mri_reface News

mri_reface 0.3.3

Christopher Schwarz — Tue, 21 Nov 2023 2:14:18 GMT

We've released mri_reface 0.3.3 with minor enhancements and bugfixes. This fixes behavior when input .nii files have inconsistent sform vs qform geometry, automatically splits dynamic PET images to allow registration across frames, and applies some minor bugfixes to nii2dicom. Please see the changelog for more details. We have provided this release in both Docker and non-Docker formats.

mri_reface 0.3.2 released

Christopher Schwarz — Tue, 02 May 2023 1:39:25 GMT

We have released our latest version of mri_reface, version 0.3.2.

Starting with this release, we are also releasing mri_reface as complete Docker image, to simplify and standardize running mri_reface in many environments. This has been our most frequent feature request. Docker releases are provided in a separate download, and the original release type will also remain available. This release also updates to use matlab runtime 2022a, so please update your installations. We also now recommend using niftyreg (reg_aladin) >1.5, due to reports of poorer registration performance with 1.3. Other smaller improvements include some tweaks for improved intensity normalization for 2D FLAIR images and improvements for images with larger slice thickness. The primary run script now also includes a DICOM->DICOM workflow suitable for some limited use cases. For a complete list of changes, please see CHANGES.txt.

mri_reface 0.3 released

Christopher Schwarz — Fri, 15 Jul 2022 2:32:59 GMT

We've released our new version 0.3 of mri_reface, for replacing faces in MRI, PET, and CT images.
Major additions include: support for PET/CT (see our new paper: https://doi.org/10.1016/j.neuroimage.2022.119357), support for multi-frame images, support for simultaneous re-facing of multiple images, matching of noise/PSF when re-refacing, and the ability to produce re-faced DICOM via our included ADIR_nii2dicom.

First public release, v0.2

Christopher Schwarz — Tue, 29 Jun 2021 1:57:23 GMT

mri_reface replaces identifiable face information in MRI scans to help prevent potential re-identification via face recognition. Faces are replaced with an average face, rather than removed, to better resemble a natural image and reduce effects on downstream brain measurement software. See our comparison/validation paper in NeuroImage: https://www.sciencedirect.com/science/article/pii/S1053811921001221. This version has several major new enhancements as well.

This software is under heavy active development and we suggest you subscribe to receive notifications of new releases. Current support is for T1 T2 and FLAIR MRI. We are working on additional modalities, along with open source and containerized releases.