In general, when EEG is performed in hospitals, electrodes are placed directly on the scalp to record EEG signals, called scalp EEG. In reality, because the scalp and skull are poor conductors of electricity, the EEG signal is bound to be severely attenuated during the passage, and coupled with the influence of electromyographic activity during seizures, the scalp EEG sometimes does not provide sufficient evidence of the origin of epileptiform discharges, and even with advanced computer processing tools such as BESA analysis, the epileptogenic foci of many epileptic patients still cannot be accurately located. In such cases, if the electrodes are placed directly on the brain surface or into deep brain structures by craniotomy, the subtle processes of EEG can be clearly displayed without interference from the scalp, skull, EMG and daily activities, with high sensitivity. Combined with the clinical seizure manifestations recorded by the video technology, the moment of seizure onset can be determined, the primary and conduction-derived abnormal discharges can be further identified, and the location and extent of the epileptogenic focus can be accurately delineated. After this task is accomplished, the brain tissue covered by the electrodes is stimulated by skillfully applying a current that is non-invasive to the human brain to these electrodes through sophisticated equipment in a fully awake condition, and the location of the brain function of the epileptogenic focus itself and important functional areas of the cortex is inferred based on the different responses of the subject patient’s consciousness, speech, limb sensation, movement, vision and other functions, so that a safe surgical excision plan can be designed. For the epileptogenic foci around the functional areas, they can be removed with confidence; for the epileptogenic foci located in important functional areas that cannot be preserved, multiple submural transverse fiber dissection can be performed as appropriate, so that the adjacent nerve cell populations cannot unite to discharge, which can reduce the possibility of clinical seizures and alleviate the condition. Compared with scalp EEG, the disadvantages of intracranial electrode EEG are that the patient must endure the pain and risk of craniotomy for placement of electrodes (together with resection of the epileptogenic focus, the patient needs to undergo at least two successive craniotomies) and the limited coverage of intracranial electrodes, which may miss epileptiform discharges beyond the coverage area. Therefore, intracranial electrode EEG cannot replace scalp EEG. A comprehensive analysis of the patient’s clinical seizure form, imaging examination, and especially scalp EEG data can make the intracranial electrode placement more purposeful and targeted, and achieve localization with less electrode placement. Intracranial electrode placement can lead to complications such as intracranial hemorrhage, cerebral edema, cerebrospinal fluid leakage and infection, with an incidence of 0.9% to 4.3% reported in the literature. We have been working to improve the surgical approach, shorten the monitoring time reasonably, and minimize the occurrence of complications. To date, there has been no precedent of death due to electrode embedding itself in our epilepsy center.