Work

QSM-CI method: iQFM (PyTorch) (v2)

Year: 2026

Type: software

Abstract: iQFM — instant tissue (local) field mapping from raw MRI phase using Laplacian-enabled deep neural networks (LoT-Unet). It is the tissue-field output of the same network/repo as iQSM: a single deep network maps raw wrapped phase directly to the local field, performing phase unwrapping and background-field removal in one pass (the iQSM submission runs the same repo with --no-iqfm to skip exactly this output). Multi-echo phase is reconstructed per echo and combined with magnitude x TE^2 weighting. Output tissue field is B0-normalized (ppm). Runs CPU-only with a CPU-only PyTorch wheel; pretrained weights (HuggingFace sunhongfu/iQSM) are baked into the image. QSM-CI reconstruction method iqfm. Browse and run it at https://qsmxt.github.io/QSM-CI/submission.html?method=iqfm.

Source: Zenodo (CERN European Organization for Nuclear Research)

DOI: 10.5281/zenodo.22091306

Authors: Gao Y., Xiong Z., Fazlollahi A., Nestor P.J., Vegh V., Nasrallah F., Winter C., Pike G.B., Crozier S., Liu F., Sun H.

Language: English

Topic: Advanced MRI Techniques and Applications

Subfield: Radiology, Nuclear Medicine and Imaging

Field: Medicine

Domain: Health Sciences

Open Access status: green

OpenAlex ID: W7204184813

Cites:

Cited by:

Related to: 10