<html><head><meta http-equiv="Content-Type" content="text/html charset=windows-1252"></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space;">Hello,<div><br></div><div>Using the Extract image filter in a streaming fashion is very efficient. Additionally the filter can be run "in-place" so no copy of the data is required. What is your file type and how is is stored?</div><div><br></div><div>Just create an ExtractImageFilter connected to connected to a reader/series reader, then only update the Extract filter not the Reader. This will cause just a region of interest to be read and stream the pipeline. As long as the ImageIO supports streaming or you have a series, the IO/memory should be quite minimal.</div><div><br></div><div>Hope that help,</div><div>Brad</div><div><br><div><div>On Mar 6, 2014, at 3:31 AM, Cyril Jaudet <<a href="mailto:drcjaudet@gmail.com">drcjaudet@gmail.com</a>> wrote:</div><br class="Apple-interchange-newline"><blockquote type="cite"><meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1"><div dir="ltr">Hi,<div> i have a similar problem to use a registration process on 4dPET images there is a multivolume explorer module architecture that seems to access in an efficient way to the multivolume data in 3Dslicer but it use vtk :<div>
<a href="https://github.com/fedorov/MultiVolumeExplorer">https://github.com/fedorov/MultiVolumeExplorer</a></div><div>Best Regards,</div><p class=""><a name="_MailAutoSig"><span style="font-size:10pt;font-family:Cambria">Jaudet Cyril, PSRPM, PhD</span></a></p><p class=""><span style="font-size:10pt;font-family:Cambria">Département
d'Ingénierie et de Physique Médicale</span><span style="font-family:Cambria"></span></p><p class=""><span style="font-size:10pt;font-family:Cambria">Institut Claudius
Regaud</span></p><p class=""><span style="font-size:10pt;font-family:Cambria">20-24, rue du
Pont Saint-Pierre - 31052 Toulouse Cedex</span><span style="font-family:Cambria"></span></p><p class=""><span style="font-size:10pt;font-family:Cambria">Tél. interne:4801
/Tél. externe : 05 67 69 63 46 </span></p><div><br class="webkit-block-placeholder"></div><div><span style="font-size:10pt;font-family:Cambria">E-mail : </span><a href="mailto:jaudet.cyril@claudiusregaud.fr"><span style="font-family:Cambria">jaudet.cyril@claudiusregaud.fr</span></a> </div></div>
</div><div class="gmail_extra"><br><br><div class="gmail_quote">2014-03-06 9:07 GMT+01:00 Polfliet Mathias <span dir="ltr"><<a href="mailto:mpolflie@etro.vub.ac.be" target="_blank">mpolflie@etro.vub.ac.be</a>></span>:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Dear ITK community,<br>
<br>
Is there an elegant, CPU and/or memory efficient way provided in the source code to store or get “subimages" ( such as a single 2D slice in a 3D volume or a single time point in a 4D series)? Some functions I use take Images as input so I cannot use Regions. For example if I were to take a 4DCT image and I want to work on a single time point ( i.e. a 3D image), I would have to use ITK::ExtractImageFilter. This method, however, becomes very memory intensive because the 4DCT is loaded into the memory and (in my case) every separate 3D slice as well. Is there an elegant or CPU/Memory friendly way to work around this?<br>
<br>
With kind regards,<br>
<br>
Mathias Polfliet<br>
PhD Student Vrije Universiteit Brussel<br>
_______________________________________________<br>
Community mailing list<br>
<a href="mailto:Community@itk.org">Community@itk.org</a><br>
<a href="http://public.kitware.com/cgi-bin/mailman/listinfo/community" target="_blank">http://public.kitware.com/cgi-bin/mailman/listinfo/community</a><br>
</blockquote></div><br></div>
_______________________________________________<br>Community mailing list<br><a href="mailto:Community@itk.org">Community@itk.org</a><br>http://public.kitware.com/cgi-bin/mailman/listinfo/community<br></blockquote></div><br></div></body></html>