An electroencephalogram (EEG) signal is often contaminated by eye blink (EB) artifact generated during eye blinks. Empirical Mode Decomposition (EMD) is an algorithm to decompose an EEG signal into multiple oscillating functions, where the slow oscillation functions belongs to the EB artifact. However, the algorithm is relatively slow for real time processing due to the iterative nature of EMD and the fact that interpolation of large number of data points consumes a lot of computer resources. In this research work, the cubic Hermite spline interpolation (CHSI) and the Akima spline interpolation (ASI) are investigated for their performance and their ability to retain the decomposition accuracy compared to the classical EMD algorithm. The ASI has produced the highest correlation coefficient, lowest Root Mean Square Error (RMSE), lowest percentage root means square difference (PRD), better Signal to Noise Ratio (SNR) and faster computation time in decomposing an artificial EEG signal. These results have revealed that the ASI technique in EMD is more accurate and faster than the conventional Cubic spline interpolation (CSI) technique. © 2016 IEEE.