Hi Bruce,

 

I have applied one of his denoising algorithms on my high resolution datasets and also got very nice results. I used it on a single subject only, however, this could be beneficial for the upcoming hires pipeline to account for lower SNR.

 

You can find his algorithms and publications here:

 

https://sites.google.com/site/pierrickcoupe/softwares/denoising-for-medical-imaging/mri-denoising/mri-denoising-software

 

Best,

Falk

 

Von: Bruce Fischl
Gesendet: Freitag, 18. März 2016 19:43
An: Freesurfer support list
Betreff: Re: [Freesurfer] Replacing / turning off denoising filter

 

that's interesting. Is it open source? If so, can you point us at it? Maybe
someone should compare it to the version I wrote 20 years ago :)

On Fri, 18 Mar 2016, Mojmír Vinkler wrote:

> Thanks!
> Just FYI - we applied newer version of NLME from Pierrick Coupé before
> running freesurfer and it greatly improved segmentation results.
>
> On Wed, Mar 16, 2016 at 3:47 PM Bruce Fischl <fischl@nmr.mgh.harvard.edu>
> wrote:
>       Hi Mojmir
>
>       we don't apply mri_nlfilter by default so there is no need to
>       turn it
>       off. You are welcome to try it out. It implement some nonlinear
>       filters
>       including what is now called nonlocal means as described in this
>       paper:
>
> http://ieeexplore.ieee.org/xpl/abstractAuthors.jsp?reload=true&arnumber=745
>       732
>
>
>       cheers
>       Bruce
>
>       On
>       Wed, 16 Mar 2016, Mojmír Vinkler wrote:
>
>       > Hi,
>       > I was wondering if it's possible to replace or turn off your
>       denoising
>       > algorithm `mri_nlfilter`. We'd like to try different denoising
>       filters and
>       > compare how they influence segmentation performance. Right now
>       we're
>       > applying filter before running recon-all, but I fear that
>       applying your
>       > filter on already filtered image might degrade our analysis.
>       >
>       > I couldn't find any mention of mri_nlfilter besides this one
>       on mailinglisthttps://mail.nmr.mgh.harvard.edu/pipermail//freesurfer/2011-October/020
>       762.
>       > html. I hope I didn't miss something anything.
>       >
>       > Thanks!
>       > Mojmir
>       >
>       >_______________________________________________
>       Freesurfer mailing list
>       Freesurfer@nmr.mgh.harvard.edu
>       https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
>
>
>       The information in this e-mail is intended only for the person
>       to whom it is
>       addressed. If you believe this e-mail was sent to you in error
>       and the e-mail
>       contains patient information, please contact the Partners
>       Compliance HelpLine at
>       http://www.partners.org/complianceline . If the e-mail was sent
>       to you in error
>       but does not contain patient information, please contact the
>       sender and properly
>       dispose of the e-mail.
>
>
>