Detection of bio-impedance gastric motility signal based on independent component analysis
Jiang Xiang-lin · Transducer and Microsystem Technologies · 2010
In order to eliminate the signal aliasing formed during the process of multi-channel surface gastric motility impedance signal acquisition,a new way which extracted gastric motility information based on independent component is studied.Four-channel detection gastric motility system is designed and gastric motility impedance information acquisition experiments of healthy volunteers is conducted.FastICA algorithm is used to implementation blind source separation for the original gastric impedance signal,the separation independent source component is analyzed,and spectrum characteristics analysis to the separation gastric motility independent source component is conducted.Application of independent component analysis can effectively eliminate the background noise added to the impedance signal,and extract gastric motility source signal.The results show that the method is very effective for multi-channel gastric motility signal blind separation.The principle of FastICA algorithm is simple,convergence speed is fast,de-noising effect is obvious,and can be effectively applied to gastric motility impedance signal detection.