IMR Press / JIN / Volume 17 / Issue 3 / DOI: 10.31083/JIN-170060
Open Access Research article
A new closed-loop strategy for detection and modulation of epileptiform spikes based on cross approximate entropy
Show Less
1 Key Lab of Industrial Computer Control Engineering of Hebei Province, Institute of Electrical Engineering, Yanshan University, Qinhuangdao 066004, China
2 School of Economics, Renmin University of China, Beijing 100872, China
3 NAAM Group, Faculty of Science, King Abdulaziz University, Jeddah, Saudi Arabia
4 Department of Mathematics, Quaid-i-Azam University, 45320, Islamabad, Pakistan
*Correspondence: (Xian Liu)
J. Integr. Neurosci. 2018, 17(3), 271–280;
Submitted: 22 August 2017 | Accepted: 28 December 2017 | Published: 15 August 2018

Closed-loop control plays an important role in the treatment of epileptiform spikes by using brain stimulation. In recent years, there have been many analytical methods for determining stimulus protocols and stimulus parameters. However, the analytical method that can start the stimulus protocol when it is needed and stop the stimulus protocol when it is not needed is rather rare. In this work, we design an analytic closed-loop control scheme which can starts control when epileptiform spikes are detected and stops control when no epileptiform spikes are detected. The neural mass model is used to simulate the generation of normal Electroencephalograph signals and epileptiform spikes. The detection of epileptiform spikes is completed via an alarm threshold which is set by using the combination of cross approximate entropy, the Pearson correlation coefficient and the fuzzy theory. If the detection result shows that there are epileptiform spikes in the neural mass model, the fuzzy proportion integration differentiation control works so that the abnormal epileptiform spikes are restored to normal EEG signals, and vice versa. The simulation confirms the effectiveness of the proposed closed-loop control scheme.

Cross approximate entropy
closed-loop control
epileptiform spikes
neural mass model
Fig. 1.
Back to top