<html>
  <head>
    <meta content="text/html; charset=utf-8" http-equiv="Content-Type">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    Thanks Simon,<br>
    <br>
    I have implemented both of the techniques you have mentioned. Nui
    scatter correction does do a reasonable job, but I have found it
    leaves significant artefacts when there is a substantial change in
    organ shape or location between scans, and I hope that MC will allow
    me to investigate the true effect of this technique when I have the
    scatter ground truth.<br>
    <br>
    WLS also has been outperforming LS in phantom experiments, but only
    when photon flux is being starved, and has been difficult to get a
    reasonable estimate of I0 from Varian data (also have yet to set up
    the scanner to even produce such a low current).<br>
    <br>
    I am doing my developments in MATLAB rather than RTK at the moment
    as I find it quicker to test things out, and utilise Toolboxes, but
    it is my hope to eventually move over once I have found more
    stability. I would love to contribute to what is now a very
    comprehensive and powerful software package.<br>
    <br>
    Cheers,<br>
    <br>
    Jonathan<br>
    <br>
    <div class="moz-cite-prefix">On 01/12/15 08:58, Simon Rit wrote:<br>
    </div>
    <blockquote
cite="mid:CAF0oig2LgKGM9jq7x+iKK95wp=9w-6tGuP1EtJUHEoUGnqnU3A@mail.gmail.com"
      type="cite">
      <div dir="ltr">
        <div>
          <div>Hi,<br>
          </div>
          For scatter and assuming you have the planning CT, you could
          consider the <a moz-do-not-send="true"
            href="http://www.openrtk.org/RTK/news/201507_press.php">solution
            of Yang et al</a> (that he implemented from Niu et al) that
          is much simpler than Monte Carlo. I know two teams that have
          already implemented it using RTK and we will try to do our own
          in the future.<br>
        </div>
        <div>For photon statistics, Cyril has recently been working on a
          weighted least square solution, see, e.g., <a
            moz-do-not-send="true"
href="https://github.com/SimonRit/RTK/commit/b133c0bb96bff6805dd28c7820a89310629ff031">this
            commit</a>. But I don't think it's mature yet.<br>
        </div>
        <div>Note that we discussed pre-processing in July in Lyon, the
          minutes are <a moz-do-not-send="true"
href="http://wiki.openrtk.org/index.php/RTK/Meetings/TrainingNov15#Pre-processing">here</a>.<br>
        </div>
        <div>Good luck, it's not an easy task,<br>
        </div>
        Simon</div>
      <div class="gmail_extra"><br>
        <div class="gmail_quote">On Tue, Dec 1, 2015 at 9:50 AM,
          Jonathan Mason <span dir="ltr"><<a moz-do-not-send="true"
              href="mailto:s1015431@staffmail.ed.ac.uk" target="_blank">s1015431@staffmail.ed.ac.uk</a>></span>
          wrote:<br>
          <blockquote class="gmail_quote" style="margin:0 0 0
            .8ex;border-left:1px #ccc solid;padding-left:1ex">
            <div style="word-wrap:break-word">Thank you Simon and Chao,
              <div><br>
              </div>
              <div>I apologise for the confusion—I can see that my
                message was not very clear. I am essentially trying to
                compensate for low photon flux and scatter, but am
                finding it difficult with pre corrected data. I will
                certainly take time to look through your suggested
                publications, so thanks for that.</div>
              <div><br>
              </div>
              <div>Without having access to low level physical data, I
                am considering simulating a scanner with MC, where I
                will have the ground truth for scatter field and photon
                statistics, and try to produce reconstructions on this.</div>
              <div><br>
              </div>
              <div>Many thanks,</div>
              <div><br>
              </div>
              <div>Jonathan</div>
              <div>
                <div class="h5">
                  <div><br>
                    <div>
                      <blockquote type="cite">
                        <div>On 1 Dec 2015, at 07:27, Simon Rit <<a
                            moz-do-not-send="true"
                            href="mailto:simon.rit@creatis.insa-lyon.fr"
                            target="_blank">simon.rit@creatis.insa-lyon.fr</a>>
                          wrote:</div>
                        <br>
                        <div>
                          <div dir="ltr">
                            <div>
                              <div>
                                <div>
                                  <div>Hi Jonathan,<br>
                                  </div>
                                  I'm not sure I fully get it either.
