Biyomedikal işaretlerin sınıflandırılmasında otomatik teşhis sistemlerinin başarımı

Abstract

In this study, the automated diagnostic systems employing diverse and composite features for Doppler ultrasound signals, electroencephalogram (EEG) and electrocardiogram (ECG) signals were analyzed and their accuracies were determined. In pattern recognition applications, diverse features are extracted from raw data which needs recognizing. Combining multiple classifiers with diverse features are viewed as a general problem in various application areas of pattern recognition. The classification accuracies of multilayer perceptron neural network, combined neural network, and mixture of experts trained on composite feature and modified mixture of experts trained on diverse features were compared. The inputs of these automated diagnostic systems composed of diverse or composite features and were chosen according to the network structures. The conclusions of this study demonstrated that the modified mixture of experts trained on diverse features achieved accuracy rates which were higher than that of the other automated diagnostic systems trained on composite features.