Lempel-Ziv complexity of the EEG predicts long-term functional recovery after stroke in rats
Non-linear complexity of the EEG signal can be used to detect abnormal brain function relating to behavioral deficits. Here, we compare the effects of experimental stroke on EEG complexity using Lempel-Ziv complexity analysis (LZC) and multiscale entropy analysis (SampEn).\n\nEEG was recorded in bilateral motor cortex at baseline and during a 30-day recovery period after distal middle cerebral artery occlusion in rats. Motor function was assessed using a single pellet reaching task. Stroke caused an acute drop in both LZC and SampEn in the ipsilesional hemisphere in wakefulness, NREM and REM sleep, as well as reduced pellet reaching success. SampEn reductions persisted for at least 10 days post-stroke, whereas LZC had returned to baseline levels by day 4. EEG complexity in the contralesional hemisphere and in sham-operated animals were unaffected.\n\nIf EEG complexity reflects post-stroke brain function, post-stroke asymmetry could be used to predict behavioral recovery. In rats, acute LZC asymmetry was significantly correlated with the amount of motor function recovery by post-stroke day 31, but SampEn asymmetry was not. EEG LZC may thus be a useful tool for predicting functional recovery after stroke. MSE could be effective in identifying cortical dysfunction, but does not reflect behavioral outcomes.