Abstracts

Rationale: EEG is one of the important tests to diagnose epilepsy. A computerized method for the interpretation of EEG, based on visual analysis followed by data entry into a computer was originally published by Oommen². The ability of the algorithm to generate accurate EEG reports and the savings in time compared to the dictated report have been demonstrated by clinical testing and was published previously(Epilepsia, 1997; 38(s3):219-20; J Okla State Medical Assoc, 2001; 94(9):400-2.). This abstract presents a web-based adaptation of the same program. It enables physicians reading EEG to generate a computerized EEG report consistent with the recommended guidelines of the American EEG Society, as they read the EEG.Methods: The reports are generated by analyzing patients demographic information as well as EEG data such as basic EEG frequencies of different areas (posterior, central and anterior) and patients normal and abnormal phenomena noted during the procedure and entering it on an onscreen Eelctroencephalographic Work Sheet . Different personnel are assigned different levels of access and operational liberties of this system for patient protection. Web programming language PHP is employed for the program, which is fast, stable, secure and easy to use in creating dynamic web pages. The data are then analyzed by the program to produce a report with appropriate clinical correlation for each of the computed normal and abnormal findings by an intelligent algorithm. The free, fast, reliable open source relational database MySQL is used for storing and searching reports. The operational aspects were tested by inputting various patient scenarios with hypothetical EEG findings.Results: The report generated by the program for each of the hypothetical scenarios was examined by the eletroencephalographer² and was found to be appropriate in both interpretations and clinical correlations.Conclusions: This project demonstrates that the algorithm originally used for EEG report generation on a local computer (see references). can be adapted for use on the worldwide web, making it within the reach of any EEG laboratory with access to the web. The user-friendly interface allows easy access and operation. Different levels of access and data input authorities for users ensure the security and integrity of this system.