[Homer-users] Kalman filter

[Bafna Tanya]

Hi,

I have a few questions about the Kalman filter for short channel regression.

1. Kaman filter is typically applicable for real time. The code in HOMER2 requires a stimulation vector to be known beforehand and it uses the stimulation vector and the hemodynamic response function to model the system. Can this code be used for real-time filtering when the stimulation vector is not known? If not, is there another way to model the system?

2. To use the code in HOMER2 for Kalman filter with short channel regression, the short distance channel signals are required as an input into the system. Are the variables uCommonMode and vCommonMode for this purpose? I could not understand where the short distance channel measurements are incorporated in the filter code.

Thank you for all your help.

Tanya
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.nmr.mgh.harvard.edu/pipermail/homer-users/attachments/20161215/5bc21e00/attachment.html