[gate-users] Image Reconstruction of Gate output
Mutanga Theodore
muathe02 at student.umu.se
Wed Jan 5 16:55:39 CET 2005
Hi Sadek,
I noticed that STIR has the two written in the code but it seems by
default the have 223. Actually both are mentioned.
How far did you go with the testing of the DST code ? I was trying to
figure out at least why I get those segmentation faults, Kris suggested I
recompile STIR in debug mode and rerun FBP to try to locate where the code
is crashing.This I'll do tonight.
Hope to hear from you soon.
Theo
> Theo and All,
>
> On the issue of the SINOGRAM_WIDTH for the DST, I believe it should be
> 249 (not 223). Below is an example of a DST scan as it is described in
> the investigater tool:
>
> radialCompression: 1
> axialResoTest*radialCompression: 249
> *thetaCompression: 1
> axialResoTest*thetaCompression: 210
>
> Kris, is there a reason why you're using 223 in STIR (as Theo reported
> in a previous email), rather than 249? Or, I am missing smth? Sorry, we
> don't have a DST, so I can't investigate this more.
>
> Cheers,
>
> Sadek
>
>
>
> -----Original Message-----
> From: gate-users-bounces at lphe1pet1.epfl.ch
> [mailto:gate-users-bounces at lphe1pet1.epfl.ch] On Behalf Of Mutanga
> Theodore
> Sent: Tuesday, January 04, 2005 10:12 AM
> To: Long ZHANG; GATE feedback and helpline for Users
> Subject: Re:RE: RE: [gate-users] Image Reconstruction of Gate output
>
>
> Hello Sadek,
>
> sorry I am bothering you again but I'd like to show you the changes I
> made to your code:
>
> 1) GEAdvance3D_Sinogram_Bin.c
>
> #define NUMBER_OF_RINGS 24 // Advance/DLS)
> #define CRYSTALS_PER_RING 420 // Discov$#define
> SINOGRAM_WIDTH 249 // (or 223 according to STIR ??)
> #define MAX_THETA 11 //
> #define NUMBER_OF_PROJECTION_PLANES 553 //
> #define FOV 600 //
> #define SLICES_PER_FOV 47 //
>
>
>
> 2)
>
> // Update the different arrays...
> //-------------------------------
>
>
>
> if (u>=0 && u<223) <<<< I changed this to 223 or 249 ? !!!!
> {
> sinogram_counts[ring2][ring1] += 1;
>
>
>
>
>
>
> 3) Gate_to_Stir.c
>
> int nRings = 24; // Number of rings in the GE Advance
> scanner
> int Num_sinos = nRings*nRings;
> int nPHI = 210; // Number of angular bins in the
> sinogram
> int nU = 223; // Number of radial bins in the
> sinogram
>
>
>
> 4)These slices per segment I got from the headers created by
> creat_proj_data
>
> int num_slices_per_segment[23] =
> {1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,47}; //Number
> of projection planes per segment
>
>
> 5)
>
> printf("Please enter the absoute number of segments you would like to
> reconstruct (0, -> 22):
>
>
> 6)
>
> //Process segments 23->2
> //-----------------------
> printf("Processing segments 23 -> 2... \n");
> for (ring1= (num_segments+1); ring1>1; ring1--)
> {
>
>
> 7) I didn't know how to change the 17 ? what does it do here ?
> r1 = ring1;
> while (r1<17)
> {
> r1 = ring1 + c;
> r2 = r1 - ring1;
>
>
>
> 8)
> //goto NEXT2;
>
> //Process segments -2->-23
> //------------------------
> printf("Processing segments 2 -> 23... \n");
> for (ring2=2; ring2< (num_segments + 2); ring2++)
> {
> int c = 0;
> r2 = ring2;
>
>
> Well Those are the things I modified, but I think there is a great deal
> more to change because when I use the manip_proj_data from Stir on the
> projection data, it doesn't look good at all.
