Module relaxometry 

R2/T2 mapping from multi-echo spin-echo (EPG), and R2’ = R2* − R2.

EPG-based fitting models imperfect refocusing (B1 < 1) so it removes the stimulated-echo bias that a mono-exponential fit suffers. [r2prime] combines the spin-echo R2 with a gradient-echo R2* (from r2star) for chi-separation, and returns the region R2 was measured over alongside the map — a spin-echo acquisition is often a slab rather than the whole head, and R2’ is undefined outside it.

With no spin-echo acquisition there is no R2 to subtract; [r2primenet] predicts R2’ from the GRE-derived R2* with a trained network (onnx feature), which is what lets the R2’-consuming separation methods run on GRE-only data.

Structs

R2EpgParams

Parameters for r2_epg dictionary matching.

R2PrimeMap

R2’ and the region over which it is a measurement.

R2PrimeNetNorm

Training z-score constants for R2PRIMEnet (xsepnet_train_patch_norm_factor_inplane_largedegree_romeo_arlo.mat): each field is (mean, std), in Dr-scaled (ppm-equivalent) units. The input is normalized (r2star/dr - mean)/std; the output de-normalized (y*std + mean)*dr.

R2PrimeNetParams

Inference parameters for r2primenet.

Constants

AUTHORS_PATCH

The authors’ patch: (D, H, W) = (x, y, z).

B0_TOLERANCE_T

How far from TRAINED_B0_T an acquisition may sit and still be accepted, in Tesla.

EPG_TE_TOLERANCE_S

How far, in seconds, r2_epg’s echo times may be from TEₙ = n·ESP (each spacing from the first, and the first echo from the spacing): 0.1 ms, which covers the rounding of echo times in DICOM and BIDS sidecars.

POOL_MULTIPLE

Four pooling levels, so every patch dimension must be a multiple of this. Built without onnx, like the other limits here, so a host can check a patch without the networks.

TRAINED_B0_T

Field strength the published R2PRIMEnet weights were trained at, in Tesla (Kim 2025).

WASM_PATCH

A patch that fits a 32-bit (WASM) heap; see the module docs for what it costs.

Functions

epg_cpmg_echoes

Simulate the CPMG multi-echo spin-echo magnitude train.

r2_epg

R2 mapping from multi-echo spin-echo magnitude via EPG dictionary matching.

r2prime

Compute R2’ = R2* − R2, clamped at zero, over the region where R2 was measured.

r2primenet

Run R2PRIMEnet: predict R2′ (Hz) from an R2* map (Hz).

r2primenet_from_magnitude

Convenience: fit R2* from multi-echo magnitude with ARLO, then run r2primenet.