Detecting variability of internal carotid arterial Doppler signals by Lyapunov exponents.

The new method presented in this study was directly based on the consideration that internal carotid arterial Doppler signals are chaotic signals. This consideration was tested successfully using the nonlinear dynamics tools, like the computation of Lyapunov exponents. Multilayer perceptron neural network (MLPNN) architecture was formulated and used as a basis for detecting variabilities such as stenosis and occlusion in the physical state of internal carotid arterial Doppler signals. The computed Lyapunov exponents of the internal carotid arterial Doppler signals were used as inputs of the MLPNN. Receiver operating characteristic (ROC) curve was used to assess the performance of the detection process. The internal carotid arterial Doppler signals were classified with the accuracy varying from 94.87% to 97.44%. The results confirmed that the proposed MLPNN trained with Levenberg-Marquardt algorithm has potential in detecting stenosis and occlusion in internal carotid arteries.