<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN"> <HTML> <HEAD> <META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1"> <META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7651.59"> <TITLE>RE: [Gate-users] [ no subject ]</TITLE> </HEAD> <BODY>  Dear MJ and others, I think what Irene et al reported is possible since benchmarkPET (and many other simulations) used blurring module (e.g., 26% resolution in benchmarkPET). This will simply smear any small difference in no blurring application like Irene did. I am also puzzled by Irene's discovery: Single's energy goes above norminal input energy without blurring in GATE version 3.1.2. First, it is not likely a coincidence sorter problem (but that less coincidence events problem probably is.). I think the problem could be due to: 1) GATE single (pulse) algorithm change? If an algorithm of single like all hits within a certain time window are collected and accounted for one pulse is implemented, it is possible to simulate pileup effect, and then we can see unreal larger energy for one single. 2) any change in Geant4? which corresponding Geant4 version? Is it possible that Genat4 loses correct track and time information for one event and wrongfully assigns those steps after a certain time to next event? To exclude those possibilities, a very low activity simulation will be helpful. Irene et al., can you do this: repeat your simualtion in v 3.1.2 at a very low activity?   Looking forward to seeing your new results. Yu Chen, Ph.D. University of Massachusetts Medical School Division of Nuclear Medicine 55 Lake Avenue North Worcester, MA 01655-0243 Phone:  (508) 856-6123 Fax:    (508) 856-4572 -----Original Message----- From: gate-users-bounces@lists.healthgrid.org on behalf of Park MinJae Sent: Mon 12/17/2007 8:20 PM To: 'Jasmine Schirmer'; gate-users@lists.healthgrid.org Subject: RE: [Gate-users] [ no subject ] Dear Irene, Let me check what you are saying... There is no problem with benchmarkPET... [In both the old and new versions of GATE, the benchmark PET was successfully completed and no difference in the energy spectra was observed.] In your simulation code, uphold value is 950 keV... [However, the upper energy threshold of the benchmark is set to 650keV while our simulations are performed at higher upper energy thresholds.] [/gate/digitizer/Singles/upholder/setUphold 950. keV] In fact, I can't get that the benchmark is same but your code is different... Can I take a look on your code and simulate your code with my gate? MJ -----Original Message----- From: gate-users-bounces@lists.healthgrid.org [<A HREF="mailto:gate-users-bounces@lists.healthgrid.org">mailto:gate-users-bounces@lists.healthgrid.org</A>] On Behalf Of Jasmine Schirmer Sent: Monday, December 17, 2007 11:24 PM To: gate-users@lists.healthgrid.org Subject: [Gate-users] [ no subject ] Dear GATE users and developers, A few weeks ago, we sent an email to the users list concerning a problem we are having with the newest version of GATE. The email was dated the 29th of November and is titled 'differences in energy spectra from different GATE releases'. At the moment, we can no longer use reliably the newest version of GATE until this issue is resolved. To recall your attention, we showed a difference in the energy spectra produced by two identical simulations using two versions of GATE (version 3.0.0 and the newest 3.1.2). The newest version gives a different and unexpected result for the energy spectra as you can see from the figures attached in the previous email. (The energy spectra does not fall to zero!) We kindly ask the developers and users to address this issue. In addition, if anyone would like to perform a similiar test on their version of GATE, we would be happy to provide the macro we used. Thanks in advance, Irene Torres and Jasmine Schirmer _______________________________________________ Gate-users mailing list Gate-users@lists.healthgrid.org <A HREF="http://lists.healthgrid.org/mailman/listinfo/gate-users">http://lists.healthgrid.org/mailman/listinfo/gate-users</A> </BODY> </HTML>