                                  What I can say is that I was a postdoc
                                  at the NKI until the end of 2009 where
                                  I was working on the Elekta
                                  reconstruction software. The way <a
                                    moz-do-not-send="true"
                                    href="http://www.openrtk.org/Doxygen/classrtk_1_1ProjectionsReader.html"
                                    target="_blank">rtk::ProjectionsReader</a>
                                  works for Elekta projections is close
                                  to what they were doing at the time,
                                  i.e., Boellaard scatter correction,
                                  cropping and simple log to go to a
                                  line integral. I know they have been
                                  improving it since (with lag and
                                  scatter corrections) but I'm not sure
                                  what's commercialized what's not. In
                                  any case, they publish / present what
                                  they do (see publications of M. van
                                  Herk, J-J. Sonke and L. Ploeger).<br>
                                </div>
                                I'm not expert of Varian data but for
                                sure, when you get pre-corrected
                                projections, it's hard to know what part
                                you're correcting (uncorrected problems
                                or miscorrections). Instead of reverse
                                engineering, which can be tough because
                                Varian does a lot of stuff on the raw
                                data I believe (look at the publications
                                of J. Star-Lack), I would start from the
                                raw data if you can access them.<br>
                              </div>
                              Good luck,<br>
                            </div>
                            Simon<br>
                          </div>
                          <div class="gmail_extra"><br>
                            <div class="gmail_quote">On Mon, Nov 30,
                              2015 at 7:50 PM, Chao Wu <span dir="ltr"><<a
                                  moz-do-not-send="true"
                                  href="mailto:wuchao04@gmail.com"
                                  target="_blank">wuchao04@gmail.com</a>></span>
                              wrote:<br>
                              <blockquote class="gmail_quote"
                                style="margin:0 0 0 .8ex;border-left:1px
                                #ccc solid;padding-left:1ex">
                                <div dir="ltr">Hi Jonathan,
                                  <div><br>
                                  </div>
                                  <div>I do not quite understand your
                                    problem.</div>
                                  <div>I think most information can be
                                    retrieved by the relationship
                                    between flux (I) and attenuation
                                    (u):</div>
                                  <div>I = I0 * exp (- integral(u * dL)
                                    )</div>
                                  <div>Of course there are other effects
                                    like beam hardening and scattering
                                    involved but this model is the
                                    basic.</div>
                                  <div><br>
                                  </div>
                                  <div>Regards,</div>
                                  <div>Chao</div>
                                </div>
                                <div>
                                  <div>
                                    <div class="gmail_extra"><br>
                                      <div class="gmail_quote">2015-11-30
                                        17:29 GMT+01:00 Jonathan Mason <span
                                          dir="ltr"><<a
                                            moz-do-not-send="true"
                                            href="mailto:j.mason@ed.ac.uk"
                                            target="_blank">j.mason@ed.ac.uk</a>></span>:<br>
                                        <blockquote class="gmail_quote"
                                          style="margin:0 0 0
                                          .8ex;border-left:1px #ccc
                                          solid;padding-left:1ex">Good
                                          afternoon RTK users,<br>
                                          <br>
                                          I have access to "raw data"
                                          from a Varian OBI scanner, and
                                          have<br>
                                          experimented with RTK for
                                          performing its reconstruction.
                                          However, it is<br>
                                          not really "raw" in the sense
                                          that the coefficients are
                                          proportional to<br>
                                          the photon flux recorded at a
                                          given sensor, but is instead
                                          proportional<br>
                                          to the Hounsfield attenuation
                                          along that path. With this
                                          data, it means<br>
                                          that one can reconstruct using
                                          FDK and other iterative
                                          techniques, which<br>
                                          assume a linear model, but the
                                          connection to the underlying
                                          physics is<br>
                                          abstract.<br>
                                          <br>
                                          The problem I then have when
                                          trying to develop
                                          reconstruction<br>
                                          techniques, is that I do not
                                          know whether I am really
                                          correcting for<br>
                                          physical distortions or just
                                          correcting for Varian's
                                          preprocessing,<br>
                                          which has been finely tuned
                                          for its own FDK method.<br>
                                          <br>
                                          My question is whether anybody
                                          has dug into the steps that
                                          manufacturers<br>
                                          such as Varian or Elekta
                                          perform to arrive at these
                                          projection images?<br>
                                          And if they think that if
                                          reverse engineered, could
                                          provide a richer set<br>
                                          of information to facilitate
                                          advanced strategies.<br>
                                          <br>
                                          Best wishes,<br>
                                          <br>
                                          Jonathan Mason<br>
                                          <span><font color="#888888"><br>
                                              --<br>
                                              The University of
                                              Edinburgh is a charitable
                                              body, registered in<br>
                                              Scotland, with
                                              registration number
                                              SC005336.<br>
                                              <br>
_______________________________________________<br>
                                              Rtk-users mailing list<br>
                                              <a moz-do-not-send="true"
href="mailto:Rtk-users@public.kitware.com" target="_blank">Rtk-users@public.kitware.com</a><br>
                                              <a moz-do-not-send="true"
href="http://public.kitware.com/mailman/listinfo/rtk-users"
                                                rel="noreferrer"
                                                target="_blank">http://public.kitware.com/mailman/listinfo/rtk-users</a><br>
                                            </font></span></blockquote>
                                      </div>
                                      <br>
                                    </div>
                                  </div>
                                </div>
                                <br>
_______________________________________________<br>
                                Rtk-users mailing list<br>
                                <a moz-do-not-send="true"
                                  href="mailto:Rtk-users@public.kitware.com"
                                  target="_blank">Rtk-users@public.kitware.com</a><br>
                                <a moz-do-not-send="true"
                                  href="http://public.kitware.com/mailman/listinfo/rtk-users"
                                  rel="noreferrer" target="_blank">http://public.kitware.com/mailman/listinfo/rtk-users</a><br>
                                <br>
                              </blockquote>
                            </div>
                            <br>
                          </div>
                        </div>
                      </blockquote>
                    </div>
                    <br>
                  </div>
                </div>
              </div>
            </div>
            <br>
            The University of Edinburgh is a charitable body, registered
            in<br>
            Scotland, with registration number SC005336.<br>
            <br>
          </blockquote>
        </div>
        <br>
      </div>
    </blockquote>
    <br>
  </body>
</html>