>
> Hope to hear from you soon
>
> Theodore
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>> Dear Sadek,
>>
>> That's great if you can implement the DST code. It looks almost
> finished.
>>
>> My idea is to build a generic framework based on your code. With this
>> framework,
>> the coding work will be greatly simplified for new scanner. It's a
>> abstracted
>> layer which can call different set of functions for a different
> scanner.
>> It's
>> pretty important when porting it to GATE, for not only the
> correctness,
>> but also
>> the optimised design.
>>
>> So it's very important to know root script characters, such as
> callback
>> functions,
>> or object oriented characters like dynamic binding, etc. Or I think a
>> input
>> parameter file or a static table is needed (like STIR). I only know
> root
>> is C/C++
>> like, but no deeper study.
>>
>> Helps is highly appreciated!
>>
>> Regards,
>>
>> Long
>>
>>
>>
>>
>>
>> In your mail:
>>>From:"Nehmeh, Sadek/Medical Physics" <nehmehs at mskcc.org>
>>>Reply-To: GATE feedback and helpline for Users
>>> <gate-users at lphe1pet1.epfl.ch>
>>>To: "GATE feedback and helpline for Users"
> <gate-users at lphe1pet1.epfl.ch>
>>>Subject: RE: RE: [gate-users] Image Reconstruction of Gate output
>>>
>>>Dear Theodore and Long,
>>>
>>>Yes I'm back, and sorry for taking me time replying to your emails.
>>>Long, I think what you have suggested in your previous email regarding
>>>evaluating the sinogram binner for the DST, and try to incorporate it
>>>with GATE, is a great idea. Meanwhile, I am also testing that for the
>>>DST configuration. Hopefully I will get back to you tomorrow with my
>>>feedback.
>>>
>>>Theodore, for your NEMA measurements, I know O. Mawlawi at M.D.
> Anderson
>>>did the NEMA2001 measurements for the DST. It was published in the JNM
>>>(J Nucl Med. 2004 Oct;45(10):1734-42); you may wana refer to that, and
>>>reproduce their analysis method, for comparison.
>>>
>>>
>>> Cheers,
>>>
>>> Sadek
>>>
>>>
>>>
>>>Theodore
>>>Sent: Monday, January 03, 2005 11:30 AM
>>>To: Long ZHANG; GATE feedback and helpline for Users
>>>Subject: Re:RE: [gate-users] Image Reconstruction of Gate output
>>>
>>>Hi Long,
>>>it seems Sadek is back, have you written to him about the code ?
>>>
>>>Cheers
>>>
>>>Theodore
>>>
>>>By the way I was simulating the ECAT exact hr+ and I would like to
>>>compare my results with someone Here is what I had for the resolution
>>>measurements , NEMA 2001 !
>>>
>>>transverse @ 1cm = 8.3 mm
>>>axial @1 cm 11.5 mm
>>>
>>>
>>>transverse radial @ 10cm = 10.59mm
>>>transverse tangential@ 10cm = 10.66mm
>>>
>>>axial @10cm 10.31 mm
>>>
>>>
>>>I wish to find out from the GATE users if these look reasonable, I was
>>>planning to do measurements myself on the ecat but I dont have acceess
>>>to one since initially I was simulating the DST. Also in the NEMA 2001
>>>there is no mention about matrix size for reconstruction , doesnt this
>>>affect the resolution ? I did not correct for anythin in the data from
>>>GATE , would it be better to correct for randoms, scatter ect , if so
>>>HOW (using GATE ?)?
>>>
>>>
>>>Cheers
>>>
>>>'Theo
>>>
>>>> Dear Kris,
>>>>
>>>> Thank you very much.
>>>>
>>>> It seems that I have mistake the sinogram width (angle position) as
>>>> sinogram length (view number, or tangential bin number, or
>>>> displacement position).
>>>> Am I
>>>> right now?
