Search

This paper presents a comparative analysis of novel supervised fuzzy adaptive resonance theory (SF-ART), multilayer perceptron (MLP) and Multi Layer Perceptrons (MLP) neural networks over Ballistocardiogram (BCG) signal recognition. To extract essential features of the BCG signal, we applied Biorthogonal wavelets. SF-ART performs classification on two levels. At first level, pre-classifier which is self-organized fuzzy ART tuned for fast learning classifies the input data roughly to arbitrary (M) classes. At the second level, post-classification level, a special array called Affine Look-up Table (ALT) with M elements stores the labels of corresponding input samples in the address equal to the index of fuzzy ART winner. However, in running (testing) mode, the content of an ALT cell with address equal to the index of fuzzy ART winner output will be read. The read value declares the final class that input data belongs to. In this paper, we used two well-known patterns (IRIS and Vowel data) and a medical application (Ballistocardiogram data) to evaluate and check SF-ART stability, reliability, learning speed and computational load. Initial tests with BCG from six subjects (both healthy and unhealthy people) indicate that the SF-ART is capable to perform with a high classification performance, high learning speed (elapsed time for learning around half second), and very low computational load compared to the well-known neural networks such as MLP which needs minutes to learn the training material. Moreover, to extract essential features of the BCG signal, we applied Biorthogonal wavelets. The applied wavelet transform requires no prior knowledge of the statistical distribution of data samples.

In this paper we present a new method for detection of spiking events caused by the increased respiratory resistance (IRR) from ballistocardiographic (BCG) data recorded with EMFi sheet. Spiking is a phenomenon where BCG wave complexes increase in amplitude during IRR. In this study data from six patients with a total of 1503 visually scored spiking events were studied. The algorithm monitors amplitude levels of BCG complexes and detects large relative increases. In this work 10 different variations of the algorithm were compared in order to find the best variation, which can cope with different recordings. The best variation of the algorithm was able to detect spiking events with 80% true positive and 19% false positive rates. The detection is not dependent on absolute waveform amplitudes and therefore does not require any recording-specific tuning prior to application. It is important to recognize spiking events in order to evaluate the severity of respiratory disturbance during sleep.

The heart disease diagnosing (HDD) system consists of a sensitive movement EMFi™-film 1 [1] EMFi is a registered trademark of Emfit Ltd. sensor installed under the upholstery of a chair. Whilst a man rests on the chair, this sensor which is sensitive to force gives us a single electrical signal containing components reflective of cardiac-ballistocardiogram (BCG), respiratory, and body movements related motion. Among different measurements of body activities, BCG has the interesting property that no electrodes are needed to be attached to the body during recording, suitable to evaluate man heart condition in any place such as home, car, or his office. This paper describes briefly our developed HDD system and especially a combined intelligent signal processing method to detect, extract features and finally cluster BCG cycles for assisting medical doctors to diagnose heart diseases of person under test. Indeed, it is a fully automatic system which is not very sensitive to the BCG latency as well as non-linear disturbance. It uses high resolution Biorthogonal wavelet transforms to extract essential BCG features and to cluster those using artificial neural networks (ANNs). Some evaluations using recordings from normal young, normal old and abnormal old volunteers indicated that our combined method is reliable and has high performance. [Copyright &y& Elsevier]