Real-time neural pulse-pattern feature extraction and recognition

Author

Tai, Jy-Der ; Gosney, W.M. ; Howard, Loin L. ; Gross, Giuenter W.

Author_Institution

Dept. of Electr. Eng., Southern Methodist Univ., Dallas, TX, USA

fYear

1992

Firstpage

1527

Abstract

Presents the algorithm and implementation of real-time feature extraction and pattern recognition for signals from a cultured living neural-cell network. Feature extraction and pattern recognition are achieved by the application of data compression techniques and an artificial neural network (ANN). The implementation consists of a 80386-based PC and a TMS320C30 digital signal processing (DSP) card that is inserted into an expansion slot in the PC. Off-line training of the ANN is done in the PC. Real-time recognition of the pulse patterns from the living neural cells is done in the DSP sub-system. The recognition results are sent from the DSP subsystem to the PC in real-time for display and recording. The feature extraction technique is introduced. The training system and the recognition system are described

Keywords

data compression; feature extraction; learning (artificial intelligence); microcomputer applications; neural nets; DSP card; cultured living neural-cell network; data compression techniques; microcomputer applications; pattern recognition; pulse patterns; real-time feature extraction; training system; Artificial neural networks; Data compression; Data mining; Digital recording; Digital signal processing; Feature extraction; Fires; Neuroscience; Real time systems; Signal processing algorithms;

fLanguage

English

Publisher

ieee

Conference_Titel

Circuits and Systems, 1992., Proceedings of the 35th Midwest Symposium on

Conference_Location

Washington, DC

Print_ISBN

0-7803-0510-8

Type

conf

DOI

10.1109/MWSCAS.1992.271070

Filename

271070