Thanks for your response.
I would like to ask how can I program the level of the noise by mfile. I am
using the below code right now but I dont know how to let the fieldtrip
know my covariance matrix is the room noise.
I look forward to hearing from you.
Thanls

cfg = [];
  cfg.covariance = 'yes'; % I dont know how to enter the noise
covariance to the data

  timelockanalysis = ft_timelockanalysis(cfg, inputdata);




On Mon, Oct 24, 2016 at 5:01 PM Georgios Michalareas <
[email protected]> wrote:

> I recommend that you should start with the empty room noise rather than
> the participant noise.
>
> Giorgos
>
> On 10/25/2016 12:56 AM, mehdy dousty wrote:
>
> Thanks for your answer. as it is resting state I can have 1 trial or keep
> the trial existed in the pre-processed signals,147trails, Based on the
> paper by Sylvain Baillet with title of "The brain resting-state activity is
> shaped by synchronized cross-frequency coupling of neural oscillations" the
> covariance matrix is empirically estimated by empty-room recording,
> therefore as the room noise and participant noise is provided by HCP I
> don't know which level of noise do I need to take, or I don't need to
> consider that and just simply did what you have proposed.
> Thanks
>
> On Mon, Oct 24, 2016 at 4:40 PM Georgios Michalareas <
> [email protected]> wrote:
>
> Hi ,
>
> I think you can still do it with ft_timelockanalysis.
>
> I assume that the resting state data you mention has only one trial.
>
> then if you just do:
>
> cfg=[];
>
> cfg.covariance='yes';
>
> pseudoavg=ft_timelockanalysis(cfg, data);
>
>
> then it will just compute the average across trials but as there is only
> one trial it will basically produce the exact same data.
>
> But in the output of timelockanalysis you ll find a field .cov whith
> dimension Nchannels x Nchannels which contains the covariance matrix of
> the MEG sensors .
>
> This covariance has been computed from the entire length of the single
> resting state trial , so I think this is what you are interested in.
>
> I hope this helps.
>
> Best
>
> Giorgos
>
>
> On 10/24/2016 10:42 PM, mehdy dousty wrote:
> > Hello,
> > I am using HCP MEG resting state to compute source localization. As it
> > is resting state signal and I am using Fieldtirp I have no idea how to
> > compute the noise covariance matrix. Mostly this matrix is computed
> > with ft_timelockanalysis with the window of [-inf 0] in task-based MEG
> > I really appreciate if anybody can help me.
> > Thanks
> > Mehdy
> >
> > _______________________________________________
> > HCP-Users mailing list
> > [email protected]
> > http://lists.humanconnectome.org/mailman/listinfo/hcp-users
> >
>
> _______________________________________________
> HCP-Users mailing list
> [email protected]
> http://lists.humanconnectome.org/mailman/listinfo/hcp-users
>
>
>

_______________________________________________
HCP-Users mailing list
[email protected]
http://lists.humanconnectome.org/mailman/listinfo/hcp-users

Reply via email to