Ballistocardiogram classification using a novel transform so-called AliMap and biorthogonal wavelets

This paper presents a new kind of mapping so-called AliMap for signal processing. This map tries to eliminate redundant information without losing relevant information incorporated in spatial, time, and frequency domains. i.e., this map is able to extract most important information and quantify thorn using scalar values, mapping from a high dimensional space to one scalar value. It has three factors to decide which information of input data is most important. This transform can be used for automatic pattern classification. In this study, we applied Haar wavelet transform to extract essential features of the ballistocardiogram (BCG) signal and AliMap to classify the BCC. Initial tests with BCG from 18 subjects (both healthy and unhealthy people) indicate that the method can classify the subjects into three classes with a high accuracy, compared with the well-known method called Starr classification of BCG. The method is insensitive to latency and non-linear disturbance. Moreover, the applied wavelet transform requires no prior knowledge of the statistical distribution of data samples.