Fuzzy oscillometric blood pressure classification

Classification of systolic, mean and diastolic blood pressure profiles using the oscillometric method is a difficult process. Generally, the algorithms aim at extracting some parameters such as height, and ratios of the pulses at certain pressure levels, which are obtained from the cuff pressure. These parameters can be used to form profiles to relate to blood pressures. The effectiveness of the classification depends on many factors, such as environmental noise, white coat effect, heart rate variability and motion artifacts. In this paper, we investigate the effectiveness of a neuro-fuzzy approach to blood pressure classification. We employ feature extraction using principal component analysis, and fuzzy sets to classify pressure profiles.