Can The Time Activity Curve (TAC) function use bi-eksponensial model?

[fajarmiraz]

Hello everyone,
I am working with a voxelized source in GATE v10 and defining its time-activity curve (TAC). I have successfully used a mono-exponential function as shown in the documentation. However, I am now trying to implement a bi-exponential function, and the simulation fails to run.

My goal is to model an uptake and washout phase using a function of the form:
Activity (t) = A0 * (1 - exp(-k_in * t)) * exp(-k_out * t)

Here is the relevant part of my script:

TAC Parametric

A0 = 16330 * Bq
lambda1 = 5.75871 * sec
lambda2 = 1.24335 * sec
k_in = 1 / lambda1
k_out = 1 / lambda2
times = np.linspace(1, 10, num=500, endpoint=True) * sec
activities = [A0 * (1 - np.exp(-k_in * t)) * np.exp(-k_out * t) for t in times]
source.tac_times = times
source.tac_activities = activities

The simulation runs without error when I use a simple exponential decay, but it fails with this bi-exponential TAC.
My questions are:
Does GATE v10 natively support bi-exponential or more complex TAC functions for voxelized sources?
If yes, could there be an issue with my implementation? For example, are there constraints on the tac_activities values (e.g., must be non-negative, must not start at zero)?
If it is not supported, what is the recommended workaround? Would I need to simulate multiple time frames with constant sources instead?

Any insight or guidance from your experience would be greatly appreciated.
Thank you in advance for your help.