OBJECTIVE: We propose a method that uses minimal redundancy and maximal relevance (mRMR) based on mutual information as criteria to automatically select references for the frequency-dependent subtraction (SUBTR) method to attenuate maternal (mMCG) and fetal (fMCG) magnetocardiograms of fetal magnetoencephalography recordings. METHODS: mRMR is calculated between all channels and mMCG/fMCG target channels and the most promising sensors are used as references to perform SUBTR. We measured the performance of SUBTR at removing interferences in two steps for different number of references in 38 real datasets. The evaluation was based on the MCG amplitude reduction. We compared the performance of the mRMR approach with random selection of references. RESULTS: Significant differences in interference removal were found when a distinct number of references were chosen by mRMR compared to random selection. CONCLUSION: mRMR provides an effective tool to automatically select a set of featured references. SIGNIFICANCE: Although we show the utility of the mRMR method to biomagnetic signals, the approach can easily be adapted to sensor array data from other applications.
OBJECTIVE: We propose a method that uses minimal redundancy and maximal relevance (mRMR) based on mutual information as criteria to automatically select references for the frequency-dependent subtraction (SUBTR) method to attenuate maternal (mMCG) and fetal (fMCG) magnetocardiograms of fetal magnetoencephalography recordings. METHODS: mRMR is calculated between all channels and mMCG/fMCG target channels and the most promising sensors are used as references to perform SUBTR. We measured the performance of SUBTR at removing interferences in two steps for different number of references in 38 real datasets. The evaluation was based on the MCG amplitude reduction. We compared the performance of the mRMR approach with random selection of references. RESULTS: Significant differences in interference removal were found when a distinct number of references were chosen by mRMR compared to random selection. CONCLUSION: mRMR provides an effective tool to automatically select a set of featured references. SIGNIFICANCE: Although we show the utility of the mRMR method to biomagnetic signals, the approach can easily be adapted to sensor array data from other applications.
Authors: J Vrba; J McCubbin; R B Govindan; S Vairavan; P Murphy; H Preissl; C L Lowery; H Eswaran Journal: Neuroimage Date: 2011-09-10 Impact factor: 6.556
Authors: Jiri Vrba; Stephen E Robinson; Jack Mccubbin; Curtis L Lowery; Hari Eswaran; James D Wilson; Pamela Murphy; Hubert Preissl Journal: IEEE Trans Biomed Eng Date: 2004-07 Impact factor: 4.538
Authors: J Vrba; S E Robinson; J McCubbin; P Murphy; H Eswaran; J D Wilson; H Preissl; C L Lowery Journal: Neuroimage Date: 2004-03 Impact factor: 6.556
Authors: J McCubbin; S E Robinson; R Cropp; A Moiseev; J Vrba; P Murphy; H Preissl; H Eswaran Journal: IEEE Trans Biomed Eng Date: 2006-08 Impact factor: 4.538
Authors: Umit D Ulusar; R B Govindan; James D Wilson; Curtis L Lowery; Hubert Preissl; Hari Eswaran Journal: Conf Proc IEEE Eng Med Biol Soc Date: 2009