Entropy Coding of Neural Firing at Hippocampus for Temporal Lobe Epilepsy Rats

Non-linear entropy coding was studied and applied to investigate particular pattern of neural firing at hippocampus of temporal lobe epilepsy (TLE) rats, which will support the hypersynchrony discharge pathogenesis for epilepsy from the view of neural firing coding. Little bins were used to split inter-spike interval of neural firing then Shannon entropy was calculated. Two kinds of typical neural firing were simulated by Chay model: bursting firing and beating firing, to verify the feasibility of entropy coding. Neural firing trains of temporal epilepsy rats and normal rats were recorded and statistics comparison was made between two calculated Shannon entropy. Results showed that entropy coding can explain neural firing coding, Shannon entropy of temporal epilepsy rats neural firing (6.36plusmn0.158 bit) is apparently lower than the entropy of normal rats (7.29plusmn0.390 bit). Entropy coding pattern of TLE rats and normal rats were completely different, entropy coded bursting firing of epilepsy neurons and hypersynchrony firing of epilepsy neurons was supported.