Research On Anesthesia Depth Monitoring Algorithm And Its Implemetation

Anesthesia is a vital part in modern medicine, and nearly an integral step of any operation. However, excessive or insufficient anesthetic will bring patients with severe physical, mental and psychological trauma, among which waking during surgery is most harmful, so the control of the reasonable amount of anesthetic has great significance.Currently, anesthetists use physiological characteristics of patients during surgery in clinical, with the depth of experience in judging the state of anesthesia. But this method is affected greatly by environmental factors and experience of anesthetists, not unsuitable for a-long-time monitoring. So the development of one proprietary anesthesia depth monitoring instrument is necessary.Based on the extensive research of commercialized anesthesia depth monitors currently on the market, we conducted research of anesthesia depth techniques based on multi-domain analysis of EEG. In this paper, MIT-BIH database was used to simulate the real depth of anesthesia state, and the first state was to extract the anesthesia-related characteristic parameters from EEG. Firstly, we proposed EEG and EOG separation technology in a single-channel noisy EEG without a reference channel, in which wavelet transform, empirical mode decomposition approach and independent component analysis were adopted. Secondly we explored the relationship between band energy ratio of EEG signal and the depth of anesthesia. Then for the first time, we have quantified the relationship between the number of eye movements and the simulated anesthesia state. And improved fast approximate entropy algorithm was to portray effectively the relationship between the simulated anesthesia state. Finally, based on the least square- support vector machine technology, we designed anesthesia depth state classifier to quantify the depth of anesthesia.Finally, the implementation on ARM platform was given. The platform set of EEG data collection, estimation of anesthesia depth in one run, and through the Bluetooth wireless data transmission, anesthesia depth values are shown in real time on slave OLED screen. Then the verification experiment showed that the instrument can effectively portray simulated anesthesia state, and this truly has laid a solid foundation for the subsequent real estimation of anesthesia depth in clinical trials.