Synchronized Averaging for Noise Reduction.docVIP

Lab 1 Exercises Report Content: Table 1. Signal-to-Noise Ratio (SNR) and Euclidean distance computed by M synchronized averaging of the Event-Related Potential (ERP) epochs. M E1 to E4 E1 to E8 E1 to E12 E1 to E24 E17 to E24 E13 to E24 SNR(dB) 2.034003 -1.373021 -0.947101 -4.271319 -12.052868 -9.292403 Euclidean distance 1.7087e+03 2.1960e+03 2.2371e+03 2.3513e+03 2.1372e+03 2.2305e+03 The original and averaged signals (all ERPs are with a 10000 amplifier gain, herein) are shown in Figure 1 to Figure 6: Figure 1. ERP Signals of EE11 to E44, and the averaged output at the bottom. Figure 2. ERP Signals of EE11 to E88, and the averaged output at the bottom. Figure 3. ERP Signals of EE11 to E1212, and the averaged output at the bottom. (a) (b) (c) Figure 4. ERP Signals of (a) EE11 to E88, (b) E99 to E1616, (c) E1717 to E2424, and the averaged output at the bottom. Figure 5. ERP Signals of E1717 to E2424, and the averaged output at the bottom. Figure 6. ERP Signals of E1313 to E2424, and the averaged output at the bottom. Summary As we all know, the larger the number of epochs or realizations that are averaged, the better will be the SNR of the result. But observing the table above, we find that the data doesn’t adapt to the theory.