Skip to content

MPAT Transition Probabilities

Tag: mpat_transitions
Category: Tractography

Builds MPAT's absorbing Markov chain (voxel-to-26-neighbour transition probabilities) by running mpat.trxviz_bridge in a Python environment with MPAT installed. Reads either a transitions ODX written by mpat --save-transitions odx, which loads in seconds, or a plain fODF ODX, which also needs tissue maps (a pyAFQ output folder, MPAT's desc-mpat_dseg, or a wired Transient WM Mask). The chain is cached as mpat_transitions_<hash>.npz and reused while the inputs and settings stay the same. See the MPAT guide.

Ports

Direction # Kind
Input 0 ODF Field (optional)
Input 1 Transient WM Mask
Input 2 Absorbing Mask
Output 0 Transitions
Output 1 CoDI (Double-ODF)
Output 2 CoDI (Single-ODF)
Output 3 CoAsy (Asymmetry)
Output 4 P(Terminate Double)
Output 5 P(Terminate Single)
Output 6 Total ODF Amplitude
Output 7 KL Divergence (Single)
Output 8 Chain Dead-State Prob

Parameters

Field Default
angle_deg 35.0
curv_sigma_deg 0.0
double_odf true
fod_power 1.0
fod_thresh 0.05000000074505806
max_length_mm 250.0
preset "Standard"
probe_radius_vox 0.0
python_path ""
step_vox 1.0
stop_mass_tau 0.0
wm_threshold 0.4000000059604645
working_dir ""

Citations

This op runs the original software rather than a TRXViz re-implementation. When you use it in a published workflow, please cite:


  1. John Kruper, Matthew Cieslak, and Ariel Rokem. Probabilistic mapping of brain connections without random sampling. bioRxiv, 2026. URL: https://www.biorxiv.org/content/10.64898/2026.11.12.766276v1, doi:10.64898/2026.11.12.766276. ↩