Acta Physica Sinica, Journal Year: 2024, Volume and Issue: 74(1), P. 0 - 0
Published: Nov. 25, 2024
Brain diseases often coincide with critical transitions in neural system and abnormal neuronal firing. Studying early warning signals (EWS) of can offer a promising avenue for predicting firing behaviors, which potentially aid the diagnosis prevention brain diseases. Conventional EWS, such as autocorrelation variance, have been widely used to detect various dynamical systems. However, these methods are limited distinguishing different types bifurcations. In contrast, EWS power spectrum shown significant advantage not only bifurcation points but also bifurcations involved. Previous studies demonstrated its predictive climate ecological models. Based on this, this study applies systems order predict behaviors distinguish classes excitability. Specifically, we compute before occurrence saddle-node invariant circle subcritical Hopf Morris-Lecar neuron model. Additionally, extend analysis Hindmarsh-Rose model, calculating both supercritical bifurcation. The contains four codimension-1 corresponding For comparison, calculate two conventional EWS: lag-1 variance. numerical simulations, stochastic differential equations simulated by Euler-Maruyama method. Then, responses detrended Lowess filter. Finally, calculated using rolling window method ensure detection points. Our results show that effectively points, mean it activities. Comparing accurately firing, excitability neurons. That is, according characteristics frequencies, during This work presents novel approach system, potential applications diagnosing treating
Language: Английский