>>>>
>>>> Regards,
>>>>
>>>> Long
>>>>
>>>> In your mail:
>>>>>From:"Kris Thielemans" <kris.thielemans at csc.mrc.ac.uk>
>>>>>Reply-To: GATE feedback and helpline for Users
>>>>><gate-users at lphe1pet1.epfl.ch>
>>>>>To: "'Long ZHANG'" <zhanglong99 at tsinghua.org.cn>,
>>>> "'GATE feedback and helplinefor Users'"
>>>> <gate-users at lphe1pet1.epfl.ch>
>>>>>Subject: RE: [gate-users] Image Reconstruction of Gate output
>>>>>
>>>>>>
>>>>>> BTW, I was pussled by the sinogram width of DST (420 crystal per
>>>>>> ring). The number 249 is larger than the 420/2 = 210. How can DST
>>>>>> get the extra LORs of 39? By interpolation? Could this help to
>>>>>> improve the spatial resolution? Thank you!
>>>>>>
>>>>>Notation: N= num_detectors per ring
>>>>>
>>>>>Why do you think that the max sinogram width is N /2? This is a
>>>>>hardware choice (recent CTI scanners tend to have
> sinogram_width=N/2).
>>>
>>>>>The 'physical' maximum of LORs in 2D sinograms is (N^2-N)/2 (i.e.
> all
>>>>>lines between pairs of detectors, excluding the degenerate cases,
> and
>>>>>identifying det pairs ij and ji). The max number of views is N/2
>>>>>(you'd expect N, but it's halved because of interleaving). So, the
> max
>>>
>>>>>sinogram width is something like (N-1)
>>>>>
>>>>>
>>>>>Kris
>>>>>
>>>>>_______________________________________________
>>>>>gate-users mailing list
>>>>>gate-users at lphe1pet1.epfl.ch
>>>>>http://lphe1pet1.epfl.ch/mailman/listinfo/gate-users
>>>>>
>>>>
>>>>
>>>> _______________________________________________
>>>> gate-users mailing list
>>>> gate-users at lphe1pet1.epfl.ch
>>>> http://lphe1pet1.epfl.ch/mailman/listinfo/gate-users
>>>>
>>>
>>>_______________________________________________
>>>gate-users mailing list
>>>gate-users at lphe1pet1.epfl.ch
>>>http://lphe1pet1.epfl.ch/mailman/listinfo/gate-users
>>>
>>>
>>>
>>>
>>>
>>>
> =====================================================================
>>>
>>> Please note that this e-mail and any files transmitted with it
> may
>>> be
>>> privileged, confidential, and protected from disclosure under
>>> applicable law. If the reader of this message is not the intended
>>> recipient, or an employee or agent responsible for delivering
> this
>>> message to the intended recipient, you are hereby notified that
> any
>>> reading, dissemination, distribution, copying, or other use of
> this
>>> communication or any of its attachments is strictly prohibited.
> If
>>> you have received this communication in error, please notify the
>>> sender immediately by replying to this message and deleting this
>>> message, any attachments, and all copies and backups from your
>>> computer.
>>>
>>>
>>>_______________________________________________
>>>gate-users mailing list
>>>gate-users at lphe1pet1.epfl.ch
>>>http://lphe1pet1.epfl.ch/mailman/listinfo/gate-users
>>>
>>
>>
>> _______________________________________________
>> gate-users mailing list
>> gate-users at lphe1pet1.epfl.ch
>> http://lphe1pet1.epfl.ch/mailman/listinfo/gate-users
>>
>
> _______________________________________________
> gate-users mailing list
> gate-users at lphe1pet1.epfl.ch
> http://lphe1pet1.epfl.ch/mailman/listinfo/gate-users
>
>
>
>
>
> _______________________________________________
> gate-users mailing list
> gate-users at lphe1pet1.epfl.ch
> http://lphe1pet1.epfl.ch/mailman/listinfo/gate-users
>
More information about the Gate-users
mailing list