<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;">Kent,<div><br></div><div>A long time ago, when I was working on IO streaming, I had modified the ImageSeries reader to populate the metadata dictionary on the GenerateOutputInformation phase of the pipeline. This ended up getting called 2 or 3 times, and it caused a significant performance penalty with the GDCM meta-data that the Slicer folks loudly complained and it was fixed [ likely here 1].</div><div><br></div><div>I have doubts about your assumption that there are not significant computation expenses with metadata IO. However I am not quite certain if my experience is relavent to this issue. </div><div><br></div><div>just an FYI,</div><div>Brad</div><div><br></div><div><br></div><div>[1] <a href="https://github.com/InsightSoftwareConsortium/ITK/commit/e92d78ee3b26069d3aeb6aa9f6460a46f20ec2de">https://github.com/InsightSoftwareConsortium/ITK/commit/e92d78ee3b26069d3aeb6aa9f6460a46f20ec2de</a></div><div><br><div><div>On Sep 16, 2014, at 1:15 PM, Williams, Norman K <<a href="mailto:norman-k-williams@uiowa.edu">norman-k-williams@uiowa.edu</a>> wrote:</div><br class="Apple-interchange-newline"><blockquote type="cite"><span style="font-family: Calibri, sans-serif; font-size: 14px; font-style: normal; font-variant: normal; font-weight: normal; letter-spacing: normal; line-height: normal; orphans: auto; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: auto; word-spacing: 0px; -webkit-text-stroke-width: 0px; float: none; display: inline !important;"><span class="Apple-converted-space"> </span>the time spent in writing and reading the MetaDataDictionary is not going to have a significant computational cost, since it only happens when you read or write a DICOM file. It’s not something done in the inner loop of a big computation.</span></blockquote></div><br></div></body></html>