1) Objective of the research:Investigation for weak signal detection and information transmission in self-adaptive neural systems.2) Novelty of the work:Being different from most literatures dealing with "stochastic resonance" [1] in biological systems to treat the neurons or networks as passive nonlinear system,whose the intrinsic properties were invariant,this study took into account the high flexibility in biological systems where modifiability is one of the fundamental characteristics.3) Methods and techniques:In the present study,we employed simple and biophysically realistic models,where neurons sensitivity could be dynamically adjusted,i.e.,the relationship between its firing activity and input stimulation is related to the changes of the cells membrane properties such as open and close characteristics of ion channels.By computer simulation,we tested this self-adaptive neurons ability for weak signal detection under various noise backgrounds.