Photon emission time

[Keyao Zhang]

Dear Dr. Fang,

I have two questions to consult you:

1. Is there a time difference in the emission of all photons? Are they all emitted at the start of the simulation? Or is it limited by hardware and the calculation of each photon is carried out gradually according to performance? Or can the emission interval be adjusted through parameter configuration?
2. If I want to simulate the propagation of femtosecond (fs) pulsed light in the brain, how should I configure tstep and tend? My understanding is that tstep and tend can be both set to 10^-8 (seconds) to ensure photons basically complete their propagation trajectories. Then, the energy value generated by the fs light pulse can be simulated by calculating an appropriate number of photons.

Thank you very much for your time and help!

Best regards,
keyao zhang

[Qianqian Fang]

hi Keyao,

1. Is there a time difference in the emission of all photons? Are they all emitted at the start of the simulation? Or is it limited by hardware and the calculation of each photon is carried out gradually according to performance? Or can the emission interval be adjusted through parameter configuration?

the propagation time (f->t) and accummulative path-length (f->pathlen) of each photon packet is reset to 0 at the time of a successful launch

because mcx/mmc simulates the impulse response (i.e. TPSF) of a photon packet, all photon packets are expected to be launched exactly at t=0; if you are dealing with temporal profiles of a source, you should use a temporal convolution of your emission temporal profile wiht the generated TPSF

1. If I want to simulate the propagation of femtosecond (fs) pulsed light in the brain, how should I configure tstep and tend? My understanding is that tstep and tend can be both set to 10^-8 (seconds) to ensure photons basically complete their propagation trajectories. Then, the energy value generated by the fs light pulse can be simulated by calculating an appropriate number of photons.

for best practices of deciding the tend/tstep settings, please check out our training tutorial (Jupyter notebook)

https://colab.research.google.com/drive/1MX4rC1my2znSYXCj8HgKmPEZzNxhZii3#scrollTo=JXabl344Y4Wa

specifically, the below section

Best Practics Guide - maximum time window

The T1 value should be decided based on the domain physical size, the scattering coefficient of the medium and the lossiness (absorption) of the medium. It is not necessarily setting the large the T1 value will bring more accurate steady-state photon distribution.

We often use a maximum simulation time window of 5 ns in our examples - in water (n=1.33) or tissue-like medium (n=1.37), this allows light to travel ~1.1 meter total path length. In tissue-like medium, due to extremely high scattering, this time should allow light to diffuse and fill a ~10 cm^3 volume and obtain a steady-state solution. One can adjust the time-window length according to the medium domain and scattering setting you have - but more than 5-10 ns may not bring much benefit in absorbing medium because photon packet weight drops exponentially along the path. Letting a low-weight photon to continue propagation results in little change to the light fluence profile while consuming computational time.

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/9f0d58a5-344d-45a7-8105-a6928aeff0c9n%40googlegroups.com.