<html>
  <head>
    <meta content="text/html; charset=ISO-8859-1"
      http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <div class="moz-cite-prefix">Hi Romain,<br>
      <br>
      This is straying into a FSL-specific question, but I'll ask anyway
      since we're on the topic - Am I correct in understanding that Eddy
      does a) motion correction and b) eddy current correction in its
      base form, with the optional bonus step of c) EPI distortion under
      Topup? <br>
      <br>
      I ask because in the manual it states Topup requires your DWI data
      and also a b0 volume with the opposite phase-encoded direction.<br>
      <pre class="moz-signature" cols="72">Kind regards,
Ivan Alvarez

PhD Candidate
Imaging and Biophysics Unit
UCL Institute of Child Health
30 Guilford Street, London, WC1N 1EH</pre>
      On 19/11/13 09:20, romain valabregue wrote:<br>
    </div>
    <blockquote cite="mid:528B2D45.4040608@upmc.fr" type="cite">
      <meta http-equiv="Content-Type" content="text/html;
        charset=ISO-8859-1">
      <div class="moz-cite-prefix">Hello,<br>
        <br>
        the artefact added with eddy_correct is more pronounce for high
        bvalue (3000) with 1000 or 1500 it should go better<br>
        For the bvec correction it is a tiny effect so it does not
        matter that much, (that's why it is not in there priority list),
        if you run eddy_correct you can use the extra script
        fdt_rotate_bvecs <br>
        <br>
        The eddy procedure is much better : this is mainly because in
        addition to motion it also correct the distorsion due to eddy
        current. (and also EPI distorsion with topup). The main
        limitation is that it require a specific acquisition (either the
        direction have to be over the all sphere or you need each
        direction with direct and oposite phase direction), but It worth
        to change your dwi acquisition.<br>
        <br>
        I hope it helps <br>
        <br>
        Romain<br>
        &nbsp;<br>
        Le 18/11/2013 23:39, Jeurissen Ben a &eacute;crit&nbsp;:<br>
      </div>
      <blockquote
        cite="mid:384E4D541D819449ABBBD507434317D190F0B7A6@xmail34.ad.ua.ac.be"
        type="cite">
        <style type="text/css" id="owaParaStyle"></style>
        <style type="text/css"></style>
        <div style="direction: ltr;font-family: Tahoma;color:
          #000000;font-size: 10pt;">
          <div>
            <div style="font-family:Tahoma; font-size:13px">
              <div style="font-family:Tahoma; font-size:13px">
                <div style="font-family:Tahoma; font-size:13px">
                  <div style="widows:2; text-transform:none;
                    text-indent:0px; letter-spacing:normal;
                    font-style:normal; font-variant:normal;
                    font-size:13px; line-height:normal;
                    font-family:Tahoma; white-space:normal; orphans:2;
                    word-spacing:0px"> Hi all,</div>
                  <div style="widows:2; text-transform:none;
                    text-indent:0px; letter-spacing:normal;
                    font-style:normal; font-variant:normal;
                    font-size:13px; line-height:normal;
                    font-family:Tahoma; white-space:normal; orphans:2;
                    word-spacing:0px"> <br>
                  </div>
                  <div style="widows:2; text-transform:none;
                    text-indent:0px; letter-spacing:normal;
                    font-style:normal; font-variant:normal;
                    font-size:13px; line-height:normal;
                    font-family:Tahoma; white-space:normal; orphans:2;
                    word-spacing:0px"> FSL's "eddy_correct" doesn't do
                    b-matrix rotation. They do offer a separate script
                    to reorient the bvecs file after running
                    eddy_correct, though.</div>
                  <div style="widows:2; text-transform:none;
                    text-indent:0px; letter-spacing:normal;
                    font-style:normal; font-variant:normal;
                    font-size:13px; line-height:normal;
                    font-family:Tahoma; white-space:normal; orphans:2;
                    word-spacing:0px"> <br>
                  </div>
                  <div style="widows:2; text-transform:none;
                    text-indent:0px; letter-spacing:normal;
                    font-style:normal; font-variant:normal;
                    font-size:13px; line-height:normal;
                    font-family:Tahoma; white-space:normal; orphans:2;
                    word-spacing:0px"> FSL5's "eddy" doesn't do b-matrix
                    rotation either and there is no script available to
                    reorient the bvecs file. On their mailinglist the
                    developers stated developing such a tool is
                    currently not high on their priority list.</div>
                  <div style="widows:2; text-transform:none;
                    text-indent:0px; letter-spacing:normal;
                    font-style:normal; font-variant:normal;
                    font-size:13px; line-height:normal;
                    font-family:Tahoma; white-space:normal; orphans:2;
                    word-spacing:0px"> <br>
                  </div>
                  <div style="widows:2; text-transform:none;
                    text-indent:0px; letter-spacing:normal;
                    font-style:normal; font-variant:normal;
                    font-size:13px; line-height:normal;
                    font-family:Tahoma; white-space:normal; orphans:2;
                    word-spacing:0px"> Hope this clarifies the state of
                    b-matrix rotation in FSL.</div>
                  <div style="widows:2; text-transform:none;
                    text-indent:0px; letter-spacing:normal;
                    font-style:normal; font-variant:normal;
                    font-size:13px; line-height:normal;
                    font-family:Tahoma; white-space:normal; orphans:2;
                    word-spacing:0px"> <br>
                  </div>
                  <div style="widows:2; text-transform:none;
                    text-indent:0px; letter-spacing:normal;
                    font-style:normal; font-variant:normal;
                    font-size:13px; line-height:normal;
