Designing a system that can accurately notify the doctor of the patient's level of consciousness in the operating room and ICU departments, has always been one of the challenges of the treatment team in hospitals. The usual method of measuring the depth of anesthesia in operating rooms is to use hemodynamic criteria, which is not satisfactory. Using bispectral index (BIS) is an advanced and reliable method to measure the depth of anesthesia, and its number has an inverse relationship with the depth of anesthesia. In this article EEG signals have been used for estimating of BIS using Time Delay Neural Network (TDNN). Using the designed software, all parameters of the extraction time domain, such as brain signal, Burst Suppression Ratio (BSR), frequency range of the bi spectrum, 95% spectral edge frequency (95%SEF), median frequency (MF), and relative delta power (RDP) have been extracted. These parameters are feed to the as input of a neural network for estimating the BIS. EEG signals during anesthesia were saved by BIS XP monitor (Aspect medical system Inc.).