Lossless compression of EEG data using neural network predictors


Sriraam, N. and Kannan, R. and Eswaran, C. (2002) Lossless compression of EEG data using neural network predictors. In: Proceedings of the 9th International Conference Neural Information Processing, 2002. ICONIP '02. IEEE Xplore, 2046 -2048. ISBN 981-04-7524-1

[img] Text
01199034.pdf - Published Version
Restricted to Repository staff only

Download (184kB)


Lossless compression of electroencephalograph (EEG) data is of great interest to the biomedical research community. Lossless compression through neural network is achieved by using the net as a predictor and coding the prediction error in a lossless manner. The predictive neural network uses a certain number of past samples to predict the present one and in most cases, the differences between the actual and predicted values are zero or close to zero. Entropy coding techniques such as Huffman and arithmetic coding are used in the second stage to achieve a high degree of compression. Predictive coding schemes based on single- layer and multi-layer perceptron networks and recurrent network models are investigated in this paper. Compression results are reported for EEG's recorded under various clinical conditions. These results are compared with those obtained by using linear predictors such as FIR and lattice filters.

Item Type: Book Section
Subjects: T Technology > T Technology (General)
Divisions: Faculty of Information Science and Technology (FIST)
Depositing User: Ms Suzilawati Abu Samah
Date Deposited: 31 Dec 2013 02:30
Last Modified: 31 Dec 2013 02:31
URII: http://shdl.mmu.edu.my/id/eprint/4692


Downloads per month over past year

View ItemEdit (login required)