<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"> <html> <head> <meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type"> </head> <body bgcolor="#ffffff" text="#000000"> Hi Yuxuan, your question has to do mainly with STIR so you could have sent it to the STIR-users mailing list as well. The difference between positive and negative ring difference has to do with their position on the Michelogram file. You are right about the relationship of the Michelogram position of the sinograms with the same absolute ring difference bu the one being positive and the other being negative. In the Michelogram format the sinograms with positive ring-difference occupy the bottom-right corner of the Michelogram 3D-array and the sinograms with negative ring difference occupy the upper-left corner. However the oblique lines of response (LORs) for a sinogram with ring difference 1 are completely different from the oblique LORs of a sinogram with ring-difference -1. They intersect each other at the level y=0 and more specifically at the line y=0, z= (z1 + z2)/2 , where z1 and z2 are the z-coordinates of the two rings that contains the two crystals whose interactions contributed to a particular coincidence event. As a result if an annihilation takes place at a point which is not contained in that particular line then we need to know which is the ring difference of the particular LOR in order to correctly estimate the proper position. In other words by switching the sinograms of positive ring-difference with the respective ones with negative ring difference you will probably not give the correct information to your reconstruction algorithm. What was the activity distribution of your projection data when you switch the position of the sinograms in the Michelograms? If it was a point source or a line source along the z-axis of the system (y=0, x=0) you will not see much of a difference. If you use a more complex activity distribution you will notice the difference. So you need to find which is the parameter that tells us if the ring difference of a LOR is positive or negative in order to be able to add that LOR to the correct sinogram in the correct position of the Michelogram array. I, together with other gate-developers, have implemented a sinogram binning algorithm for GATE and STIR. We consider the positions of the crystal interactions (we have 2 crystal interactions that contribute to a coincidence event) within each ring with respect to the projection view as the parameter that defines if a ring difference is positive or negative. To be more specific, if an interaction takes place at a crystal which is located at the lower half of the projection plane (with respect to the projection view) then we consider that the ring where that crustal belongs to is ringA Else we consider that ring as: ringB Now the ring difference is defined by the equation: Ring_difference = ringB - ringA As a result if one crystal belongs to the lower half-plane and is located at ring number 12 and the other crystal belongs to the upper half plane and is located at ring number 15 then the ring difference of the LOR created by these two singles is : Ring_difference = 15 - 12 = 3 Both crystals cannot be in the same half-plane (low or up) at the same projection plane. They will always be one at the lower and the other at the upper half plane. We determine lower half plane by looking the projection plne from the transaxial view, so whichever crystal is below the y=0 level is considered to belong to the lower half-plane We have to note here that the criteria we use to determine the position of the crystal within each projection plane takes into account the projection view each time The 3D reconstruction algorithms require information about sinograms with positive and negative ring difference. So we need to place those sinograms within the Michelogram to the correct position where these algorithms expect to find them. Hope that helps Best regards Nicolas __________________________________________________________________ Nicolas Karakatsanis, MEng Nuclear Imaging Medical Group Biomedical Simulations and Imaging Sciences Laboratory National Technical University of Athens, Greece   <a class="moz-txt-link-abbreviated" href="mailto:yuxuan.zhang@di.mdacc.tmc.edu">yuxuan.zhang@di.mdacc.tmc.edu</a> wrote: <blockquote cite="midOF3BEA7F70.FF2AFB5D-ON86257282.0058D6A8-86257282.005CC376@mdacc.tmc.edu" type="cite"> Hi, Dear Kris and STIR users, I have a question about the format of 3d sinogram. When I was trying to bin my raw data into a 3d sinogram, I don't know what's the exactly definition of positive and negative ring difference. For example, if there's a crystal pair between crystal_i and crystal_j with the axial position of z_i and z_j. From the raw data, I could get two combinations of the same coincidence pair as (C_i, C_j) or (C_j, C_i). Obviously these two type of combinations should goes to the same oblique sinogram. But how to calculate the ring difference in this case? Ring difference could be either (z_i-z_j) or (z_j-z_i). Which one is correct? If I choose one of them, for example, z_i-z_j, and with the value of z_i=1, z_j=2, then the ring_difference = -1, if I choose the other, the ring_difference = +1. I think the difference of these two definitions are the same as switch the cells in the Michelogram between up-left corner and bottom-right corner. I have tried to bin the sinogram in both way and do the reconstruction by OSEM program, and the results looks almost the same. Does that mean the absolute direction of the ring difference is not important? or there's something else I misunderstood? Thanks. Yuxuan <pre wrap=""> <hr size="4" width="90%"> _______________________________________________ Gate-users mailing list <a class="moz-txt-link-abbreviated" href="mailto:Gate-users@lists.healthgrid.org">Gate-users@lists.healthgrid.org</a> <a class="moz-txt-link-freetext" href="http://lists.healthgrid.org/mailman/listinfo/gate-users">http://lists.healthgrid.org/mailman/listinfo/gate-users</a> </pre> </blockquote> </body> </html>