[gate-users] PET simulations duration

[Corinne J Groiselle]

Hi,

I am running GATE on an AMD Opteron (64-bits) 1.4 GHz cluster with 2 Go 
of RAM per dual-processor.

To have a clinically realistic simulation of an F-18 acquisition (@ NEC 
peak - 2 min scan), it takes 15h to 24h depending on the size of the 
object on 120 CPUs.

For example, I have ~24 million total coincidences in a 70-cm long, 
20-cm diameter cylinder filled with water and 5.8 mCi of F-18 for a 
2-min acquisition (the simulation takes 15h x 120 CPUs).

Hope this helps,
Corinne

pasantos at bwh.harvard.edu wrote:

>Dear all,
>
>I am trying to decide if I should use GATE to simulate PET acquisitions
>with I-124 but, before I can actually install GATE and try it myself, I
>would need to have an idea of the duration of a simulation either with
>F-18 or I-124.
>
>Could you experienced users please tell me what kind of workstations you
>are using and how much time it takes to run a simulation of a determined
>number of events considering simple object geometries?
>
>I would really appreciate all the information you could provide me with.
>
>Thank you in advance,
>Paula
>
>
>
>  
>

-- 

Corinne J. Groiselle, PhD
Research Instructor

Department of Radiology / Division of Nuclear Medicine
University of Massachusetts Medical School
55 Lake Avenue North
Worcester, MA 01655
USA

_Phone_: (508)-856-6123 - fax: (508)-856-4572
_Email_: Corinne.Groiselle at umassmed.edu _
Professional homepage_: http://users.umassmed.edu/corinne.groiselle
Official UMASS faculty page: 
http://www.umassmed.edu/faculty/show.cfm?faculty=731
_Personal homepage_: http://corinne.groiselle.free.fr


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opengatecollaboration.org/mailman/private/gate-users/attachments/20041203/d806e247/attachment.htm>