Re: [Tirius-users] New user of Tirius

[Liliana C. <llc...@gm...>]

Hello.

Thank you so much for you help. It was really helpful. I am sorry that
I used your personal email, but I was getting somehow desperate. From
now on, I will only use tirius-users.

This week I will try to get on with changing your code, now with
better insight. I will keep you updated on my evolution and my
problems.

Once again, just keep the goodwork.

Thank you again,
Liliana

On Sun, Jun 28, 2009 at 7:49 PM, Jean-Daniel Leroux<le...@ho...> wrote:
>
> Hi Liliana,
>
> Sorry for the long delay... I just realized that messages sent with the
> tirius-users mailing list address were going directly to my junk mail
> folder... not a good thing when you are the administrator! Now it is fixed.
>
> First, thanks for reporting the bugs about the loading of the sinogram and
> for the Gate single data... I will try to fix them for the next release of
> Tirius that maybe could be in octobre or novembre.
>
> About the SetCameraDefaultParams() function, the ring diameter is computed
> from the front face of the crystals. In fact, the 150 cm you not about the
> Sherbrooke scanner is probably measured from the front face of the detector
> casing and not from the front face of the crystal within the casing which is
> the reference I use. About the number of ring you should use the total
> number of ring which is in your case 13*4=52 for your scanner.
>
>  For the FindCoincDetectorAdd and FindSingleDetectorAdd function, it could
> be a little bit tricky depending on how you defined your scanner in your
> .mac file for GATE. The key is to follow in which order you use your
> repeater for repeating the crystals to build your scanner geometry.
>
> Lets have an example by assuming that you defined your block of crystals
> first. Then you repeat it around the ring lets say 8 times. And finally you
> repeat the 4 block axially. Then,
>
> volumeID5 could go from 0 to 168 (position in block 13x13)
> volumeID4 could go from 0 to 7 (position in the ring)
> volumeID3 could go from 0 to 3 (axial position of the block)
>
> Then, you should choose a given convention for ordering your crystal in the
> scanner. I suggest you to order them counter-clockwise from the bottom ring
> to the top ring. Your detector address could then be found using:
>
> axialRingPos = volumeID3 * 13 + (volumeID5 / 13);
> crystalPos = volumeID4 * 13 + (volumeID5 % 13);
> detectorAddress = axialRingPos * numberCrystalInRing + crystalPos;
>
> This is an example only so do not cut and paste it directly in your own
> code. You will have to check if the detector address correspond well to what
> you expect looking at the (x,y,z) position of the hit in the GATE list mode
> file.
>
> Good luck,
>
> Jean-Daniel
>
>
>
>
>
>
> ---------- Forwarded message ----------
> From: Liliana Caldeira <llc...@gm...>
> Date: Thu, Jun 18, 2009 at 10:40 PM
> Subject: New user of Tirius
> To: tir...@li...
>
>
> Hello everyone.
>
> I'm a new user of Tirius. Just start last week. I have used STIR and
> Qt before, so the installation went fine and the basic user interface
> was also good. First, let me congratulate you in the great work.
>
> I have noticed one bug that you might want to take a look. When I am
> trying to convert GATE data if you do it two times in a row, it gives
> you a Segmentation Fault in the second one. I try to understand where
> the problem is and aparently it is in Sinogram.cpp in member function
> CreateSinogram. In the second if, the option else when trying to
> delete mSinogram-.>noCrystal1. I don't know if I'm doing something
> wrong or it is really a bug. Anyway, what i do for now is just start
> tirius over again, because I don't know how to fix it.
>
> else  // Then we only recompute the sinogram address
>        {
>                if (mSinogram->noCrystal1 != NULL) delete []
> mSinogram->noCrystal1;
>                mSinogram->noCrystal1 = new IFormat [mNbTORinScanner];
>                if (mSinogram->noCrystal1 == NULL)  {
> ShowError("CreateSinogram:
> Unable to allocate memory for sinogram"); return ERROR; }
>
>                if (mSinogram->noCrystal2 != NULL) delete []
> mSinogram->noCrystal2;
>                mSinogram->noCrystal2 = new IFormat [mNbTORinScanner];
>                if (mSinogram->noCrystal2 == NULL)  {
> ShowError("CreateSinogram:
> Unable to allocate memory for sinogram"); return ERROR; }
>        }
>
>
>
> I also had to change some line in GateSingleData.cpp because when
> there is less than two run in the gate simulation (I have one), the
> mAcqTimeSlice is zero and it gets into a endless loop. I can send you
> a example file if you would like. As far as I can see, I have 3
> options and I would like you to tell me which is the most correct one:
> 1- Initialize mAcqTimeSlice with value of 0.01
> 2- Change the if clause in cGateSingleData::ReadCoincInfo(char
> *FileName), we have:
> if (bTimeSliceFound == false && D1[i].runID == 1) {
>                                ...
>                        }
>
> and change to if (bTimeSliceFound == false && D1[i].runID == 0) {
>                                ...
>                        }
> and put a if inside if (mAcqTimeSlice <0.01) mAcqTimeSlice=0.01
>
> 3- Just put another if before this one
> if (bTimeSliceFound == false && D1[i].runID == 0) {
>                                ...
>                        }
>
> I think the first one is the most elegant, but I'm afraid it will
> affect other things. Can you tell me your opinion?
>
> Now, I'm working on adding my GATE scanner to the converter. I add all
> the files necessary and I already see my scanner in the options. The
> only thing i need to do now is change the parameters in
> SetCameraDefaultParams(). But I'm not sure how you calculated the
> RingDiameter, because instead of 150 * 2 you used 157*2. Should I had
> some extra mm to my RingDiameter too? As my Gate Scanner is different
> from SherPET, I'm also not sure which number of ring i should put,
> because i have 4 rings, but each ring has a block of 13*13 repeated
> around the ring. Should I put 4, or should I put 13*4=52?
> In the functions, FindCoincDetectorAdd and FindSingleDetectorAdd, if
> it is a cylindrical PET (6 columns for volumes) i leave it like that,
> but if it is a ECAT , i should change the number 6 to the number of
> columns of the volumes. Do you think i need to change something else?
>
> I also want to say that I can help in testing for some other data that
> I have acess, such as Siemens scanners or GATE simulations. I'm more
> than glad to help.
>
> Thank you so much for any help that you can give me.
>
> Best Regards,
> Liliana Caldeira
> PhD student
> University of Lisbon
> Institute of Biophysics and Biomedical Engineering
>
>
> ________________________________
> Internet Explorer 8 makes surfing easier. Get it now!