                    font-family:Tahoma; white-space:normal; orphans:2;
                    word-spacing:0px"> Cheers,</div>
                  <div style="widows:2; text-transform:none;
                    text-indent:0px; letter-spacing:normal;
                    font-style:normal; font-variant:normal;
                    font-size:13px; line-height:normal;
                    font-family:Tahoma; white-space:normal; orphans:2;
                    word-spacing:0px"> Ben</div>
                  <div style="widows:2; text-transform:none;
                    text-indent:0px; letter-spacing:normal;
                    font-style:normal; font-variant:normal;
                    font-size:13px; line-height:normal;
                    font-family:Tahoma; white-space:normal; orphans:2;
                    word-spacing:0px"> <br>
                  </div>
                  <div style="widows:2; text-transform:none;
                    text-indent:0px; letter-spacing:normal;
                    font-style:normal; font-variant:normal;
                    font-size:13px; line-height:normal;
                    font-family:Tahoma; white-space:normal; orphans:2;
                    word-spacing:0px">
                    <div>-<br>
                      <div>
                        <div>
                          <div>
                            <div> <font size="1">Ben Jeurissen, Ph.D.</font></div>
                            <div>
                              <div><font size="1">Post-doctoral
                                  researcher</font></div>
                              <div><font size="1">Vision Lab, Dept. of
                                  Physics</font></div>
                              <div><font size="1">University of Antwerp</font></div>
                              <div><font size="1">Universiteitsplein 1,
                                  N.1.18</font></div>
                              <div><font size="1">B-2610 Wilrijk,
                                  Belgium</font></div>
                              <div><font size="1">Phone: +32 3 265 24 77</font></div>
                              <div><font size="1">Email: <a
                                    moz-do-not-send="true"
                                    class="moz-txt-link-abbreviated"
                                    href="mailto:ben.jeurissen@ua.ac.be">ben.jeurissen@ua.ac.be</a></font></div>
                              <div><font size="1">Url:&nbsp;<a
                                    moz-do-not-send="true"
                                    title="http://visielab.ua.ac.be/people/ben-jeurissen
                                    Ctrl+Click to follow link"
                                    href="http://visielab.ua.ac.be/people/ben-jeurissen"
                                    target="_blank">http://visielab.ua.ac.be/people/ben-jeurissen</a></font></div>
                            </div>
                          </div>
                        </div>
                      </div>
                    </div>
                  </div>
                </div>
              </div>
            </div>
          </div>
          <div style="font-family: Times New Roman; color: #000000;
            font-size: 16px">
            <hr tabindex="-1">
            <div id="divRpF964956" style="direction: ltr;"><font
                color="#000000" face="Tahoma" size="2"><b>From:</b> <a
                  moz-do-not-send="true"
                  class="moz-txt-link-abbreviated"
                  href="mailto:mrtrix-discussion-bounces@public.nitrc.org">mrtrix-discussion-bounces@public.nitrc.org</a>
                [<a moz-do-not-send="true"
                  class="moz-txt-link-abbreviated"
                  href="mailto:mrtrix-discussion-bounces@public.nitrc.org">mrtrix-discussion-bounces@public.nitrc.org</a>]
                on behalf of Ivan Alvarez [<a moz-do-not-send="true"
                  class="moz-txt-link-abbreviated"
                  href="mailto:ivan.alvarez.11@ucl.ac.uk">ivan.alvarez.11@ucl.ac.uk</a>]<br>
                <b>Sent:</b> 18 November 2013 13:31<br>
                <b>To:</b> Donald Tournier<br>
                <b>Cc:</b> mrtrix mailinglist<br>
                <b>Subject:</b> Re: [Mrtrix-discussion] Motion
                correction best practise<br>
              </font><br>
            </div>
            <div>
              <div class="moz-cite-prefix">Hi Donald,<br>
                <br>
                Thank you for all your comments. <br>
                <br>
                I was not aware FSL did b-matrix update, will definitely
                look into this. <br>
                <br>
                It's interesting that you mention we may be loosing more
                than we are gaining with motion correction, particularly
                with the introduction of artefacts and poor registration
                at high b-values. Is there a rule of thumb for when are
                such procedures useful/detrimental? For example, in fMRI
                movement &gt;5mm within a single run (i.e. 5-10 minutes
                continuous scanning) is considered worrisome but
                salvageable and &gt;10mm is often too much to be rescued
                by registration. Is there something like that for dMRI?<br>
                <br>
                PS The Gaussian process code is now in FSL under the
                name EDDY ( <a moz-do-not-send="true"
                  href="http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/EDDY"
                  target="_blank">
                  http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/EDDY</a>). I
                haven't personally tried it, but would love to hear from
                other users!<br>
                <pre class="moz-signature" cols="72">Kind regards,
Ivan Alvarez

PhD Candidate
Imaging and Biophysics Unit
UCL Institute of Child Health
30 Guilford Street, London, WC1N 1EH</pre>
                On 18/11/13 12:15, Donald Tournier wrote:<br>
              </div>
              <blockquote type="cite">
                <div dir="ltr">Hi Ivan,
                  <div><br>
                  </div>
                  <div>I'm probably not the best person to ask about
                    this... There's a few other people on this list
                    whose opinion on the matter is much better informed
                    than mine, I cordially invite them all to chip in.
