For recent years, patients are more likely to accept sedation during endoscopic examination due to better quality and experience. Modest anesthetic state for different kind of procedure is important, but it is also not easy to maintain in the same state all the time. Thus, to prevent awareness during the procedure, monitoring the brain state is an important issue. Traditionally, anesthesiologists assess patients' brain state from heart rate, blood pressure and muscle tone. All these vital signs are indirect hint for the real brain state. Thus, electroencephalogram(EEG) was considered to be potential tool to directly monitor our brain during anesthesia. In 1937, EEG study was started in anesthetic patients. Until 1994, Bispectral Index, the first commercialized processed EEG monitor, was published for intra-operative brain state assessment. To perform EEG study on a surgical patient is not easy due to the complex circuits and lots of interrupts during the procedure. This might be the reason why most of the EEG studies under anesthesia are collecting from healthy volunteers. It takes longer preparation, better team cooperation, and support from hospital to start this study in endoscopy room.

The investigators would like to start from a preliminary study, 5 patients in each groups, and collect the EEG during peri-endoscopic period. The sedation steps follow the routine protocol of our practice. Flumazenil and placebo are given to the groups separately on the end of the examination. From this study, further understanding of human brain during sedation especially in such short periods is expected. And furthermore, to give safer and more comfortable anesthetic experience for the patients.

Inclusion Criteria:

• ASA I or II
• plan to do colonoscopy

Exclusion Criteria:

• existed CNS disease or injuries
• known allergy to propofol, midazolam or alfentanil

It's double-blind study. We make every participant an envelope in which a randomized result of GABA antagonist or normal saline is declared to the care provider during the colonoscopy. The participant and the investigator are not aware of the medication.