pathological states were developed. CONCLUSION: The methodology gives us the possibility for the more detailed quantitative and qualitative analysis of a noisy fluctuating blood flow data.