                    ;)</div>
                  <div><br>
                  </div>
                  <div>I've no experience with ExploreDTI, but you can
                    ask Alex Leemans directly through the ExploreDTI
                    mailing list. That said, I'm fairly confident it
                    does the b-matrix update...</div>
                  <div><br>
                  </div>
                  <div>As to FSL, as far as I know it also does
                    volume-wise affine registration, and it is possible
                    to do the b-matrix update, although I'm not familiar
                    with the procedure needed to do this.&nbsp;</div>
                  <div><br>
                  </div>
                  <div>I'm not familiar with any package that does
                    slice-by-slice registration, but my gut feeling on
                    the matter is that there's probably not a great deal
                    of point to doing this as slice-wise
                    mis-registration is generally accompanied by
                    through-plane motion, which will cause signal
                    corruption due to spin history effects. For this
                    reason, I'd consider the entire volume affected to
                    be corrupt in this case: even if there is no obvious
                    signal drop, the chances are there will be some
                    corruption. That said, it might be worth doing if
                    your downstream processing pipeline has a way of
                    handling outliers, etc.&nbsp;</div>
                  <div><br>
                  </div>
                  <div>I'd also like to highlight that these approaches
                    are still far from perfect. I've already raised the
                    issue on this list, but based on my limited exposure
                    to FSL's eddy_correct (which seems to be what most
                    people use), I think it often creates more problems
                    than it solves. I hasten to add that this problem
                    may also apply to other approaches, but so far I've
                    only been exposed to data processed using
                    eddy_correct. I've come across many cases where the
                    coregistration introduces artefacts, even when the
                    original data wasn't particularly affected in the
                    first place. These artefacts typically consist of
                    the DW images being stretched along one or more
                    axes, probably because the algorithm tries to match
                    the parenchyma bit of the DW volumes to the
                    parenchyma+CSF parts of b=0 volume. This is
                    particularly pronounced with high b-value data, but
                    I've also seen it happen in run-of-the-mill b=1000
                    data too (as recently as a couple of weeks ago, in
                    fact). All this to say, if you use these tools,
                    please don't treat them as a black box, do check
                    that they're working as expected.&nbsp;</div>
                  <div><br>
                  </div>
                  <div>On the upside, Jesper Anderson recently proposed
                    a new approach based on Gaussian processes, which I
                    think has now made it into FSL5. If any other users
                    have tried using it, feel free to comment on this...</div>
                  <div><br>
                  </div>
                  <div>Cheers,</div>
                  <div>Donald.</div>
                  <div><br>
                  </div>
                </div>
                <div class="gmail_extra"><br>
                  <br>
                  <div class="gmail_quote">On 18 November 2013 03:06,
                    Ivan Alvarez <span dir="ltr"> &lt;<a
                        moz-do-not-send="true"
                        href="mailto:ivan.alvarez.11@ucl.ac.uk"
                        target="_blank">ivan.alvarez.11@ucl.ac.uk</a>&gt;</span>
                    wrote:<br>
                    <blockquote class="gmail_quote" style="margin:0 0 0
                      .8ex; border-left:1px #ccc solid;
                      padding-left:1ex">
                      <div bgcolor="#FFFFFF">Hi Donald,<br>
                        <br>
                        I wanted to bring up motion correction again,
                        particularly what is recommended for and against
                        in MRtrix. I am aware the issue has been raised
                        in the mailing list before, but it might be
                        useful to have an idea of what is generally a
                        good or bad idea.<br>
                        <br>
                        So far, I have seen people doing
                        motion/eddy-current correction in either
                        ExploreDTI or FSL. The documentation for both is
                        these is somewhat scant and I am trying to piece
                        together what do they exactly do. This is my
                        naive reading so far, please feel free to
                        correct me:<br>
                        <br>
                        ExploreDTI<br>
                        * Affine registration<br>
                        * Whole volume at a time<br>
                        * Updates B-matrix<br>
                        <br>
                        FSL<br>
                        * Affine registration<br>
                        * Slice-by-slice<br>
                        * Does<i> not</i> update B-matrix<br>
                        <br>
                        &gt;From what I understand in the discussion,
                        the slice-by-slice registration is preferable to
                        avoid smearing artefacts across the whole volume
                        while updating the B-matrix can generally
                        improve results (<a moz-do-not-send="true"
                          href="http://www.ncbi.nlm.nih.gov/pubmed/19319973"
                          target="_blank">http://www.ncbi.nlm.nih.gov/pubmed/19319973</a>).


