Detecting electrical stimulation of cortical mapping from recorded awake surgery sounds

Recently the surgical operation is recorded by the video camera and the sound recorder, so a number of intraoperative records are stored. There are, however, no indexes for searching surgical events, so these intraoperative records are not exploited sufficiently for analysis of surgical procedures. In our study, we target the records of awake surgery performed in the Intelligent Operating Theater of Tokyo Women's Medical University by using Intraoperative Examination Monitor for Awake Surgery(IEMAS) system. There are a number of useful intraoperative records of surgical procedures. In this paper, we propose an automatic indexing method for electrical stimulation of corical mapping which is of paramount importance in awake surgery. The electric sounds are output from the instrument while the surgeon performs direct electrical stimulation of cortical area. It is, therefore, possible to index automatically to detect these sounds. The power of this sound is, however, weak because the instrument is located aside from the sound recorder. We adopt short-term autocorrelation analysis for detection of the signals and dynamic thresholding for indexing electrical stimulation automatically. In the experiment, we applied the proposed indexing method for actual intraoperative records. The proposed approach is able to index the surgical events successfully.