Detector intersection area

12 views
Skip to first unread message

Vytautas Gradauskas

unread,
Nov 4, 2025, 5:42:43 AMNov 4
to mcx-users
Hello prof. Fang,

Is there a function in MCX that calculates the "active" intersection (between zero and non-zero label media) area of a detector?

Thank you
Vytas

Qianqian Fang

unread,
Nov 7, 2025, 11:28:16 PMNov 7
to mcx-...@googlegroups.com, Vytautas Gradauskas

hi Vytas,

no, mcx does not have such function. it is actually not easy to do - tessellating 3D points in to an open surface is surprisingly difficult problem.

over the past two days, I tried a few approaches, including modifying a CRUST algorithm, or using alphaShape() provided by matlab, the results were unsatisfactory. also, neither of them work for co-planar (2D) point cloud.

for example, running the colin27 built-in example

https://github.com/fangq/mcx/blob/master/mcxlab/examples/demo_colin27_atlas.m

you can see the point cloud detected at the 4 detectors

by calling a modified crust function, I can get something following the shape, but there are a lot of holes - with some places missing big patches. alphaShape() failed in different ways.

in the attached figure, top shows the detected photon positions for the 4 detectors in the colin27 example; the bottom panels showed the crust-recovered surface mesh for det#1 and det#2. both of them showed partial recovery, thus calculating area is not accurate


Qianqian

--
You received this message because you are subscribed to the Google Groups "mcx-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mcx-users+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/mcx-users/1103f28d-304a-44fd-9faa-3fe05953033fn%40googlegroups.com.
detpt_tess.png
Reply all
Reply to author
Forward
0 new messages