                        Is this roughly correct? If so, are there any
                        other considerations specific to MRtrix?<span
                          class="HOEnZb"><font color="#888888"><br>
                            <pre cols="72">-- 
Kind regards,
Ivan Alvarez

PhD Candidate
Imaging and Biophysics Unit
UCL Institute of Child Health
30 Guilford Street, London, WC1N 1EH </pre>
                          </font></span></div>
                      <br>
                      _______________________________________________<br>
                      Mrtrix-discussion mailing list<br>
                      <a moz-do-not-send="true"
                        href="mailto:Mrtrix-discussion@www.nitrc.org"
                        target="_blank">Mrtrix-discussion@www.nitrc.org</a><br>
                      <a moz-do-not-send="true"
                        href="http://www.nitrc.org/mailman/listinfo/mrtrix-discussion"
                        target="_blank">http://www.nitrc.org/mailman/listinfo/mrtrix-discussion</a><br>
                      <br>
                    </blockquote>
                  </div>
                  <br>
                  <br clear="all">
                  <div><br>
                  </div>
                  -- <br>
                  <div dir="ltr"><b><font color="#990000">Dr J-Donald
                        Tournier (PhD)</font></b><br>
                    <div><font color="#990000"><br>
                      </font></div>
                    <i><font color="#990000">Senior Lecturer,&nbsp;</font></i><i><font
                        color="#990000">Biomedical Engineering</font></i>
                    <div><i><font color="#990000">Division of Imaging
                          Sciences &amp; Biomedical Engineering<br>
                          King's College London</font></i>
                      <div><i><font color="#990000"><br>
                          </font></i></div>
                      <div><i><font color="#990000"><b
                              style="font-family:Calibri,sans-serif;
                              font-size:15px"><span
                                style="font-size:10pt">A:</span></b><span
                              style="font-family:Calibri,sans-serif;
                              font-size:10pt">&nbsp;Department of Perinatal
                              Imaging &amp; Health, 1<sup>st</sup>&nbsp;Floor
                              South Wing, St Thomas' Hospital, London.
                              SE1 7EH</span><br>
                          </font></i></div>
                      <div><i><font color="#990000"><b>T:</b> +44 (0)20
                            7188 7118 ext 53613</font></i></div>
                    </div>
                    <div><i><font color="#990000"><b>W:</b>&nbsp;<a
                            moz-do-not-send="true"
href="http://www.kcl.ac.uk/medicine/research/divisions/imaging/departments/biomedengineering"
                            target="_blank">http://www.kcl.ac.uk/medicine/research/divisions/imaging/departments/biomedengineering</a></font></i><br>
                    </div>
                  </div>
                </div>
              </blockquote>
              <br>
            </div>
          </div>
        </div>
        <br>
        <fieldset class="mimeAttachmentHeader"></fieldset>
        <br>
        <pre wrap="">_______________________________________________
Mrtrix-discussion mailing list
<a moz-do-not-send="true" class="moz-txt-link-abbreviated" href="mailto:Mrtrix-discussion@www.nitrc.org">Mrtrix-discussion@www.nitrc.org</a>
<a moz-do-not-send="true" class="moz-txt-link-freetext" href="http://www.nitrc.org/mailman/listinfo/mrtrix-discussion">http://www.nitrc.org/mailman/listinfo/mrtrix-discussion</a>
</pre>
      </blockquote>
      <br>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <br>
      <pre wrap="">_______________________________________________
Mrtrix-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Mrtrix-discussion@www.nitrc.org">Mrtrix-discussion@www.nitrc.org</a>
<a class="moz-txt-link-freetext" href="http://www.nitrc.org/mailman/listinfo/mrtrix-discussion">http://www.nitrc.org/mailman/listinfo/mrtrix-discussion</a>
</pre>
    </blockquote>
    <br>
  </body>
</html>