An effective novel patient specific Gaussian template based scheme for somatosensory evoked potential detection

Somatosensory evoked potentials (SSEPs) have been widely used for intra-operative neurophysiological monitoring (IONM) in surgeries for scoliosis and spinal cord related surgical procedures. Current technological trends show that at least 200-300 trials are required to generate a readable SSEP signal. A novel signal processing approach is outlined in the article, which aims at reducing the number of trials required for SSEP generation using only 30 trials. The analysis was performed on data recorded in seven patients undergoing surgery, where the posterior tibial nerve was stimulated and the SSEP response was recorded from scalp EEG using two bipolar electrodes, C₃-C₄ and C_Z-F_Z. The time latencies of the P37 and the N45 peaks are detected along with the peak-to-peak amplitudes. The time latencies are detected with a mean accuracy greater than 95%. Also, the P37 and N45 peak latencies and the peak-to-peak amplitude were found to be consistent throughout the surgical procedure within the 10% and 50% clinical limits set for the time latencies and the peak-to-peak amplitude, respectively.