2015 年 43 巻 1 号 p. 22-62
Many guidelines for digital EEG (dEEG) have been published in the current dEEG era since 1990s in which they have described and defined its mechanical and software-based functions. However, little has been done for practical guidelines to record, read and interpret dEEG so far exclusively from the viewpoint of EEG technologists and EEGers. Therefore, this guideline aims to provide us with the information how to use dEEG most effectively by knowing 1) how to record and manipulate dEEG to maximize its function, and 2) how to read dEEG to maximize its function and to minimize the interpretation error and the spent time.
It is very important for EEG technologists to record dEEG with good quality. They should display and monitor dEEG in the most suitable condition, being apart from the fixed recording condition. Namely, they should change any display conditions including montage, filter setting, time window, and amplitude depending on the patients’ condition, as done previously in the analogue EEG era. Thus they could inspect and record the best dEEG while recording. As a result, (1) it provides EEGers with the most suitable display condition to read EEG immediately after recording, and (2) it gives us the suitable dEEG for further data processing such as frequency analysis, power analysis, topographical mapping, time-frequency analysis, source estimation, etc.
For EEG reading, although EEGers could anecdotally choose any display conditions (i.e., montage, filter setting, time window, and amplitude), it is usually the most effective way to read EEG as it was inspected and displayed while recording by the EEG technologists as long as dEEG was well inspectionally displayed. When needed, they could change display conditions any time. However, it should be avoided that they arbitrarily select only one or two montages to complete dEEG reading. They should read dEEG by means of at least two montages or more in combination to complement the features of each montage as so-called rational montage selection, e.g., referential montage with ipsilateral earlobe reference easily causes earlobe activation (reference activation) in the temporal abnormality and thus it is carefully interpreted or not selected in patients with temporal abnormality. Prior to final interpretation, they should make sure that they selected appropriate montages based on the clinical information, e.g., averaged reference is rather recommended to extract and localize the maximum active area properly unless the interested EEG activity is generalized, and it is not applied to read generalized activity.
While reading, EEGers could effectively use the tools of data processing such as frequency analysis, power analysis, topographical mapping, time-frequency analysis, source estimation, etc. Thus it helps them inspect dEEG visually with more accuracy, and increases both sensitivity and specificity to detect the abnormal finding.
Finally, this guideline could provide any trainees with practical learning way to read dEEG most effectively because dEEG is the useful, self-learning tool how to read EEG. Teaching files of dEEG with EEG reports enormously enhance this process.