Reflectance Calculation in PMCX
Shiru Wang
Qianqian Fang
hi Shiru,
see my replies below
Hi Dr.Fang,
I hope this email finds you well. First, I wanted to express my appreciation for PMCX - it's proving to be a valuable tool for our work.
I'm encountering some challenges while using PMCX to simulate and calculate time-resolved surface reflectance (backscattered photons) of a geometry using a planar source. After reviewing the intensity curve over time in my simulation results, I have concerns about potential issues in my approach. I would greatly appreciate your guidance on two specific questions:1. Where the source should be located? Should the source start at the air layer above the defined volume or should it start at the top of the volume?
either way is ok - when a photon is launched outside of the
volume (we call it a wide-field source), mcx performs an initial
propagation to advance it to the first non-zero voxel, then start
propagating it based on the optical properties. The space outside
of the volume, as well as 0-valued voxels are defined as
cfg.prop(1,:), which is by default mua=0/mus=0/g=1/n=1, simulating
air.
If your source is collimated, pulling it back along the launch
direction won't change the final solution, although it adds a
slight overhead due to the initial ray-tracing.
I do want to add two details
1. cfg.voidtime flag controls whether the photon's travel time
includes the time traveling before entering the first non-zero
voxel, by default (cfg.voidtime=1), the timer starts when a photon
is launched; if set to 0, the timer starts when it enters the
first non-zero voxel
https://github.com/fangq/mcx/blob/v2024.2/mcxlab/mcxlab.m#L270
2. cfg.isspecular flag controls whether a photon's initial weight
is reduced when entering from air to the first non-zero voxel, by
default (cfg.isspecular=0) does not reduce the initial weight
(i.e. w0=1 when entering the volume), if set to 0, it reduces
based on 1-specular-reflection = transmission using the
angle/optical properties of the interface of the voxel.
2. Could you clarify how to determine the photon termination point? Specifically, I'm uncertain whether to look for a label value of 0 or optical property [0 0 1 1].
a photon is terminated by one of the 4 conditions
1. moving from a non-zero voxel to a zero-voxel
2. transmitting through the exterior bounding box of the volume, and
3. total travel time exceeds cfg.tend
4. if cfg.minenergy is defined, a photon is terminated if it does
not survive the Russian roulette after its weight drops below
cfg.minenergy
for your question, it uses the label instead of optical properties to terminate. this is why when you simulate a cavity, you should set the voxels in the cavity to a non-zero label with air properties instead of setting it to 0.
let me know if this is clear
Qianqian
It would be great if there's a demo that we could look into. Hope to get your reply and thanks for the help!
Best,Shiru
--
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/15e8f0d9-3ea3-4325-90c5-e696d164c4b6n%40googlegroups.com.
Qianqian Fang
hi Shiru,
the diffuse reflectance output is accumulated by tracking photon packets that escape from the desired surface, there is no need to manually track those.
unless you need to estimate something that is exiting-angle-dependent, then you should save the photon's exiting direction using the cfg.bc flag, and do this calculation manually. otherwise, diffuse reflectance (accumulating escape weights regardless of exiting direction) is sufficient for most purposes.
by the way, diffuse reflectance can only be saved in the 0-valued voxels immediately outside of the object boundary (non-zero valued voxels), so you must pad a layer of zeros outside of the surface where you want to save it. see example code
https://github.com/fangq/mcx/blob/v2025/mcxlab/mcxlab.m#L401-L402
reading your code, I don't fully understand why you set
vol[:,:,0]=3, and read the output from Data[:,:,1]
Qianqian
Hi Dr.Fang,Thank you for your detailed guidance on the previous discussion. I've been implementing the photon simulation using pmcx following your instructions, and I have another question regarding the surface reflectance calculation.
Currently, I'm estimating the surface reflectance by collecting photon data from the top layer of my geometry. I'm not sure if this could be considered as a reasonable estimation. I wanted to ask your opinion on this approach and whether there might be a more accurate method, such as specifically tracking upward-traveling photons. Attached is my code for the simulation.
Hope to get your reply. Thanks so much for your time and consideration!
Best,
Shiru
To view this discussion visit https://groups.google.com/d/msgid/mcx-users/f0efc1ae-17f4-430a-bc5a-85dec049724dn%40googlegroups.com.
saiei yu
Qianqian Fang
Dear Saiwei,
mcx-users has nearly 400 subscribers from all over the world.
While I perfectly fine reading your questions in Chinese, out of
the respect to all other subscribed users, please ONLY post
questions in English, so everyone understands what you asked
about.
If you feel that you can not express your question clearly in English, please feel free to use Google Translate, or email me individually.
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/b45ab0b3-38fa-45e9-83c1-22fa0e0b7effn%40googlegroups.com.
saiei yu
Dear Professor Fang,
First of all, I sincerely apologize. This is my first time posting here, and I am not yet familiar with all the rules. I accidentally posted a letter in Chinese, which may have caused you some inconvenience. I am truly sorry for that.
Additionally, I really appreciate your patience in responding to my question. I will carefully read your reply and follow your guidance to resolve the issue.
Thank you once again.
Best regards,
saiwei
Qianqian Fang
no worries, I hope my response to your question earlier was
helpful - feel free to follow up if you have any additional
questions.
To view this discussion visit https://groups.google.com/d/msgid/mcx-users/ecd758e3-399e-4957-81b3-8fdac3692e00n%40googlegroups.com.